The Reality of a Pick and Place Lifestyle

[rx8pilot]

I wanted to start a thread about the general process of PCB assembly. There are quite a few topics about specific machines, solder paste, ovens, etc.....but none that discuss the big picture reality of owning and effectively operating a pick and place line.

I started my pick and place experience on eBay a few years back and the details of the story are thick and rich. The learning curve has been intense. The overall benefit is hard to put on paper. Sometimes I love it and sometimes I want to sink the whole system in the middle of the Pacific Ocean. It occurred to me that my personal experience is full of problems and challenges that are likely self-inflicted. I purchased a used/broken machine with zero documentation. I had no previous experience.

The earliest lessons were that the pick and place machine is only a fraction of the process. The process starts in the PCB design phase - assigning part numbers, creating an accurate BOM, ordering the right parts, organizing the right parts, making sure the parts are packaged in a way appropriate for the limitations of the machine.

Then I had the learning experience of solder paste, stencils, ovens, hand placed parts, etc. Solder paste is a non-trivial decision, stencil design is important, stencil printers can wreck your day, printing small single PCB's is easy - large fine pitch panels is another story.

All that babble to say this: I have no idea if I am doing it well. I have no idea if I am efficient. I have no idea if all of it is worth it. At any given moment - I feel like the system is never good enough. I am always short something like feeders, or something needs adjustment, or whatever. My board designs keep getting harder and harder to assemble. Smaller parts, higher density, double sided = zero process forgiveness.

I am very curious if anyone is willing to share successes or failures in the realm of PCB assembly.

[rx8pilot]

Today's party:

Just need to run a quick batch of 16pcs of a single design.
2 of the coveted precision feeders that can handle 0402/0201 parts were acting up. Each needed about an hour for fiddle and fix.
4 parts are in 4 different size tubes,, but my vibe feeder does not have the right plates - so I have to hand place those fine pitch parts by hand.
2 electrolytic caps would not fit in the feeders and I did not want to figure out why - so I hand placed those as well.
I had a bunch of cut tapes of 0402, diodes, smd fuses. The 0402 parts made a mess when I had the nozzle bumping them during pickup. Took a while to get that sorted.
This could go on for quite a while.....but in the end.....these PCB's ended up taking a very long time. It took all day when I planned for 3 hours. Hell - the 1st one took 3hrs all by itself. The last one took about 10 minutes.

A gazillon mispicks, errors, etc, etc. I would swear that my machine has a personality and it can sense my fear and frustration.

Perhaps a total of 500 curse words screamed at the top of my lungs and all is well - PCB's assembled on a holiday when all assembly houses are closed.

Sure hope this is worth it, because it appears to be taking years off my life.

I need to improve part organization, solder paste printing, feeder maintenance, machine maintenance, make or buy a vibe feeder that can handle any size tubes, and spend a whole day just doing experiments to understand how to reduce the number of missed pickups. The lack of consistent performance is deadly for a small operation. I never know what to expect.

[EEVblog]

We have discussed this a lot on the AmpHour and also many places scattered across the forum.
The general consensus is that there is a quite narrow windows of usefulness for a personal PnP machine.

Mike's Electric Stuff is a classic example. He does short-ish runs of boards in short lead times that very often use thousands of LED's and other parts repeated on the board, so he derived great benefit from having his own PnP for those uses. Most people will not fit into that category though.

[RoGeorge]

All ears and wanting to hear more, please go on.

[mikeselectricstuff]

4 parts are in 4 different size tubes,, but my vibe feeder does not have the right plates - so I have to hand place those fine pitch parts by hand.

Cut the top of the tube away for 1 part length, so the parts are picked from the end of the tube - no need for any part-specificness on the vib feeder.
 

[mikeselectricstuff]

My number one tip is to take the time to write software to make the process of getting from your PCB software to the P&P machine as quick and error-free as possible. This would include such elements as :

Have fiducial components in your library, so fid positions get set automatically
Have all your PCB library parts with consistent orientation with respect to taping, so rotations just work correctly without needing to check.
Physical differences (height and vision parameters) mean you need to have different PCB library parts for the same nominal package, e.g. for 0603 you may have resistor, thin cap, thick cap, ferrite bead, LED etc. and each may have different parameters for vision, speed,height etc.
 
For parts that can come in tape or tube (Mostly SO ICs), you need seperate library parts as the orientations are different.

Dealing with panelisation - step/repeat, intelligent handling of fids ( PCB may have fids per circuit or per panel, so detect the outermost and only pass these to the P&P system)

Other things :
By far the biggest factor in getting good assemblies is stencil print quality. You need a tensioned stainless stencil which is accurately flat against the whole PCB surface.

Have as many feeders as possible - loading parts into feeders is takes the  longest time in setup ( if other things take longer you're doing it wrong).
Buy jellybean parts in full reels and keep them in the feeders.
 
It is almost never worthwhile to machine place a single board, unless it has more than about 100 parts, or most parts are the same.
With a foot operated vacuum pen and tape holders you can typically manually place at 10-30cpm

Unless it really matters and someone is actually going to check, use leaded solder. It has a much bigger margin between "reflow" and "incinerate"

[mikeselectricstuff]

If you think you can build a useful pick and  place machine you are almost certainly wrong.
If you think you can do it for less than $1000 you are definitely wrong
More so if you start by making an XY gantry. The hardest part is the picking, not the placing. Feeders are everything. A P&P without reel feeders is a toy.

A reliable pick & place process ( the machine is only part of the process) is not one big problem but a a huge collection of small problems, any of which can result in lost productivity, wastes parts & trashed boards
 
If you want a fun project, go for it but treat it as an educational experience which is unlikely to result in a useful tool. If you need a cheap machine, go buy a used dinosaur or Chinese machine and be aware it will take time and tweaking to make it useful.

Any pick and place machine, at any price, will take you a few weeks to get a process and running well. If you need some boards made quickly, go to a subcontractor
 
One example of the sort of issue you can hit  that immediately comes to mind : I had a crazy timescale project - ten boards with about 2500 LEDs, plus firmware etc. to play videos from a MicroSD card. Timescale : about a week and a half, from nothing.
 
This didn't particularly faze me - done this sort of thing loads of times before - thought I had at least a day contingency.

However the only LEDs that anyone had enough stock of at a reasonable price came on really thin tape. When this went through my feeder, about half the time, the LED would jump out of the tape when it was advanced.
It took most of my contingency day to modify the feeder to the point that I only lost about 5%, fortunately Digikey had enough extra stock to cover this and there was enough  time to get another reel to make up the wasted parts

[mikeselectricstuff]

And another one - LEDs with soft silicone tops, sticking to the nozzle ( my machine doesn't have blow-off)
https://twitter.com/mikelectricstuf/status/994172467232018432

[DerekG]

I wanted to start a thread ..... but none that discuss the big picture reality of owning and effectively operating a pick and place line.

Excellent idea.

At any given moment - I feel like the system is never good enough. I am always short something like feeders, or something needs adjustment, or whatever. My board designs keep getting harder and harder to assemble. Smaller parts, higher density, double sided = zero process forgiveness.

For all these reasons, I have decided to outsource our SMD assembly.

But, I'm always wanting to know if I have made the right decision.

[Smallsmt]

Have all your PCB library parts with consistent orientation with respect to taping, so rotations just work correctly without needing to check.

Funny thing but the feeder setup solve this problem!
Because if you have feeders on different side of the machine this can't help, especially if you use feeder cassettes.
You need to define the angle deviation inside feeder setup.

[Smallsmt]

And another one - LEDs with soft silicone tops, sticking to the nozzle ( my machine doesn't have blow-off)
https://twitter.com/mikelectricstuf/status/994172467232018432

Blow off is normal on our pnp machines but sometimes a special teflon nozzle solve this problem too.
We use another parameter to close vacuum in advance on Z axis to push the part to paste surface using the blow off.

[Smallsmt]

I wanted to start a thread ..... but none that discuss the big picture reality of owning and effectively operating a pick and place line.

Excellent idea.

At any given moment - I feel like the system is never good enough. I am always short something like feeders, or something needs adjustment, or whatever. My board designs keep getting harder and harder to assemble. Smaller parts, higher density, double sided = zero process forgiveness.

For all these reasons, I have decided to outsource our SMD assembly.

But, I'm always wanting to know if I have made the right decision.

If you always want to cover 100% of part placement it is a hard job.

In our prototype assembly I found out it it helps a lot if machine place the standard parts already on feeder so manual assembly for some special parts took not too much time.

Definitely the stencil printing and paste type are most important for good results.
And cheap bend PCB's cause a lot of problems too.
We use mostly gold plated pads on our pcb's because it's more reliable if you store the PCB over longer time.

One of the biggest problems is to care the right feeder setup! If you put a part in wrong orientation  or wrong part in feeder you receive a lot of rework.

Mostly I do a first PCB PNP run and check everything after finished. Then start the remaining pcb's.

In one PNP job I received 68 wrong placed TQFP100 because of 180 degree turned JEDEC tray. It took 2 days to repair the boards.

[Kjelt]

Excellent stories and such bad experiences. 

Just need to run a quick batch of 16pcs of a single design.

Is this not the kind of job where you and Mike each and everytime in a topic say you do NOT use an P&P for ? 

It took all day when I planned for 3 hours. Hell - the 1st one took 3hrs all by itself. The last one took about 10 minutes. 

Exactly and the last 100000 will take each a couple of minutes, you have the quircks ironed out and are ready for massproduction.
Oh wait you only had 16 boards. Perhaps it was quicker to do it manually as you say so many times yourself.

For these kind of jobs you don't need a P&P machine made for massproduction that runs ultrafast.
You were better off with an intelligent machine that can visually check it picked the component and orientation, then checks the placement after.
Takes time yes but who cares with <50 boards.

However the only LEDs that anyone had enough stock of at a reasonable price came on really thin tape.
When this went through my feeder, about half the time, the LED would jump out of the tape when it was advanced.

For modern P&P machines there are probably feeders suited for this kind of tape ?
You were lucky you could modify the feeder in time, I think no other party would have accepted this contract in that limited amount of time.

[mikeselectricstuff]

Have all your PCB library parts with consistent orientation with respect to taping, so rotations just work correctly without needing to check.

Funny thing but the feeder setup solve this problem!
Because if you have feeders on different side of the machine this can't help, especially if you use feeder cassettes.

Only if the p&p software is so dumb that it doesn't know where the feeders are, so it can automatically apply an appropriate rotation.
if you move a feeder, the only thing you should need to do is tell the machine where it is - the machine should know to do any rotation to compensate.

[mikeselectricstuff]

And another one - LEDs with soft silicone tops, sticking to the nozzle ( my machine doesn't have blow-off)
https://twitter.com/mikelectricstuf/status/994172467232018432

Blow off is normal on our pnp machines but sometimes a special teflon nozzle solve this problem too.
We use another parameter to close vacuum in advance on Z axis to push the part to paste surface using the blow off.

My machine went obsolete long before silicone LEDs were a thing. if it was a major issue I'd mod it to blow, or make teflon nozzles, buf for now a spray of PTFE lube spray on the nozzle every so often fixes it.

[mikeselectricstuff]

Excellent stories and such bad experiences. 

Just need to run a quick batch of 16pcs of a single design.

Is this not the kind of job where you and Mike each and everytime in a topic say you do NOT use an P&P for ? 

Depends entirely on the job. Once you have a good process, a single panel can be worthwhile but you'd typically hand place anything there was only ohne or two of per circuit, unless it was already in a feeder.
Subbing a single panel gets expensive due to their setup, and the fact that (IME) subcontractors always insist on using full-sized stencils that fit their machine, despite the fact that a small, cheap frameless stencil is just fine for 1-off jobs

However the only LEDs that anyone had enough stock of at a reasonable price came on really thin tape.
When this went through my feeder, about half the time, the LED would jump out of the tape when it was advanced.

For modern P&P machines there are probably feeders suited for this kind of tape ?

Quite probably. The feeder design on mine is maybe 25+ years old and far from ideal. It mostly works, most of the time, and doesn;t need too much tweaking. if I used my machine mopre than a few times a month I'm sure I'd have moddied it to use some Yamaha or simlar feeders

[rx8pilot]

Just need to run a quick batch of 16pcs of a single design.

Is this not the kind of job where you and Mike each and everytime in a topic say you do NOT use an P&P for ? 

[/quote]

Like everything in the P&P world...it depends.
In this particular case the PCB has about 125 placements, about 30 unique 0402 passives among other parts. In addition, this is the third time I have run these boards so the setup was almost complete. I think I had to swap just a few parts in feeders to get started.
Even after the programming was done and proven on previous runs and the feeders all loaded.....all the stupid problems still ate my lunch. One feeder was getting stuck, another feeder had an alignment issue. The Z-pickup heights all needed some adjustment that I was not expecting. The laser scanner that measures and orients the parts on the nozzle had a smudge on it which increased the reject rate even though the parts were perfectly positioned. On that one, I was chasing it down in software thinking the part profile was messed up. The lens of the scanner is very hard to see and it was the last thing that came to mind - not even sure how it would get dirty. Perhaps I manually changed a nozzle at some point - who knows.

I did the initial prototype of this design by hand which took about 4 hours or. That included all the time needed to double check part values, re-calculate things that did not 'seem right' just to make sure, etc. My guess is that it would by about 90 minutes each on average if I manually assembled these. After all the delays and difficulties with the P&P - the average time was about 30min/each. Considering that 30min is the disaster scenario for P&P - that is not the end of the world.

While this setup is a sad story, I do have some good ones for my other designs. The designs I do the most of (and keep all the feeders loaded for) are easy to run just a handful of boards without much trouble. Boards that took 60 minutes on the best of days with my manual setup are done in about 5-6 minutes on the P&P. That is pretty good, but in fairness, it took many hours to get it all working like that.

For these kind of jobs you don't need a P&P machine made for massproduction that runs ultrafast.

Totally true. My machine has an IPC rating of 3600CPH but I never run it that fast. I estimate that when I am at 'full speed', the machine is only placing 1,000CPH. The whole thing is limited by the processes before and after the machine and I am the only one doing it. If I had an 8 head machine doing 30,000CPH it would not make much difference unless I also got a fully automatic printer, a production oven, an AOI machine. A LOT of my time is burned dealing with cut tapes and only having a few spare parts available. One of my designs has 3 chips that are about $20-$30/each. I cannot just let those get dumped in the bin and keep guessing. I have go fetch the part, put it back in place until the adjustments are made. Ceramic capacitors are getting into this range as well - many of the caps in my power stuff are $.75-$1 each so I have only a small number of spares and have chase down any bad picks. A small issue will put $100 of parts in the dump bin pretty quick.

And another one - LEDs with soft silicone tops, sticking to the nozzle ( my machine doesn't have blow-off)

For the first month - I had no idea my machine had blow-off. Many of the parts were sticking to the nozzles just enough to shift them out of position during the release. It took me a while to figure out that blow-off was built in and I only needed to connect compressed air.

Is this not the kind of job where you and Mike each and everytime in a topic say you do NOT use an P&P for ? 

Depends entirely on the job. Once you have a good process, a single panel can be worthwhile but you'd typically hand place anything there was only ohne or two of per circuit, unless it was already in a feeder.
Subbing a single panel gets expensive due to their setup, and the fact that (IME) subcontractors always insist on using full-sized stencils that fit their machine, despite the fact that a small, cheap frameless stencil is just fine for 1-off jobs

My decision to get into P&P was dominated by deadlines that CM's or other assembly options could not accommodate. For the times that the schedule technically permits - it is still a rush job and priced accordingly. Even if I was willing to pay the price, I still have to organize the project data precisely and kit all the parts and get them delivered to the assembly house. That is, in many cases, half the work of making a P&P run if not more. On top of that....I still have the stress worrying about component rotation or some other simple error that will throw the project into a tailspin. At least when I run my own PCB's, I can take the first one to the bench, inspect and power it up before I run the rest of them.

[mikeselectricstuff]

My decision to get into P&P was dominated by deadlines that CM's or other assembly options could not accommodate.

Me too. And it is not just simply timescales. Doing it in-house means a lot less time documenting things - I can sometimes have a whole job assembled in-house in less time than it takes prepping everything for an assembly house, waiting for quotes, dealing with queries etc.

 

[rx8pilot]

4 parts are in 4 different size tubes,, but my vibe feeder does not have the right plates - so I have to hand place those fine pitch parts by hand.

Cut the top of the tube away for 1 part length, so the parts are picked from the end of the tube - no need for any part-specificness on the vib feeder.

I sorta hacked something like this together a number of months ago - and it sort-of worked. It seems I am fighting a poor design overall. The parts were getting stuck or they were vibrating right out of the feeder. There is an adjustment for frequency and amplitude, but it is tricky to set it. My guess is that a fairly simple machining effort should allow some improvement. My latest effort was laser cutting some thin plastic sheets to guide the parts. That was helpful, but still not quite reliable enough for prime time. SOIC-8 parts are pretty good, but DFN-10 are not. More fiddling ahead.......

My number one tip is to take the time to write software to make the process of getting from your PCB software to the P&P machine as quick and error-free as possible.

This was, bar none, the biggest single process improvement I have made! It wasn't even all that difficult relative to other areas of improvement.

What this forced me to do is create and enforce a part numbering scheme that covers the whole process from design to delivery. I built a spreadsheet (all I had at the time) to create the part numbers based on pre-defined criteria. That number is used in the CAD software, purchasing, storage bins, feeders, P&P software, etc. When I have completed a design, I can one-click the output with the part number X-Y and Theta to the P&P and that is it. I have it setup to do top and bottom of the PCB in separate files so there is no programming penalty for double-sided designs. HUGE time saver.

The part numbering scheme was designed so that when it is sorted - it naturally groups part families together and in sequence of values.

Example:
E20-040-0960
E='ELECTRONIC COMPONENT'
20= Resistor
-040= SIZE 0402
-0960 = 10k (part number has no relation to the real value)

during prototyping or maintenance, I can just scan through the bins quickly to get any part I need. At the moment, I think I have about 500 unique parts in use and I can find them very fast. I can look at a feeder and know what is in it. I can look at a DigiKey invoice and know what the part is and what products it is used in. It seems so simple, but without a logic part numbering and organizing system, I would spend a lot of time looking and guessing.

The next major upgrade would be developing a database - Filemaker is what I am looking at. The excel spreadsheet has long been outgrown and is very difficult to maintain.

Have fiducial components in your library, so fid positions get set automatically

I forgot to put fiducials on my last design!!! What a pain. The latest Eagle version allows user-defined design templates which I think I will use to avoid forgetting in the future.
Initially I thought I could use some vias or pads - but that did not work very well. All the vias had connected traces or they were stitching vias. The connected vias are no longer a circle. The stitching vias had no copper edges - just the solder mask which is not nearly accurate enough.

I had to manually align each board and verify it before any placements. Probably only a minute per PCB, but that seems like an eternity when it is usually totally automatic and takes 1 second.

In one PNP job I received 68 wrong placed TQFP100 because of 180 degree turned JEDEC tray. It took 2 days to repair the boards.

Oooof! That is my nightmare scenario.

Re-work can undo the efficiency of the fastest pick and place system in the world. The faster your machine is, the faster it can break things.

[mikeselectricstuff]

The part numbering scheme was designed so that when it is sorted - it naturally groups part families together and in sequence of values.

Example:
E20-040-0960
E='ELECTRONIC COMPONENT'
20= Resistor
-040= SIZE 0402
-0960 = 10k (part number has no relation to the real value)

This may not work for everyone, but I always treat part types and values seperately - e.g. SMALL0603R is the PCB footprint and P&P machine package type for vision & height, "1K" is the part value. ("SMALL" is a historical artifact)
My machine also treats them seperately, so you don't end up with ridiculous numbres of part types. The part type determines dimensions and vision params, the value is the label it gives the feeder you loaded it into

[rx8pilot]

The part numbering scheme was designed so that when it is sorted - it naturally groups part families together and in sequence of values.

Example:
E20-040-0960
E='ELECTRONIC COMPONENT'
20= Resistor
-040= SIZE 0402
-0960 = 10k (part number has no relation to the real value)

This may not work for everyone, but I always treat part types and values seperately - e.g. SMALL0603R is the PCB footprint and P&P machine package type for vision & height, "1K" is the part value. ("SMALL" is a historical artifact)
My machine also treats them seperately, so you don't end up with ridiculous numbres of part types. The part type determines dimensions and vision params, the value is the label it gives the feeder you loaded it into

Without a doubt.....part number schemes are like underwear. You don't want someone else's underwear. :-)

I made of list of personal goals for my part numbering system (not important to everyone)
1. Fixed length if possible
2. Sorting the list automatically groups similar parts
3. Easy to human read with modest amount of memorization
4. Syntactically easy to use in the various systems at play
5. Scalable beyond any current expectation

My machine is kind of similar - but requires a unique part number for each part. You start with a defined package. When you add a part, you have to pick the package which determines the feeder, nozzles, dimensions, etc. In the end....I now have a very long list of parts in the P&P that match the part numbers in the rest of the system. The length of the list is not much of a challenge since I never really directly use the list. When I import a new design - it queries the list and imports the data. For new parts, it gives me a list and I add them as needed. I can also add them with an offline utility as well.

I have 'generic' parts too, but so far have not used them. By the time I am assembling the PCB, I have been forced into creating the new part numbers already for purchasing and storage.


On the topic of solder paste......

I used leaded solder for quite a while. As Mike pointed out, it is easier to use and a lot less likely to melt stuff. As I started using smaller parts, I moved to a Type 5 solder paste Loctite GC10. It is un-leaded and I have to be much more precise with the oven profile. Does anyone know of a Type 5 leaded paste? I have no requirements to use lead-free at the moment, and would love to have Type 5 leaded. Easy to print and easy to re-flow.

[Kjelt]

Type5 is called AGS or AF5  from Multicore usually sold in syringes AFAIK because the syringe needles require the smallest particle size to pass.
But they should be sold in jars as well, can not find the leaded type 1,2,3 seems like the entire industry went Rohs.

[jmelson]

I have a Philips CSM84 (made by Yamaha).  It is quite old-school, no vision, chuck alignment jaws to center small parts, and a mechanical alignment station that centers large parts on the nozzle by bringing the part over to that station.

It is built like a tank, though.  I've had a few problems with it, but it mostly just keeps on running.  But, without vision, the centering is not so great for fine-pitch parts.
It will do SOIC parts with perfect accuracy, but .65 mm quad packs are a little iffy, and .5 mm definitely have to be adjusted by hand.

I recently did just TWO boards for a customer, but they had about 400 components each, and I did have the whole setup already saved.

I wrote a little C program that takes the placement file from my CAD package and converts it for the P&P machine.

Jon

[jmelson]

I forgot to put fiducials on my last design!!! What a pain. The latest Eagle version allows user-defined design templates which I think I will use to avoid forgetting in the future.
Initially I thought I could use some vias or pads

My Philips CSM84 can use pads as fiducials.  I'm not sure they are as accurate as a fiducial pad, but it works pretty well.  There is a setting for normal or reverse fiducial.

Jon

[rx8pilot]

I forgot to put fiducials on my last design!!! What a pain. The latest Eagle version allows user-defined design templates which I think I will use to avoid forgetting in the future.
Initially I thought I could use some vias or pads

My Philips CSM84 can use pads as fiducials.  I'm not sure they are as accurate as a fiducial pad, but it works pretty well.  There is a setting for normal or reverse fiducial.

Jon

My machine is supposed to do it, but rarely does. The worst part is that it will get a false position and not throw an error - then 'correct' all the positions into oblivion. It becomes a big mess pretty quick.
Lesson learned - fiducials are just as important as food and water, lol.

[rx8pilot]

For 0402/0201 placement.....which Yamaha feeders are the best?

I wonder if it is possible to make an adapter or modification for my Quad. The precision Quad feeders are hard to find and very expensive.

[rx8pilot]

We have discussed this a lot on the AmpHour and also many places scattered across the forum.
The general consensus is that there is a quite narrow windows of usefulness for a personal PnP machine.

Mike's Electric Stuff is a classic example. He does short-ish runs of boards in short lead times that very often use thousands of LED's and other parts repeated on the board, so he derived great benefit from having his own PnP for those uses. Most people will not fit into that category though.

Yep - here and there. I am curious how many people have small systems, how many people think they need a small system, why they think they need a small system, what kind of learning curves and issues do people face in reality, etc, etc.

I think my operation (or hopeful operation) is a good fit for having a system in-house. I will freely admit that I was very, very naive about what it actually would take to pull it off. Even after many hours of reading and a decade of owning/operating a CNC machine shop - nothing really prepared me for the avalanche of details. At this moment, I don't regret it - but that has not always been the case. There have been many times that I was kicking myself. At the same time, I could never figure out how to finish my work any other way.

If I did not get the P&P, I think I would have needed to totally re-think my small biz. When I first began to look for machines - I started at the bottom. I had little to no money so the cheap Chinese no-vision machines were on my radar. I could easily see the limitations right away but at the time - my boards were almost all 0805 passives, SOIC-8's, and other easy parts. I was thinking that for $5k I could get the machine, printer, and cheap oven and it would at least be a step up from 100% manual assembly. At this time, I had pretty much maxed out my manual assembly setup as much as possible without buying some new hardware. I forgot the model number, but I was only a few days away from ordering one of the ultra-low-cost machines from China.

That was when I had a conversation here on the EEVBlog about used machines including the Quad IV-C. I had scoured the internet for information about them and generally informed myself about the system. 1990's design, multiple ownerships, big, heavy, reliable, were all take-aways from the initial look. The scary part was that they were sn MS-DOS machine made with parts and software that was long ago obsolete and discontinued. Then I learned that a company in PA had purchased the Quad business and refurbished them with Windows 7 and all new software around 2009/10 or so. I called them and got a quote - the machine and a modest batch of feeders was MSRP $50k. Maybe haggling could get me to $40k-ish.

Not possible. I did not have the $$. Back to the Chinese hobby machine.....until.......eBay. A listing popped up with very little details. Quad IV-C with Feeders BUY IT NOW $5,000 about 60 miles from my house. I was in bed about to fall asleep and scanning eBay. I saw this and asked my wife what she thinks...she said go for it. So I clicked the button with zero understanding of what I was committing myself to. I used two credit cards to cover it, rented a trailer and went to pick it up. It was extremely dirty with junk piled all over. The PC had broken off it's rack rails. But I noticed real fast that this was one with the Windows 7 upgrade. There were about 50 feeders. All the parts were there.

Another machine that was the MS-DOS version was sitting next to it. The guy that purchased that one was asking if he could also buy the one that I just purchased. I answered no and loaded it on the trailer and took it home. There is a thread (I will try to get the link) of the disaster I just purchased, but for this conversation, it is important to know that it took about 6 months and another few $k to get the first part placed.

It was, perhaps, the most difficult possible entry into pick and place. I am trying to forget that part of the experience.

Skip to present day - I am VERY happy I never had to mess with the limited use desktop hobby machines but with that said - the used Quad has not been a picnic in any way. It fights me at every opportunity even if I win every time. I have been able to go from 0805 to 0603 to 0402 and I already have my sights set on 0201 (which the machine technically has the capability of doing if everything is completely perfect). Total cash in the machine is about the same as a Neoden 4, total time invested.....too much time.

If the N4 was available when I got my machine - It would have been high on the list of consideration.

Type5 is called AGS or AF5  from Multicore usually sold in syringes AFAIK because the syringe needles require the smallest particle size to pass.
But they should be sold in jars as well, can not find the leaded type 1,2,3 seems like the entire industry went Rohs.

Yep - the market for leaded solder is probably shrinking. That may be especially true for a type 5 ball that is usually on the finer and fancier PCB's. Still looking.

[jonnod]

Referring to rx8pilot’s original post - what a great post and topic.   In fact it has prompted me to make my first ever post.  I don’t post things because of time pressure – why – because I too have set up my own small in-house smt line so don’t have any time!  But I can feel rx8pilots pain and I share it! 
Hey, looking over my post, I see it is very long – but it is a difficult subject and I hope it helps others who may be considering setting up their own SMT line?

I have had a very similar experience to rx8pilot.  I decided to install my own smt line after terrible experiences with local board loaders. Also I didn’t want to get my boards done in China due to risks of IP loss and part substitution.   20 months on and I still can’t say if it was a good idea or not. 

I also totally underestimated the huge learning curve, yeah, by a factor of 1,000 I think!
I purchased brand new items.  Stencil printer, SmallSMT P&P VP2800HP, Reflow Oven Puhui T-937.   Each item represents a continual source of frustration and difficulty, but of course the P&P is the most complex item and therefore the biggest hassle.

I also had to modify my entire design process from the schematic up to match the smt line.  A great deal of time and effort is required to ensure that the P&P file is correct for the particular P&P machine.   

Fine pitch parts like the hideous DFN/QFN, VSOP, leadless 0402, BGA present great difficulty for manual stencil printing & placement.   It can take hours to align the stencil printer and check the print quality.   Solder choice is vital for print quality.  I have to use leadfree solder and after a great deal of research chose the very expensive GC10 paste.  Even so, the solder needs to be at the right temperature, squeegee pressure & angle “just right” to print properly and I often have to readjust the stencil alignment which is a difficult and time consuming process (only req’d if fine pitch parts are used).

The oven was a major hassle.  Leadfree means it is very easy to burn the boards & kill parts!  There is far less margin than with leaded soldering.   The oven’s own temperature measurement is wildly inaccurate and it takes days of careful checking and calibration to get the solder profile correct.  This must be done using the target PCB and adjusted for each different PCB if the size and thermal load changes.   Fortunately I did chose an oven that allowed custom temperature profiles to be created on a PC and uploaded.  (The software supplied is not good!)  My first new purchase would be a better oven but I’m now wary of manufacturers claims.  I also only have single phase power available.

I have learnt that the choice of component package and the packaging – tape & reel vs. tube vs. tray is vital, plus one is often faced with huge MOQ’s that can be very expensive.
Cut tapes and loose parts are a major hassle and to be avoided.

Then there is the the biggest problem of them all that dwarf’s all the other problems.   The Pick and Place machine itself.   What a nightmare.   
Firstly the good bits:  Not long ago, P&P machines were way to expensive to contemplate for small business.  Modern software, imaging and manufacturing have reduced the price to a reasonable level.   
The software supplied with my machine is surprisingly good  - despite the rather obtuse menus due to language and translation issues.   The support from the manufacturer is also very good, but here again language is a big barrier.  (I need to learn Chinese and German!).  It can place most standard parts accurately and consistently but even this takes continual monitoring and adjustment.
Now the not so good bits that keep me awake at night and want me to just grab it an throw the machine into the Pacific Ocean right on top of rx8pilot’s machine… It took about 9 months to get the machine working to an acceptable level to place my first board.  There were a number of manufacturing faults and breakdowns.  (yes, you get what you pay for).   The initial set-up was very difficult, a very steep learning curve, it took months.    Sure, one could place a few 0603 chip resistors pretty quickly on the demo board but my boards use a lot of different parts and it is this that is still an issue, almost 2 years later.

The feeders are a nightmare on the machine I have.  Each one has to be hand adjusted via fiddly springs and if the tape changes, it has to be done again.  The 8mm wide tape feeders only took weeks to get working, but the 12, 16 and 24mm tape feeders are a huge problem.
The single biggest problem is that the feeders on each bank interact.  Get one feeder working and others stop working.   Months go by working on this.   Cassette type feeders probably work  better but they are more expensive.
Tall parts are a problem for most of the entry level P&P machines.  Anything over 7mm presents a collision risk and involves time consuming work-arounds and P&P file manipulation.
Forget about speed and claimed parts-per-hour. in practice these are rubbish.  Continual stoppages and breakdowns are the ultimate arbiter of machine performance.  (I have only placed a handful of PCB’s without the machine stopping and requiring intervention, approx 0.01%).  I have to stand and operate the machine continually, or employ someone to do so, including hand manipulation of the wider tapes to try to prevent stoppages.   This is not something I had ever considered in my original decision to go “in house smt”.   I had not factored the cost paying a wage for an operator to stand by the machine while it ran just to produce 2-3 boards per hour.  (180 parts per board).

*Sigh*  Would I do it all over again?  Probably not.  I would spend a fraction of the time vetting and working closely with a local board loader who has a “real” SMT line.   But now, I’ve put about $30k into my line and for now, I need to recoup some of that money by making boards.

Hope this is of some assistance.

[Dubbie]

This is fascinating reading. Someone a couple of years ago asked my advice about buying their own pick and place. I said “no way, don’t do it”. They did it. I’m not sure they have ever made a single successful run of boards from the thing. This thread vindicates my advice somewhat. Thanks to those who have honestly documented their experiences here.

[Kjelt]

IMO, I think it would be very helpfull for anyone wanting to setup his own SMT line to spent one day in an actual SMT assembly house from scratch to final product.
I have done this for my work a couple of times and I will never forget the panic when the fastest flying probe tester went down. They had to deliver 20000 boards the end of the week and this tester was essential for guaranteeing quality. The same evening an engineer was flown in from the manufacturer from another country, he repaired the tester at nightime and at 7.00 production could continue. I don't know what the bill was they sent but it was high 4 figures for sure.
The manager told me they would not make much profit on that run that week.

Just looking at all the people working there from the logistics planning to the operational persons, to people manually inserting some large components, the different type of machines from stencil, P&P, reflowovens, testers, flowsoldering baths, programmers and stickering.
Still 4 persons that work full time to correct boards manually or try to find why some defects occurr.
It was an ant hill, beautifull but I can not understand someone would really think he could do this all by himself esp. under time pressure.

 

[mikeselectricstuff]

Something to note is that having your own P&P for only your own jobs is a very different situation from being a subcontract manufacturer doing work for multiple clients, all with their own systems, requirements, quirks etc.

Something that you absiolutely must not even consider is paying for the machine by running jobs for other people. That;s a whole different ballgame.
 
For people regularly doing small runs of different boards, doing it in-house with a low-cost (<$20K) setup is definitely worth considering.

If you regularly do runs of the same job, or have higher volumes,where placement time is a significant limiting factor,a subcontractor will almost certainly be a better option.   

[rx8pilot]

This is fascinating reading. Someone a couple of years ago asked my advice about buying their own pick and place. I said “no way, don’t do it”. They did it. I’m not sure they have ever made a single successful run of boards from the thing. This thread vindicates my advice somewhat. Thanks to those who have honestly documented their experiences here.

With all the pain and suffering....I still don't think I could have what I have today without the P&P. The more I use it, the better I get. The better I get, the more my little business grows. I REALLY REALLY REALLY wish it was easier....but no pain no gain I guess. My direct competitors all have slower time to market and higher costs overall. The difficult learning curve has earned me a competitive advantage in the end.

Something to note is that having your own P&P for only your own jobs is a very different situation from being a subcontract manufacturer doing work for multiple clients, all with their own systems, requirements, quirks etc.

Something that you absiolutely must not even consider is paying for the machine by running jobs for other people. That;s a whole different ballgame.

 

Seriously correct. A few people have asked if I can help them assemble. The answer is so easy.....NO. No way. Impossible. Unless they buy out my entire business.
It is, like Mike stated, a totally different universe trying to make boards for other people. I have my little internal system that closely matches my designs - and even that is a very big challenge.

I would rather have my machine sit powered down for months than try to 'keep it busy'.

IMO, I think it would be very helpfull for anyone wanting to setup his own SMT line to spent one day in an actual SMT assembly house from scratch to final product.

There is some wisdom there, but only if you are setting up a commercial operation. A small in-house system that only does in-house designs is at least a more simplified task. Manual printing is fine, no need for AOI or flying probes, etc. I am not sure I know of any good ways to gain an understanding of the PCB assembly reality without actually doing it. That is the majority reason for this thread. Hopefully not a depository of complaints, but something that will paint a realistic picture of what it is like. There are a lot of discussions that focus on machines and specification when the most important thing to focus on is the WHOLE process.


My personal progression into to P&P:
Started with Weller WES51 soldering station, manual solder paste syringe. Manually placed with tweezers, soldered each joint with the iron.
EFFECTIVENESS: Totally F'ing stupid

Found a small batch oven and started ordering prototype stencils. Still placing with tweezers from a box full of parts.
EFFECTIVENESS: Got through some simple prototypes

Added a vacuum pen and fabricated some cut tape holders to organize parts. The vacuum pen is still in use today and I love it. Tweezers are rotten. Also got an Amscope stereo microscope.
EFFECTIVENESS: This really sped up the process and reduced the mental anguish. Now I could do pilot runs of small designs. 10-20 pcs before I went crazy. The microscope immediately revealed tiny bridges and other flaws that were causing problems.

Added a stencil printer, color-coded part trays, and color-coded assembly guides that I printed and had as PDF on a computer monitor.
EFFECTIVENESS: This allowed more complicated designs to be hand assembled. The stencil printer is a piece of junk, but far better than the loose prototype stencil taped to my bench. The parts organization was a huge boost for speed and accuracy. This is where I achieved my max speed assembling manually. At the time, I had (3) double-sided boards that needed to be assembled. I did them in batches of 5-8 at a time

NOTES: Other important gear I picked up along the way that was important - pneumatic syringe dispenser for solder paste and SMD adhesives, hot air machine, upgraded the Weller WES51 to a JBC (life changer), a solid variety of vacuum pen heads.

This is when I purchased the P&P machine but continued to manually assemble PCB for many months (right next to the machine). As the machine slowly came to life.....it was clear I had to take all my previous tools and knowledge and create a system of sorts. I was now in a position where I was more process/skills limited than hardware/tools limited. I did not log the time it took to fix and modify the machine. I also have no real idea how much learning curve I have put into all of this. I do know that I would fail miserably trying to manually assemble my current designs even with my low volumes. I have too many small and fine pitch parts.


Becuase of the path I chose, I have very little experience sending boards out for assembly. The limited experience I do have showed me that it is not easy or fast or cheap.

[Dubbie]

These guys had none of the aspects that made the endeavour worthwhile in the end for you.

They made very short runs. Maybe 6 boards in total of a design. They had simple boards, maybe 30 components each. And lastly they only had the budget for the crappiest Chinese PnP with useless feeders.

This is fascinating reading. Someone a couple of years ago asked my advice about buying their own pick and place. I said “no way, don’t do it”. They did it. I’m not sure they have ever made a single successful run of boards from the thing. This thread vindicates my advice somewhat. Thanks to those who have honestly documented their experiences here.

With all the pain and suffering....I still don't think I could have what I have today without the P&P. The more I use it, the better I get. The better I get, the more my little business grows. I REALLY REALLY REALLY wish it was easier....but no pain no gain I guess. My direct competitors all have slower time to market and higher costs overall. The difficult learning curve has earned me a competitive advantage in the end.

[jmelson]

Yep - here and there. I am curious how many people have small systems, how many people think they need a small system, why they think they need a small system, what kind of learning curves and issues do people face in reality, etc, etc.

Well, I got my Philips CSM84 in 2007, and was using it in production in less than 2 weeks.  I have never even SEEN a P&P machine before, but I had retrofitted CNC to a manual mill.  There was some learning curve, but not real bad.  There were a few error messages that were not in the book, so I have to invent what the cause was and what to do about it.  The biggest part of the learning curve was coming up with how to size solder paste apertures to get good soldering with minimal bridging.  I have that pretty well figured out now.

I have done over 1000 boards on my machine.  I do small batches of some fairly complicated boards.  By small batches, I'm talking about anywhere from 10 to 60 boards at a time.  Contract manufacturers will charge a LOT for such short runs, as their setup time is the same for 1 or 1000.  Most of my boards have parts on both sides, and can have up to a few hundred parts/board.

My Philips CSM84 is not a "small system", it is 5 x 7 feet and weights 700+ kg.  It cost the previous owner over $100K.  We have about 15 different board designs we do on it.  My machine was running production less than 2 weeks before it was shipped to me.  The seller was getting 2 brand-new machines, and had to get the old ones off his floor.  So, I didn't have anything to fix when I got it.

Jon

[jmelson]

I also had to modify my entire design process from the schematic up to match the smt line.  A great deal of time and effort is required to ensure that the P&P file is correct for the particular P&P machine.   

strangely, I did not have to do this.  The first boards I built on the P&P were designed for manual assembly, and were fairly low-density.  So, I went straight from manual assembly to P&P on the very same designs.  As time went on, I did a few things differently to accommodate the P&P better, but it was not a big deal.

Fine pitch parts like the hideous DFN/QFN, VSOP, leadless 0402, BGA present great difficulty for manual stencil printing & placement.

Yes, the leadless chip-scale parts are a nightmare!  I did ONE board with some CSP components and had dozens of shorts/board.
I will stay with lead-ed parts from now on, so I can inspect and rework any shorts.

  It can take hours to align the stencil printer and check the print quality.   Solder choice is vital for print quality.  I have to use leadfree solder and after a great deal of research chose the very expensive GC10 paste.  Even so, the solder needs to be at the right temperature, squeegee pressure & angle “just right” to print properly and I often have to readjust the stencil alignment which is a difficult and time consuming process (only req’d if fine pitch parts are used).

The oven was a major hassle.  Leadfree means it is very easy to burn the boards & kill parts!  There is far less margin than with leaded soldering.   The oven’s own temperature measurement is wildly inaccurate and it takes days of careful checking and calibration to get the solder profile correct.

I print by hand, with a hand-aligned stencil, and it works pretty well.  Yes, there are good and bad solders.  I have to get some of this magic GC10 stuff.  The Indium Corp. solder I got recently does not perform well, the paste just smears out a couple minutes after printing and causes shorts.

As for the oven, I use a converted GE toaster oven, but the trick is to poke the thermocouple into a PTH on the board, so the controller measures ACTUAL board temperature.  This works great, but I still get a darkening of the silkscreen, and sometimes even a slight darkening of the whole board.

Cut tapes and loose parts are a major hassle and to be avoided.

I buy whole reels of generic R's and C's, but for specialty parts (big value or HV caps, for instance) and expensive ICs I buy cut tape.  The Yamaha feeders on the Philips really handle cut tape very well.  You only lose a couple parts at the beginning, and can use all parts right to the end of the tape.

  It can place most standard parts accurately and consistently but even this takes continual monitoring and adjustment.

I have had to make some adjustments to the nozzle offsets and other alignment values on my machine.  But, after that, I have never "adjusted" my placement files to correct for the machine.  I run a program that I wrote that pulls the XYR from the CAD file and formats it for the P&P machine, and load that file into the machine.

When I went to doing a board with 0603 parts, I had to set up a crosshair and adjust the feeders so that the pick-up location was more accurate and consistent.  I found that dome-top LEDs and 0603 LEDs would not pick up well.  But, 0805 LEDs and up, as long as they have a flat top, work fine.  Yes, feeders are "fiddly", but I do not have to ADJUST them for each tape.  I do have a problem where 0.1uF caps have thicker tape, and that tends to stick a bit and not advance.  I put a couple C-clamps on the tail of the tape and that helps it pull through the feeder.

The Philips CSM84 can only handle parts up to 6.5 mm height, so that is a limitation.  I have had a few breakdowns with it, but I managed to fix them all.  A sensor went out and it would randomly throw completed boards off the end of the conveyor onto the floor!  The rotation motor's commutator was packed with copper dust and shorting out, causing faults.  Sometimes the vacuum generator/valves get contaminated with oil and you get weak vacuum, and you have to tear them down for cleaning.

But, all in all, I can't IMAGINE running my business at the level it is now without this machine!  I've built 1000+ boards, and mounted well over 100K parts with it.

Jon

[DerekG]

I have too many small and fine pitch parts.

Yes, I learnt this very early on when using contract SMD assembly shops.

Each component loaded costs money & fine pitch parts cost more to load than non-fine pitch.

Working with the assembly shop during the design process can save a lot of money. This includes optimal pad apertures & solder paste stencil thickness & even ensuring larger SMD components do not shade the smaller ones.

Oh ......... and reducing the parts count by designing in micros etc can save you lots.

With the continual drive for more miniaturisation from the component manufacturers themselves, part footprint selection becomes more challenging as each month goes on.

[grimmjaw]

Interesting thread..

Never owned/operate P&P myself but being struggling with the process.
Typically I have a board with >800 parts, >90 unique part, and always have subcon the assembly . Cost about 1500 USD and 3-4 weeks lead time.
Always wonder if cheaper & faster to buy a P&P,, at least for prototype



 

[rx8pilot]

Interesting thread..

Never owned/operate P&P myself but being struggling with the process.
Typically I have a board with >800 parts, >90 unique part, and always have subcon the assembly . Cost about 1500 USD and 3-4 weeks lead time.
Always wonder if cheaper & faster to buy a P&P,, at least for prototype

$1,500 for a single prototype?
Do you have to kit the parts for the assembler or do they source the parts?

When my work was dominated by prototypes - I seriously considered one of the manual machines. Some of the ones I saw were pretty clever and would make a board like yours a practical effort. Perhaps not something that you are excited about, but something you could do in 4 hours instead of 4 weeks. I was able to sustain about 400-500CPH average when I was organized and had all the parts laid out nicely in holders and bins. In your case - factor in the prep work, paste printing, restroom breaks, etc....4hrs is reasonable. 5 hours would be safe.

Of course, if you needed 10.....yikes. A P&P may be helpful if the learning curve has not scared you away.

The 3-4 week lead time for outsourcing would have me looking for a very tall bridge. I have never had that much time for a prototype.

[xaxaxa]

I generally outsource production runs and hand assemble prototypes; the last board I did has ~350 parts, ~1200 pads, and 56 BOM lines. The assembly for 50pcs costed $800 total, which is $16 per board. Lead time is about 1 month.

The same design takes me about half a day to assemble using stencil + paste and tweezer for pick and place. I have custom software that will display one BOM line at a time and a picture of the pcb with all the locations for that component highlighted which makes it much easier; most of the time is spent gathering the parts rather than PnP, so a manual PnP machine wouldn't really give me any benefit.

Nowadays for prototypes I use a low volume PCBA service that will solder just the passives, so I only need to solder the chips and connectors. The cost is about $60 for the design I mentioned earlier for 5pcs and this includes PCB fabrication (10x10 cm, 4 layers), PCBA, and parts (passives). With this there is no reason for me to consider getting a PnP system.

[Gary.M]

Tell me more about your custom software?

I have custom software that will display one BOM line at a time and a picture of the pcb with all the locations for that component highlighted which makes it much easier; most of the time is spent gathering the parts rather than PnP, so a manual PnP machine wouldn't really give me any benefit.

[rx8pilot]

I generally outsource production runs and hand assemble prototypes; the last board I did has ~350 parts, ~1200 pads, and 56 BOM lines. The assembly for 50pcs costed $800 total, which is $16 per board. Lead time is about 1 month.

That is pretty good pricing. Do you trust the house to source parts - or do you prepare and send kits?

Since you are able to tolerate the lead-time - it sounds excellent. Maybe I should consider using my pick and place machine to build the immediate needs so that I can outsource the rest. Overall - that may be a great way for me to maximize my time and still accomplish the quick delivery of the product.

[mrpackethead]

Having my own PNP line has transformed how i cna do things, but it was not just a matter of getting some machines and seeing things transform overnight!

two years ago i bought two 2nd hand Yamaha YV100-ii pnp machines at Auction.         These are seriously well engineered machines, but older.    They are good for placing down to .4mm pitched QFN's and 0402's.  ( can do 0201 but i dont' need to ).      I wont' commetn much on the amount of time it took to learn what i needed to know, but it was a lot of time.  Its not just the PNP machine, is the solder paste application and the reflow.

However after a lot of work, its transformed what we do, and means its economic for us to to low volume, quick turn boards.. I can run 6-8 different jobs in a single day if need be. The key was building an end to end systm to manage parts and libarys.   we are an altium Shop.   

Firstly we transformed all our Altium librarys, to use an online DBLIB.    If its not in the company Dblib you are not allowed to use the part in your design.   If you want to add a part you have to submit it into the database.  This means that you collect all the info about that part, like who sells it and were we can order it from.    Yes, its quite a bit of work.  So theres a big incenttive to use parts that are already in teh libary.  As it turns out we can make an awful lot of thigns with a library of only about 300 differnet items.     This is the key to our sucess.   We've been able to minimize the actual inventory of parts to an absolute minimum.

As stock is received, we inventory it into our stock control systme ( same database that contains the altium library ) so we know were it is.  We actually print a QR code that tracks the parts through our warehouse. So we can just put reels in the next avaible bin space.. Habing a hand held scanner has solved lots of problems.

We have a custom PNP, and BOM job that is exported out of altium..  That gets processed by some python scripts to create a Picking List, and also the Yamaha VIOS files.. It splits it up across the two machines and top and bottom. It also lets me set up panels etc.   Specific compoent information is set up in the PNP machines about the compoentns..  and we use the parts match functions there.
For most of the jobs we do, we probably only need to swap a few feeders of specialist parts over..   Most of my generic jelly bean parts never get swapped out untill they get run out of parts.

Sometimes if i have a large complex part that i only need to place a few of, i'll do it by hand at the end of the run before I reflow it. Its sometimes not worth the effort to set up.

Before i had my own PNP line, it simply was'nt economic to run 2 panels of somethign.. Now its almost press print.


The attached file is a Picking list i just ran against a new design; this one has quite a few new parts in it, but i'm building a lot of them, so Its ok.

the sections are quite explantory..     A list of parts that are alrady on the PNP,   Were they are location   NICK_15 is the 15th slot on machine "NICK",  the  2nd nubmer is the total needed, the last nubmer is the number we have in stock

In stock, is the same, but they are on the shelf.       NP_C3_3     Northpoint - Shelf C3, bin 3        Bin-B is a special bin :-)     

And a list of items to order.

When i place an order of items,   through the web interface,  we'll get another section that tells me that they are ordered, and were they are comign from.

The custom database and reports will never end. but being able to ask a question like " how many ABC's can i make with the stock i have on hand " is helpful.

As time peromits i add new modules to speed thigns up.

I'm never going to take on adhoc assembly work for others.  Thats a mugs game. The effort it takes to design pcbs that are manufacturable is quite big and peopel just dont' do it.
( I used to be one of them ).   My desings are massively designed for Manufacture. Its the little thigns that make SUCH A massive difference.


I did some interesting reports this week.

I've found that i have about 3x as many problems with QFP's as compared to QFNS..   QFN's are so much easier to handle, and reflow effectively.  You would think not having legs might be a problem. In fact it helps!

[julianhigginson]

Really interesting thread! Thanks for being so open with your trials and tribulations.

The first electronics company I ever worked for had its own P&P line. And a guy who ran it every day. As a junior engineer, I probably wasn't making the best designs, but he never had a problem with what I sent through. And from memory, the 5 or 6 engineers in our team had free reign to spec what we wanted, and there wasn't any database in the engineering side (though anything chucked over the fence to operations ended up in parts database) though the designs were all kind of similar, so our free choices probably weren't massively unique....

This was late 90s, so the gear was probably very solid and expensive and not so many years old at the time. I think something broke once, and he had a few days down while parts were express couriered from Europe, but normally he was just pushing out production batches and prototype runs day after day... He made it look easy.

It's interesting to think he might have had all the run to run issues you guys talk about, but just quietly made it happen. Though maybe having a bigger budget and newer gear (at the time) made a big difference?

[D3f1ant]

For me having a in house PNP meant we could turn out prototypes and low volume orders in very short time frames. I couldn't run my business without that capability now,it's possibly the best thing I ever invested in,  along with a CNC mill and a 3D printer.

The rest of the lifestyle experience is about the same as everyone else. The learning curve is steep, the support from the supplier limited to non existent even on a new machine.

[xaxaxa]

Tell me more about your custom software?

https://gist.github.com/xaxaxa/9a57705e701738ec38ea01df3699b7d2
It's just a simple script that takes a gEDA pcb file, bom file, and a png render of of the board, and generates a series of png images; this is an example image:

[grimmjaw]

$1,500 for a single prototype?
Do you have to kit the parts for the assembler or do they source the parts?

When my work was dominated by prototypes - I seriously considered one of the manual machines. Some of the ones I saw were pretty clever and would make a board like yours a practical effort. Perhaps not something that you are excited about, but something you could do in 4 hours instead of 4 weeks. I was able to sustain about 400-500CPH average when I was organized and had all the parts laid out nicely in holders and bins. In your case - factor in the prep work, paste printing, restroom breaks, etc....4hrs is reasonable. 5 hours would be safe.

Of course, if you needed 10.....yikes. A P&P may be helpful if the learning curve has not scared you away.

The 3-4 week lead time for outsourcing would have me looking for a very tall bridge. I have never had that much time for a prototype.

Is the price too high? Legit question..the range I'm getting is between 1kUSD-1k5USD.The board has  few BGA & 0402.We actually kit the part
Prototype is really a pain.Can easily spend 8h assembling board with no guarantee the prototype will work (either mistake in assembly or design) .

If not mistaken , manual P&P can cost as much as desktop P&P. I'm looking for solution since placing part with tweezer  for 8h will be the death of me.
From the sound of it, owning P&P seem to be not much better 

[NorthGuy]

Having my own PNP line has transformed how i cna do things, but it was not just a matter of getting some machines and seeing things transform overnight!

This is an excellent process. But how about feeders jamming, moving out of place, tapes not moving or moving too fast losing parts, parts falling from the nozzle or sticking to it. All those horrible problems that other people encounter. Do they affect you not?

[xaxaxa]

That is pretty good pricing. Do you trust the house to source parts - or do you prepare and send kits?

In my first production run I decided to let the board house source almost everything (there were just 3 parts they couldn't get that I had to provide);

Generally I'm not very concerned about counterfeit ICs; I think parts fall into 3 categories:
* high end/difficult to clone parts - FPGAs, high speed ADCs, high speed op-amps, RFICs, etc - these are the least concern, and any source (even aliexpress/taobao) is considered fine; most chips in my design fall into this category
* counterfeit-prone parts and jellybean parts - low speed op-amps, voltage regulators, microcontrollers (stm32/atmel/PIC), passives, nrf series, anything vintage, etc - these I would take greater cautions and develop rigorous functional tests
* connectors - these are the highest concern; a bad quality connector can be impossible to tell from the real thing and pass all tests just fine (and cause reliability problems down the road), so these would be the first candidate for supplying myself

I also provided a test jig that does functional tests and checks for RF characteristics which would catch out-of-spec parts.

[Selectech]

Years ago, started with tweezers and soldering iron, then got a hot air tool.

For a while had a CNC machine that put down paste, manual place with vacuum pen and a reflow oven ( toaster oven with controller ).

With low cost SS stencils now available, can't afford time & fuss to use anything else.

Got a TM220 a few years ago, largely set up with a standard set of parts ( 0603 & 0805 R & C, SOT23, 2 mm square Fets, SOD323, couple of TSSOP16 )  that are common across the roughly 20 board types I make. I have a manual p&p machine for the rest of the items that get placed, small qty per card, SOIC8,14,16 QFN20, QFN48, TQFP100 etc.

The TM220 was a bit fiddly to get set up, but has been a big help for the cards that I make in batches of 10 to 20 at a time, or new proto cards that are not yet ready for volume.

For larger qty of my main production items, I build in lots of 100 units or more with a local assembly house ( Ottawa, Canada ) using my parts kits ( 500 to 700 placements per card ) . They have been quite competitive in terms of assembly cost and assembly quality has been excellent.

I am waiting for delivery of an OpenPlacer but that is really more of an experimental aid for low volume R&D assembly rather than replacement for assembly house.

[rx8pilot]

Having my own PNP line has transformed how i cna do things, but it was not just a matter of getting some machines and seeing things transform overnight!

Nice story! Thanks for sharing....you are where I am planning to be. My progress is slow because I am still a one-man-band....but getting there.

This is an excellent process. But how about feeders jamming, moving out of place, tapes not moving or moving too fast losing parts, parts falling from the nozzle or sticking to it. All those horrible problems that other people encounter. Do they affect you not?

This is a very good question. It refers to a daily struggle in my personal experience. After everything is planned, located, programmed, and loaded.....the feeders mess up. The nozzles are sticky. Some fiddly part of the machine needs attention. It only takes ONE single feeder messing up to eat a lot of time. If you are lucky enough to have a matching spare feeder - perhaps the problem will be solved in 10 minutes. If it is the only available feeder for that type of part.....you either have to hand place that part or fix the feeder.

If you look at percentages - my system appears to be very reliable. The reality is that 99% working means 1% is not. If that 1% is a critical feeder - it takes a lot of time.
Some tapes have really sticky adhesive holding the cover film.....that can cause problems for some of my feeders but not all of them. Some feeders are more accurate than others. Some periodically don't advance. Jamming happens but quite rare with Quad feeders. I think what I need to do is spend a day or 2 going through my top feeders for maintenance and ranking. Most of the problems are easy fixes, but I do have to disassemble and inspect/measure. I have spent a lot of time calibrating and doing preventative maintenance on the machine itself....but I only mess with the feeders when they mess up. That is only noticed in the middle of a job when I am in a huge hurry so they generally just get a band-aid and thrown back into the fight.

Real feeder maintenance program....IMPORTANT. I have 300 feeders now. If each one takes 30min for evaluation and adjustments, that is nearly a month of full-time work. Yikes!!!!! Hopefully, I can average much faster and only prioritize 200 feeders.

It's interesting to think he might have had all the run to run issues you guys talk about, but just quietly made it happen. Though maybe having a bigger budget and newer gear (at the time) made a big difference?

My guess is that if I ran my machine all day, every day.....I would be MUCH better at dealing with all the quirks. It is a periodic process in my world and it takes quite a while to get back up to speed when I set up a new job. My brain has already archived what I know about P&P and I have to restore the archive to get anything done. Over time, I have made up carious cheat-sheets that help me remember little details. I have a bunch of videos that I made showing how to do infrequent tasks like calibrations, adding a new type of feeder, prgramming a new type of nozzle or whatver.

If I ever hire someone....all those notes and videos will be important in training.

If not mistaken , manual P&P can cost as much as desktop P&P. I'm looking for solution since placing part with tweezer  for 8h will be the death of me.
From the sound of it, owning P&P seem to be not much better 

In the end....it is a detailed decision based on factors that will never be prosecuted on a forum.
The really nice manual systems seem to cost as much as the entry-level  / low-capability P&P machines. For BGA-0402-prototype - a nice manual machine seems like a better option. I kinda wish I had both full-auto and manual. Manual for one-off prototypes where I am still working out values of R's and C's as I am assembling. Full auto once I have the bugs worked out.

As for tweezers......I at least doubled my placement speed when I got my cheap fish pump vacuum pen setup. With that and a good set of bent tips and a smooth rotating surface, I could place rather quick rarely having to touch the PCB itself to accomodate rotations.

[D3f1ant]

Don't discount the usefulness of using a machine in combination with hand placing for prototypes. Let the machine do all or the majority of the caps and resisitors and any hard to place fine pitch parts, have machine loaded with common value parts and jellybean transistors and diodes.
For low volume and prototypes you want a machine that is easy to setup for new parts or odd parts, and that don't need everything on feeders, ie it can pick from strips double sided taped to the placement deck

[mrpackethead]

This is an excellent process. But how about feeders jamming, moving out of place, tapes not moving or moving too fast losing parts, parts falling from the nozzle or sticking to it. All those horrible problems that other people encounter. Do they affect you not?

Yes, i do have some of those problems. however not nearly as many as the folks that are using the $3-5k machines.  teh amount of engineering in these machines is quite amazing. and i have a regular maintaince schedule where i go and check the alignments,   make adjustments etc.      Usint the right nozzles for parts is critical.  Knowing how to configure the pick up and drop is really important.   DId i say it took a LOT of time to learn all this stuff...  but its about solving those problems learnign what caused them and avoiding them in the future. 

[mrpackethead]

Don't discount the usefulness of using a machine in combination with hand placing for prototypes. Let the machine do all or the majority of the caps and resisitors and any hard to place fine pitch parts, have machine loaded with common value parts and jellybean transistors and diodes.
For low volume and prototypes you want a machine that is easy to setup for new parts or odd parts, and that don't need everything on feeders, ie it can pick from strips double sided taped to the placement deck

This is very sage advice.    I have some oddball parts that just are too hard to place,   ( a huge inductor for example ).. It easy to hand place..    I sometimes have some paste in pin Connectors.   they only add a few seconds per part to put on.

[Kean]

I have a TM240A.  I almost certainly wouldn't have purchased it, but it kind of came as a bonus along with a manual stencil printer, when I moved into new business premises.  Long story...

I had been considering getting a Neoden 4 or something around that price range, but I really don't do enough volume to justify it.  I would often hand assemble small batches of boards for my clients - never more than 10 at a time.  Many of these are quite small and tightly packed PCBs (0603, 0402, but no BGA), and the small ones are often a double sided load.

After acquiring the TM240A it sat in a corner for 6 months, until I had a job worth running - five panels of 6 PCBs.  It took a long day to get familiar with the machine, mostly finding the right way to load tapes and formatting the data file, but also making sense of the controls, and correct setting of part rotation.  Training a staff member on the basics took a few hours, including him generating the data file for his design.

I had no one to show me anything, other than some YT videos, but I do have many years of experience with CNC equipment and manual assembly.  Surprisingly I found handling a 3x2 panel was easy, but the instructions were lacking on how to restart at a particular component on a PCB in the panel other than the first.  That was the one time I reached out for support from Neoden.

I do have the usual issues, like too few "feeders", and mispicks with components getting dropped around the tape handling area.  Most issues I've found to be due to tapes getting jammed due to leader tape splices, or poor handling of the cut tape from the supplier.  Cut tape is one of those things you just have to deal with when using "expensive" parts in low volume manufacture.  I now buy as many parts that I can in full reels.

I certainly don't try to place every part with the machine, just the majority of small & cheap parts.  It isn't much hassle to hand place a large capacitor or inductor during inspection before reflow.  Usually a few of the placed parts need a bit of a nudge into position, but the reflow is the bottleneck in the process anyway.

I think I've only run about 6 different designs on it in the last 8 months, and I don't imagine using it more than one day a month on average, if that.  For that first design in a panel of 6, I've run about 30 panels, and the only real issue was reworking the first couple of panels where I had a DC/DC converter chip rotated 180.  Didn't show up in my initial low voltage tests, but blew most of the components when powered from 24V!  That was a pain to rework, as it was a TSSOP-8 package and up tight against a large diode and inductor.

I definitely haven't got the process streamlined to the point where it is economical compared to my local assembly house, but it makes a lot of sense for the small runs of prototypes or low-volume designs.

My local assembler requires me to kit parts for any runs, but at least he is pretty flexible about cut tape and using my 300x400 stencils.  Most other assemblers I've dealt with require large unframed stencils and charge $600+ each for them.

[PartialDischarge]

What about a semiautomated pick and place, like the LPKF Protoplace, haven't seen them discussed here, and make better sense for small runs. However the prices may not be that far apart

[Kjelt]

They are discussed, but indeed pricewise probably something you want to build yourself.
Still the biggest issue remains parts storage/sourcing/feeding.
The LPKF IIRC has a circular rotating storage unit, but that only covers what 100 parts?
R's and C's alone in different sizes would take 600 trays.

[mikeselectricstuff]

The very fact that there are hardly any manual P&P systems on the market, and none AFAIK from China, where labour is cheap, suggests to me that few people think they are worthwhile.

[rx8pilot]

The very fact that there are hardly any manual P&P systems on the market, and none AFAIK from China, where labour is cheap, suggests to me that few people think they are worthwhile.

The problem I saw with the manual machines is that by the time they were good enough to be useful....they had become very expensive, big, and complicated. I made a brief attempt to design and build a small machine to assist me. Essentially a dampened steady rest for my hand with a camera and lights. A fairly simple concept, but still a lot of work to bring it to reality. After machining a few pieces - I put it off to the side.

By the time it would have been useful - I would have been barely 1 step away from adding full motion control to it.

[rx8pilot]

I should add that if money and space were less of an issue - I would like to have a really nice manual machine.

[NorthGuy]

What about a semiautomated pick and place, like the LPKF Protoplace, haven't seen them discussed here, and make better sense for small runs. However the prices may not be that far apart

I don't have a PnP machine. I was thinking about getting a PnP machine for prototyping. It seems that the manual machines would target people like me. But I don't think they are useful for me. What I want is inserting a board, pressing a button, and returning back after some time to a whole placed board - this would save me lots of time.

An automated machine would do this for me, but even $10K machines don't seem to be reliable enough for this (based on this and other threads) - they're designed for speed, not for reliability or ease of use.

Manual machines have a potential to save me some time (assuming the machine is very reliable), but the benefit wouldn't be big enough to worth considering. I liked how they can place big ICs, BGAs etc., but it's very few of these I need to place, and I can rework later if something goes wrong. For such a small benefit, the manual machines are extremely overpriced.

Long story short, I will not be getting any PnP machine. The technology is not there yet.

[Kjelt]

Long story short, I will not be getting any PnP machine. The technology is not there yet.

Ah c'mon technology is not the issue, the issue is price. As Mike already said the XY gantry is a small part of the puzzle. Feeders at a price of >€100 a piece and needing at least 50 of them will set you back high four to five figures.
So what you want is a "slow" reliable machine that can pick components visually assisted in a random position from small containers or any other container then place them on the pcb and visually check the result. It does not matter if it takes 30 minutes for one pcb as long as it is ok. This is possible with openpnp I think but indeed not ready as is sold.

[rx8pilot]

What I want is inserting a board, pressing a button, and returning back after some time to a whole placed board - this would save me lots of time.

That process is called "Send the job to an assembly house"
ALL P&P systems, regardless of price, require a lot of work to build the skills and then do the job. Obviously, the newer, fancier, and more expensive machines will more reliably pickup and place overall - but they still need skilled labor in abundance. Skills and monitoring take time. Troubleshooting takes time.

I had a call with the Europlacer rep last week. The smallest machine they offer with the same number of feeders that I have for my old-school machine is about $180,000 or so. My guess is that it is easier and better in just about every way......but still a lot of work. Let us pretend it is half the time to setup (being generous), it would still take a lot of time to recover the $180k. Maybe it can go for a full day without much attention, I don't know for sure. Even if there are no mis-picks - it will still run out of parts, trays and cut tapes are part of my reality. So even when my machine is tip-top shape - there is always a part that needs to be replenished.

The used/refurbished machine I have now was about $10k by the time it was being used for the first time and it slashed my assembly efforts overall. Still a pain in the neck - but far better than manual assembly. [GUESSING] I have a couple of hundred hours total fiddling, fixing, and learning on top of the effort to just setup and run the system. If I had to guess, I would say I have probably got that time back and I am getting faster and better every time I assemble a batch of boards.

I have no expectation of 'push a button and come back later' performance from any machine. At least not anything like I experience with CNC machining. In the CNC world, the machines just make parts over and over with very little fuss once they are setup. Very little monitoring and tweaking needed if you are configured well.

When I hear assembly lead times that on the order of a month or more....my eyes get really big. The challenge of an in-house assembly system seems like the lesser evil.

[NorthGuy]

So what you want is a "slow" reliable machine that can pick components visually assisted in a random position from small containers or any other container then place them on the pcb and visually check the result. It does not matter if it takes 30 minutes for one pcb as long as it is ok. This is possible with openpnp I think but indeed not ready as is sold.

Yes, such thing would be ideal for me. I did think about DIY machine, but it would not be easy to design/make feeders. Picking from the containers has problems to solve too - parts may sit on top of each other, may lay on a side, or may need to be flipped, the nozzle may not be able reach places close to the container wall or in the corner (especially if you make walls tall). The worst, if a part falls from the nozzle into a different container, you're doomed. So, it'll take considerable time and effort to design too. In the end, even though the use of containers avoid feeders, it might not be very reliable neither.

[rx8pilot]

So, it'll take considerable time and effort to design too. In the end, even though the use of containers avoid feeders, it might not be very reliable neither.

Shortcuts are indeed hard to find in this space.
If I was looking for a hobby project - it would be a ton of fun to attack the challenge [with zero expectation of making it 'worth the effort]. Every time I struggle with some detail of assembly - it reminds me why countless engineers over the course of decades have arrived at the extremely complex solutions we see in the professional environments.

Lot's of effort to reduce cost and complexity and they all quickly throw capability, reliability, and usability out the window as they chase a low price tag.

With all of that said....the commercial machines have a huge focus on raw speed. It the focus was for a slow-ish or modestly fast machine that was VERY easy and VERY reliable while capable of placing 0201 and BGA - I feel like that would lower the cost enough to be practical for small businesses. I can be totally successful with a precise and reliable process that is slow by commercial standards. New machine placement rates are out of this world......but I only need 1000-1500CPH to be completely satisfied.

Single nozzle, single gantry, slow pitch ball screws, lighter weight chassis - not a problem.

[khs]

I have no expectation of 'push a button and come back later' performance from any machine. At least not anything like I experience with CNC machining. In the CNC world, the machines just make parts over and over with very little fuss once they are setup. Very little monitoring and tweaking needed if you are configured well.

This is excactly my experience.

My old cnc mill runs alone.
Running my pnp, one finger is always on the stop button ready to press.

The reason may be a mill uses quite large and heavy tools and work pieces,
a pnp handles tiny parts and small pc boards.

Additional a pnp is much faster. As a rule of thumb my milling jobs take about half a hour to one hour, so I can go into another room and make some development work. (For me there is no way to work in the garage, because it's a common basement garage  )

The time my boards are in the pnp is about two minutes (70 parts). This is 10 times shorter (or faster) compared to my mill jobs.

Making a prototype with my pnp over all takes about two hours from the layout software to the soldered board.

If I would send the boards to an assembly house it would not save much time. The parts must be ordered, the orders must be placed, the parcel must be received, opened  and inspected and the invoice must be paid.

At the end of the day, it was one of my best investments to pay about 5K€ for a CSM60 with 50 feeders including a (new) low noise compressor and move it all (in parts) one stair down.

[mrpackethead]

I have a DIMA FP600 manual PNP.  Once everythign is setup a good operator can place 900 parts per hour.     Its the setup time that will kill you.

[JPlocher]

> What I want is inserting a board, pressing a button, and returning back after some time to a whole placed board - this would save me lots of time.

So would we all...

Assume the machine works perfectly - you still run out of parts, have board/solder paste issues, inventory management and physics to get in the way...

As with any precision machine (anything that slaps around 0402's and 0603's at a "thousands per hour" rate *is* a precision system, or it is garbage), routine maintenance and recalibration is required.  Feeders wear out or get bumped, springs get sprung, cameras collect smudges, waste tape curls around and finds its way back into the moving parts, and - even after tweaking everything, something will get out of alignment as soon as you turn your back.  Google, Facebook and Apple have nothing on these machine's level of (malevolent) AI technology...

With all that, 6 months in, after several hundred boards and a dozen designs, I'm still happy with my CHMT48.  The friction tape cover takeups and the pin-pulling part advance are not nearly as convenient as swappable Yamaha pneumatic feeders, and I'm still improving my workflow from board design to machine (AMEN! to the posts discussing parts management processes, they are an absolute must-have!), but it makes prototyping and short run production possible for the hobby work I do (model train electronics).  My hobby time is my scarce resource, and the PnP system pays itself back at a 10x to 20x rate (1-2 hours on the machine -vs- 2-3 weekends mixing board assembly with family time...); granted, I could do the same with a fab service (and have...); from a time perspective, I spent almost as much time kitting parts, creating assembly and test instructions and doing the running around as I did hand assembling things in the first place - sigh.

From a financial-only perspective, if you need to ask, the answer is "send it to a fab service".  At the going rates here for PTH and SMT hand assembly services, you could get a 500 or 1000 boards built for just the capital cost of a low end machine, not counting stencil printers, ovens, solder stencils, solder, the space used for all the above and the time out of your life it will consume.

• Lesson #1:  Reels are much better than cut-tape segments
• Lesson #1A:   The parts you really need to place won't fit into your machine because <perversity>
  
• Lesson #2:  Reels look like they hold an infinite quantity of parts - but they run out when least expected!
  
• Lesson #3:  When loading a reel of new parts, be sure to sample and store sufficient loose pieces so that you can rework/repair PnP failures during board inspection - it is a PITA to extract single components from a reel after it has been loaded into the machine/cassette!
  
• Lesson #4:  The dandruff that falls out of a PnP machine is only salvageable if you don't value your time.
  

[rx8pilot]

After going it alone for a couple of years..... I no longer feel alone. The good and bad I have experienced seem to be rather similar to others that have chimed in.

Really nice to read all these experiences and insights.

Short and misplld from my mobile......

[Kjelt]

I have no expectation of 'push a button and come back later' performance from any machine. At least not anything like I experience with CNC machining. In the CNC world, the machines just make parts over and over with very little fuss once they are setup. Very little monitoring and tweaking needed if you are configured well.

This is excactly my experience.
My old cnc mill runs alone.
Running my pnp, one finger is always on the stop button ready to press. 

Weird comparisson but if you want to make it my thoughts about this are different:

- a cnc needs each and every workpiece mounted manually over and over again unless you have a $1M workcenter with workpiece library where you can pre-install 6 or so workpieces. Tools can break, motors can loose steps when feed is too high and stop so a cnc machine can stop now and then and you have to clean clearances, change the tool and restart where you left off. I never saw a full automated cnc line where aluminium blocks are clamped by robots and the finished product is unloaded in boxes without human handling, have you?

- a P&P machine when installed properly and maintained can run 100s of boards sequential without stop. The last line I visited begin of this year had shifts of 10 persons personell monitoring 7 pcb lines and most they had to do is visually inspecting the output (more monitoring), putting new reels/feeders on the machine before the parts were running out (orange light). In the two hours I visited them, no red light occurred (full stop due to placing failure or obstruction or whatever) and the 7 lines produced over I believe 1500 boards. There were no persons with their hands over the ES button

[rx8pilot]

Weird comparisson but if you want to make it my thoughts about this are different:

- a cnc needs each and every workpiece mounted manually over and over again unless you have a $1M workcenter with workpiece library where you can pre-install 6 or so workpieces. Tools can break, motors can loose steps when feed is too high and stop so a cnc machine can stop now and then and you have to clean clearances, change the tool and restart where you left off. I never saw a full automated cnc line where aluminium blocks are clamped by robots and the finished product is unloaded in boxes without human handling, have you?

- a P&P machine when installed properly and maintained can run 100s of boards sequential without stop. The last line I visited begin of this year had shifts of 10 persons personell monitoring 7 pcb lines and most they had to do is visually inspecting the output (more monitoring), putting new reels/feeders on the machine before the parts were running out (orange light). In the two hours I visited them, no red light occurred (full stop due to placing failure or obstruction or whatever) and the 7 lines produced over I believe 1500 boards. There were no persons with their hands over the ES button

The comparison to CNC is indeed imperfect. They are very different in so many ways. My comparison is just the relative 'fiddliness' of each one. Based on your comments, it appears your perception of the CNC process is limited. As for your example..... a full-time line making the same board 1500 times with a few $million in equipment (AOI, spare feeders, etc) is a different animal compared to a small business with an in-house line. If you visit the HAAS factory you will see about 1 operator per 10-15 CNC machines.

After 10 years owning/operating a CNC shop and the last few years getting up to speed in PCB assembly.....I can positively, without the slightest hesitation, say that the CNC process is a fraction of the learning curve and has a far higher process reliability overall. As an example....my first CNC machine arrived before I even knew how to load a tool in it. In fact, I did not even know how to turn it on. A week later, I was making parts that were ready for release. It was a long week.......but I was making parts. My capability in terms of complexity, precision, and efficiency rapidly improved from there. After about a year, the learning curve transitioned to a more nuanced level where I am splitting hairs over speed, surface finishes, process efficiency and such.

I learned CNC/machining from the ground up and it was not easy and it is not simple in any way. It is a constant challenge and a never-ending learning curve. PCB assembly, in my experience so far, has more moving parts, more variables, and requires far more inspections and handholding from end to end. PCBA requires me to manage hundreds of unique raw materials and deal with extreme precision distributed over hundreds of feeders. 2 years after getting my machine commissioned and 'ready' - I still struggle with each job in some way. There is always something that is not as good as it should be - and it only takes one problem to eat a lot of time.

In my CNC world - it is not without problems but they are much easier to see issues from a mile away. I have broken one tool in the last 8 months. One tool was broken during setup - not while running. The majority of the parts I make pass inspection on the first try. It never mis-picks a tool.

Perhaps if I spent $500k+ on all new SMT equipment - automated printer with inspection, modern P&P, feeders for every imaginable part, slick process control software, AOI, oven, and electrical test.......I would expect far fewer error rates and higher productivity (After a very long year of learning). That would also require a big investment in part management and storage along with purchasing and inventory control.

[Kjelt]

Yes my cnc experience is limited but I found it also a steap learning curve.

I am guessing here but probably because you have the CNC experience you think differently, if

After 10 years owning/operating a CNC shop

Was
"After 10 years owning/operating a SMT line"
It was the other way around, I dunno. 

Modern 3D cam software that generate code with tool simulation possibilities is miles away from the past.
The modern CNC machines that then get their G code directly fed from a PC with look ahead etc are completely different beasts from the 80s Haidenhahn machines.

I wonder how much easier a modern P&P machine would be compared to your DOS based oldie.

If the smt board shops would have so much problems with running series they would have a tough business case. They ask big $ yes but they get the job done repeatedly, consistent and in time.

[rx8pilot]

"After 10 years owning/operating a SMT line"
It was the other way around, I dunno. 

Well, if I normalized for experience.....after 2 years of CNC I had already moved on to 5 axis work and had 3 machines that I was able to keep busy with 1 person. The process was not easy or simple - but it was at least predictable with 99% accuracy. Machine-related issues almost never stopped the process. The machines I had at the time were considered 'entry level' by commercial standards. Almost every limitation was based on my personal set of skills and understanding. As time marched on....I learned to push the machines faster and faster. The setup times shrunk. Scrap parts nearly disappeared. I don't see equivalent progress in my P&P efforts. It seems that to rapidly improve the process - I need to have a team and perhaps all new equipment.

After 2 years of P&P, I really struggle to anticipate issues and the process is very hard for me to schedule. To be fair, my P&P equipment is fairly old - although it is in very good condition at the moment.
I have a Skype session in a couple of weeks with a sales engineer at Europlacer. They are going to walk me through the setup and run process so I can see for myself how much different/better it is to setup and tweak on a modern machine.

Assuming that goes well.....I will visit an actual line using the same machine. That miniature education will be very interesting! It is hard for me to understand how much of my limitations and issues are directly related to the specific machine I have. The claim is that it only takes 30 seconds to load a feeder - we will see. The claim is that it can run through some enormous number of picks of an 0201 part without a mis-pick - we will see. My guess is that it is true as long as everything is completely and totally perfect. I will be trying to learn, more than anything, is what goes wrong - how often - what causes it - and what the remedies are.

I would not want to be overly reliant on service technicians or remote sessions everytime a problem pops up.

[mikeselectricstuff]

The claim is that it only takes 30 seconds to load a feeder - we will see.

Bet that's for a new reel with full leader, not a cut tape

The claim is that it can run through some enormous number of picks of an 0201 part without a mis-pick - we will see. My guess is that it is true as long as everything is completely and totally perfect.

..and the machine gets serviced every 6 months under a maintainance contract...

[khs]

I never saw a full automated cnc line where aluminum blocks are clamped by robots and the finished product is unloaded in boxes without human handling, have you?

Yes. It was a combination of a lathe and a mill. Left side a (big) rod of aluminum came in and from the middle of the machine the finished parts came out. If I'm not wrong, it was in the range of 100K€

- a P&P machine when installed properly and maintained can run 100s of boards sequential without stop.

This is true. No doubt.
But I'm quite sure, the people of an assembly house have spent a lot of time until all pcb lines are working without problems. It's is a different business. If one the feeders in a line doesn't work, it costs a lot of $$$/€€€.

If one of my feeder don't work, I place the part manually. The additional working time (cost) is minimal.
This is the difference.

But back to the simplified comparison:

A simple cnc mill has
- one workpiece
- a xyzuw head with the milling tool.

A pnp has: 
-   1 main workpiece (the pc board)
- +10 additional workpieces (the parts)
- +10 feeders
- one or more  xyr(z) heads with placing tools (nozzles).

To run a cnc mill you have to adjust two parts: The workpiece and the tool.
To run a pnp you have to adjust, the pc-board, the heads and all feeders.

So even with a small pnp there is a factor of 10 more work to do (sure, very very simplified).

My comparison is just the relative 'fiddliness' of each one.

This gives an additional factor depending on part size (1/part size).

[rx8pilot]

The claim is that it only takes 30 seconds to load a feeder - we will see.

Bet that's for a new reel with full leader, not a cut tape

Yes. I asked about cut tape handling and the promises quickly departed.....

The claim is that it can run through some enormous number of picks of an 0201 part without a mis-pick - we will see. My guess is that it is true as long as everything is completely and totally perfect.

..and the machine gets serviced every 6 months under a maintainance contract...

Yes - this an interesting point on the business end. For a dedicated assembly house, it is a fairly straightforward calculation to weigh the benefits of an expensive preventative maintenance program. For a low-use, in-house system it probably does not make any sense. Curious if the manufacturers are willing to answer your phone calls if you are not on a maintenance program. Maybe they have timers on board that can drive the maintenance effort instead of a calendar.

[Kjelt]

Look in your service manual for the service and replacement times.
If you run low series in long intervals you might need to run other schedules for some cleanings.

[rx8pilot]

Look in your service manual for the service and replacement times.
If you run low series in long intervals you might need to run other schedules for some cleanings.

Totally true....although I have no manuals 
I just finished cleaning and greasing the ball screws and linear guides. The linear encoders were cleaned, which is very delicate and hard to reach. The optics on the cameras were cleaned. The down camera lens had a loose front element which looked like erroneous behavior from the machine. I thought it was a motion control problem which I traced, racked, and tested for two days before I realized the loose lens was just making it look like the machine was not hitting its marks. I checked backlash, absolute accuracy and repeatability which are all in good shape. A number of belts were re-tensioned, the PCB shuttle bearings replaced.

Next up is a critical alignment of the feeder docks and JEDEC trays - I think they are all crooked which only matters when you are setting up the machine. For a new setup - if all the feeders are perfectly aligned to the XY axis - almost no software offsets are needed. Each feeder has an internal offset that lines it up - all of mine are a little different so each feeder needs to be adjusted in software everytime I place it. Some of those alignments should pay for themselves.

I have replaced about 100 peel rollers that pull the cover tape through the feeders. The vacuum path probably needs to be cleaned - but I don't even know how to do that yet. Based on the muck I have cleaned from the nozzles - it is a safe bet the vacuum system is also mucky.

There is a leak in the air blow-off system - minor but needs to be addressed.

All in, I would estimate it needs about 10 hours of labor remaining for maintenance right now and I would think I would be good for about a year. Some of the recent issues I have addressed will probably outlive the machine. I tend to look and listen for tiny differences in general. When something does not look right or sound right - it is usually not good. It would be nice to have a manual - but by now I have taken just about every section apart at some point. I have learned to calibrate and measure it.

[Kjelt]

You really are a CNC person 
The vacuum and pressure system are important.
In my manual of an old Philips P&P the vacuum pipes needed cleaning each 100 hours or so, checking the vacuum and pressure with a barmeter (the values are in the manual tolerance of 0,05 bar), vacuum filters etc. inspection of the nozzle, small wear could mean many mispicks etc.

Is there no-one on the forum that has the manual for you or a copy, could be some things in there you have not thought of that might make a big difference.

[rx8pilot]

The vacuum and pressure system are important.
In my manual of an old Philips P&P the vacuum pipes needed cleaning each 100 hours or so, checking the vacuum and pressure with a barmeter (the values are in the manual tolerance of 0,05 bar)

How do you clean the vac lines?

Is there no-one on the forum that has the manual for you or a copy, could be some things in there you have not thought of that might make a big difference.

PPM has the pneumatic and wiring diagrams in what they call a manual - they want $150 for that. It will be a bargain if it is rather complete. I fear that it may leave me disappointed. They have answered a pile of phone calls for free, but for some reason are protective of the manuals. The operation manual they offer, for example, only covers the software and it is not a well done manual at that.

I may just buy the $150 mystery manual and hope for the best.

[Kjelt]

How do you clean the vac lines?

On the Philips machine you remove the nozzle, inside is a small pipebrush that acts like a prefilter.
You remove the pipebrush, insert a new one , rag it a few times keep it out. Then on the pressure manifold of that head there is a cleaning setting which blows low pressure 0,6 bar ultra clean filtered air through the vacuum tube and head so a reverse blow action. Then replace the pipebrush and the nozzle.

This differs per P&P machine ofcourse but most have multiple filters, one close to the head, one before the pump. If you have cleaned your system, measure the vacuum on the head so you have a reference when something is not right.

[Kjelt]

The machine has a calibration nozzle in combination with calibration software it determines if the system is still within specifications.
Btw here is some example of a maintenance schedule for this machine.

[mrpackethead]

I can load a 8mm Yamaha Pnematic feeder in about 3 minutes with a new reel of parts, or in about 4 minutes with cut tape. ( provided its not super short.      This assumes you have the feeder and hte materials on the bench.

[rx8pilot]

I can load a 8mm Yamaha Pnematic feeder in about 3 minutes with a new reel of parts, or in about 4 minutes with cut tape. ( provided its not super short.      This assumes you have the feeder and hte materials on the bench.

I would say that is a safe estimate for Quad feeders too. Could be faster if my life depended on it, but 3 minutes is a good average for new reel.

I could save some time if I had some better splicing tape for the cover film along with a dispenser that is easy to pull short tabs from. At the moment, I am using Kapton tape and a terrible dispenser. Any suggestions for some thin, easy to dispense tape with aggressive adhesive would be nice. The same stuff the OEM's use on new reels - not sure where to get something similar.

[Bassman59]

Perhaps if I spent $500k+ on all new SMT equipment - automated printer with inspection, modern P&P, feeders for every imaginable part, slick process control software, AOI, oven, and electrical test.......I would expect far fewer error rates and higher productivity (After a very long year of learning). That would also require a big investment in part management and storage along with purchasing and inventory control.

And then you’d find that you were now a contract assembly house, not a product design business.

[tautech]

I could save some time if I had some better splicing tape for the cover film along with a dispenser that is easy to pull short tabs from.

Could you lap join the cover film and so use a double sided tape ?

This stuff really sticks like poo to a blanket and although I use it mainly for business card to brochures I've used for finger pull blocks on sliding glass windows. Sort of stuff that when it's stuck, it's stuck for good !
http://multimedia.3m.com/mws/media/765005O/iatd-product-info.pdf

[rx8pilot]

Perhaps if I spent $500k+ on all new SMT equipment - automated printer with inspection, modern P&P, feeders for every imaginable part, slick process control software, AOI, oven, and electrical test.......I would expect far fewer error rates and higher productivity (After a very long year of learning). That would also require a big investment in part management and storage along with purchasing and inventory control.

And then you’d find that you were now a contract assembly house, not a product design business.

Exactly.....

[mrpackethead]

I can load a 8mm Yamaha Pnematic feeder in about 3 minutes with a new reel of parts, or in about 4 minutes with cut tape. ( provided its not super short.      This assumes you have the feeder and hte materials on the bench.

I would say that is a safe estimate for Quad feeders too. Could be faster if my life depended on it, but 3 minutes is a good average for new reel.

I could save some time if I had some better splicing tape for the cover film along with a dispenser that is easy to pull short tabs from. At the moment, I am using Kapton tape and a terrible dispenser. Any suggestions for some thin, easy to dispense tape with aggressive adhesive would be nice. The same stuff the OEM's use on new reels - not sure where to get something similar.

I use these.  they make them in 8/12/16 mm width.. 500 in a box. Easy as to use.  Keep some tape from old reels,  ( and keep some old reels, they all come in handy ) to use to splice on.   


https://www.aliexpress.com/item/SMT-Double-splice-Tape-8mm-blue-500pcs-box/32750742195.html?spm=a2g0s.9042311.0.0.27424c4dyPYXGS

[D3f1ant]

0201 is no problem at all. ROFL.
 

[D3f1ant]

For spicing leader on, or joining two strips together (why they sometimes supply two short lengths instead of a continuous strip?),  I use 6mm wide masking tape, took a few goes to find one that is sticky enough, some brands don't stick to the top tape particularly well.

That splicing tape looks interesting, might have to try it, thanks for the link.

[mrpackethead]

There s multiple places selling them,  I bougth a variety of widths.. They work really well, as they stick on both sides, makign the tape spliced nice and tight, and because it folds over in half, its all straight without any real issues. ( if your splice is not straight, your in for trouble! )

[rx8pilot]

I use these.  they make them in 8/12/16 mm width.. 500 in a box. Easy as to use.  Keep some tape from old reels,  ( and keep some old reels, they all come in handy ) to use to splice on.   


https://www.aliexpress.com/item/SMT-Double-splice-Tape-8mm-blue-500pcs-box/32750742195.html?spm=a2g0s.9042311.0.0.27424c4dyPYXGS

Ordered those. Bargain. Thanks for the tip!

0201 is no problem at all. ROFL.
 

If you have enough money - 0201 is no problem
If you have steady hands - 0201 is no problem
If you have a bad temper - 0201 IS a problem

[jmelson]

Assuming that goes well.....I will visit an actual line using the same machine. That miniature education will be very interesting! It is hard for me to understand how much of my limitations and issues are directly related to the specific machine I have. The claim is that it only takes 30 seconds to load a feeder - we will see.

Well, on my Philips CSM84, if I have to load a new reel or fiddle with a balky feeder/tape, it usually only takes me 1-2 minutes to have the feeder back on the machine and running.  Presumably, my feeders are about as old as my machine, now over 20 years old.  The CSM series used mechanical advance from the head for 8 and 12mm feeders, so they are really basic.

I would not want to be overly reliant on service technicians or remote sessions everytime a problem pops up.

No, this would be a disaster for the small shop!  I've managed to fix everything that went wrong with my machine so far.  And, any tech would likely have ordered terribly expensive replacement parts, when in several cases things just needed cleaning.

Jon

[jmelson]

I could save some time if I had some better splicing tape for the cover film along with a dispenser that is easy to pull short tabs from. At the moment, I am using Kapton tape and a terrible dispenser. Any suggestions for some thin, easy to dispense tape with aggressive adhesive would be nice. The same stuff the OEM's use on new reels - not sure where to get something similar.

I use masking tape.  For narrow tape, I just wrap the masking tape over the cover tape as many times as needed until there is no remaining sticky side exposed.
This generally flows over the puller mechanism pretty well.  (There are masking tapes with weak adhesive, this stuff seems to be the more sticky type.)

Jon

[rx8pilot]

What manual printers is everyone using? Is it practical to have a reasonably functional printer <$1k USD? The fancy manual printers look amazing but at $4k-$6k new.

Safe to say....any printer is probably better than what I am using now.

After getting a printer, I have guessed at technique, squeegee choice, angle, speed, single vs multipass, etc....
With small, single PCBs I can nail fine pitch prints all day. With panels, fine pitch is scary and I have too many flaws overall. Would love to improve consistency in printing fine pitch panels in a single pass. It is a bit of an art form that relies on feelings and experience, but I bet an equipment upgrade would at least make a contribution to success.

[JPlocher]

The "under $US500" manual printer world seems to fall into three categories -

• stuff aimed at an occasional user making small lots with frameless stencils, usually less than $US200
• basic, no frills and little adjustment 30x40cm framed, single sided board focus, ($US300-$400) and
• basic, hints of frills and a bit more adjustment 30x40cm framed, with "support" for dual sided boards ($US400-$550)

The Neoden printer (resold by many at widely different price points) is in the last category; it supports board standoffs and thus 2-sided boards, with X, Y, Z and horizontal twist adjustment.  I have this printer, and it works well for small to mid volumes and 0602 jellybeans with little to no alignment registration problems; it takes me about 5-10 minutes to align and set up a new board/stencil set, and 2-3 minutes to do the same with something I've set up before.

Similar looking ones in the $200-$300 range sometimes are only for single sided boards - look to see if the stencil hinge is in the same plane as the board mounting platen instead of ~1/2" above it...

I haven't used the "$US 1k+" ones, but they look to offer more and better frills and features - chief among which seems to be better/repeatable framed stencil registration / alignment and less slop/play/backlash in the hinge assembly to work better with fine pitch.

Key for me (I've got 10x-20x board designs in my "product line") is to design for fabrication consistency - I use a common board frame "component" with similar fiducials and alignment pin/holes across all my designs to minimize the board to stencil registration/alignment adjustments when swapping between jobs.   

[rx8pilot]

This is not my pick and place lifestyle, but I am drooling for just a moment.....




Now....back to our regularly scheduled programming.....

[forrestc]

Somehow missed this thread starting, guess it's better late than never....

My story goes something like this:

Back in 2006, I decided to start selling some products I had designed.  For ease of assembly (so I thought), I designed these products with through hole technology.  Initially, I was hand soldering each board.   Then someone told me about dip soldering, so I was dipping boards in a pot of molten lead or more accurately a electric skillet modified so it would get hot enough to melt solder in it.   That was very nasty.   A year or so later, I bought a wave solder machine.

Fast forward to 2010 and I had a really good year, well at least on paper.   Because of the details of the US tax code, I had about $80K or so of profit according to the books and very little cash to show for it - since most of the $80K went into inventory which you can't call an expense until you actually sell it.     I had also been looking at/for a pick and place machine for some time - mainly because through hole was becoming so obsolete it was hard to find certain types of components in through hole.   Because of another part of the US tax code, one way to resolve these issues is to go out and buy a big piece of equipment, write it off in a single year,  and then pay the loan off over time, spreading the tax impact over several years.   So, I went out to a vendor which I had been looking at for some time and bought a surface mount line consisting of a Manncorp/Autotronk MC384 P&P machine, a small batch oven, and a dry box.   

This was one of the best purchases I've made for the business.    I typically use the same parts over and over in my designs - I have around 30 different boards we assemble, and all of them use the same power supply section, I use the same capacitors, etc.   Because of this, once a part goes on the machine it stays one the machine.    There are some very very low volume parts that we do hand place, but almost everything else is just one the machine.   (we probably hand-place around 100 parts/month, if even that many).

I did spend some extra money on a high-precision dot dispenser for solder paste on the machine.   This allows us to do a certain amount of assembly without a stencil.   For instance, all of the prototype boards which come in are done on the board, and I don't bother with a stencil.   There are a few very low volume products (like we build maybe 20-30 a year) that we also use the dispenser for.   I wouldn't recommend this for high volume runs (we have a stencil printer for those), but for prototypes it's great.

I will say that the machine has relatively trouble free.  I suspect a lot of this is because it was a brand new machine.  Not to say it isn't without it's problems.   I think the biggest two I'm aware of are issues where the cover tape on a reel will remain stuck to the carrier tape and the feeder misfeeds when it reaches the pick location.   And the other are odd vision issues which prevent proper placement - usually caused by some component change or difference such that the vision doesn't do a good job of figuring out the actual alignment of the part on the head.

Every once in a while I'll send out one of our higher-volume products for an assembly quote just to check to see how we're doing cost-wise.   Although it's getting closer to what it costs us to do assembly every time I do this, so far it really hasn't come close enough to seriously consider it.   Plus, I have to consider the somewhat intangibles such as control over the process, and also that ability to do one-off prototypes using the actual machine and parts we'd be using in production.

[SimonD]

wow !!!
Is there somebody who knows the prices of this wonders ?
( Just for our imagination ...  )
is a USA company ? in their site i see offices in USA, Europe and China.

[mikeselectricstuff]

On the subject of printers, I use the Eurocircuits one, which has no adjustments except tension and height. It uses a pin registration system, which works very well, but is ridiculously over-engineered.
I think with some clever design, something similar could easily be done for between a quarter and a third of the price.

[mrpackethead]

On the subject of printers, I use the Eurocircuits one, which has no adjustments except tension and height. It uses a pin registration system, which works very well, but is ridiculously over-engineered.
I think with some clever design, something similar could easily be done for between a quarter and a third of the price.

Is that one that uses frameless?    The main lissue i have with framed stencils is they take up a lot of space to store.

[mikeselectricstuff]

On the subject of printers, I use the Eurocircuits one, which has no adjustments except tension and height. It uses a pin registration system, which works very well, but is ridiculously over-engineered.
I think with some clever design, something similar could easily be done for between a quarter and a third of the price.

Is that one that uses frameless?    The main lissue i have with framed stencils is they take up a lot of space to store.

Yes - you can clamp and tension any size of frameless stencil
https://www.eurocircuits.com/ec-equipment-ec-stencil-mate/

[khs]

What manual printers is everyone using?

Because I did not find a suitable one (for my pnp) , I made one myself (with my cnc).

If I'm not wrong, it's the only printer in the world capable of printing small circuit boards using stencils without a frame.

If interested, you may take a look at


At 0:38  you can see how to adjust very small pc-boards.

Unfortunately it's out of stock at the moment, but the new batch in nearly finished.

(It was designed for very small businesses, but reality showed that in many cases larger companies were those who had one ordered...)

[julianhigginson]

Because I did not find a suitable one (for my pnp) , I made one myself (with my cnc).

If I'm not wrong, it's the only printer in the world capable of printing small circuit boards using stencils without a frame.

really nice! It looks very solid and neat and simple and efficient, and I am keen to know when they are available.

I notice it holds boards with a channel in the edges of the side pieces.
How does it handle different thickness boards?

And do you get flexing issues when printing on wide boards?

[khs]

It looks very solid and neat and simple and efficient, and I am keen to know when they are available.

The milling and drilling work is done, I expect the stencil printers in about a week back from anodizing.

I notice it holds boards with a channel in the edges of the side pieces. How does it handle different thickness boards?

There are two channels, 1.0 mm and 1.6 mm.

And do you get flexing issues when printing on wide boards?

It depends on the thickness of the pc-board and the force on the stencil. Because the squeegee is supported by the both board holders, the force on the pc-board can be quite low.

But the stencil printer is more or less a vise for the stencil with a massive holder for the board, so it is more or less impossible to move the stencil relative to the board. So, when the pc-board is bent, the stencil and the pc-Bard are bent together.

To minimize the forces on the stencil, I started with a solder paste with low viscosity. Surprisingly, the print quality increased with increasing viscosity.

Increasing viscosity means that at the end the solder paste has the viscosity of plasticine, with the appropriate forces on the stencil / pc-board. It was a bit like filling holes in a wall with smoothing cement. As a result the solder paste was formed to tiny 'bricks' on the 0.5 mm pitch pins and my 144 pin FPGA was soldered perfectly. (I think this could  work with other stencil printers too).
 

[mrpackethead]

Interesting design for a printer

At small board sizes I can see this working alright. But once the boards get much bigger than 100mm wide or so it might be problematic. You need mod point support pins or something to keep the boards
From flexing.

How does it cope with double sided?  Normally there is some warp in the boards after reflow so they are are not perfectly flat

[briandorey]

I use a https://blundell.co.uk/product/manual-screen-printer-foil-type/ for unmounted stencils and we have made thousands of boards using it in the past few years.

I made a shoort video of our first go with it on . It wasnt very cheap but it is well made.

[Dubbie]

That looks very nice khs . I might have to build myself something similar.

[khs]

At small board sizes I can see this working alright. But once the boards get much bigger than 100mm wide or so it might be problematic. You need mod point support pins or something to keep the boards From flexing.

The stencil printer is designed for small prototypes typical up to euro card size 100 mm x 160 mm.

Here in Europe there is a company adding free stencils to their prototype boards. So for a small pc board you get a small stencil. With this stencil printer you can print with it in minutes. This saved me lot of time and money.

With increasing board size professional (an a lot more expensive) printers are better.
Every tool hat it's limitations..

How does it cope with double sided?  Normally there is some warp in the boards after reflow so they are are not perfectly flat

I have not seen it yet. If the pc-board is not flat, it would be very difficult to print with a stencil due to the gap between the pc-board and the stencil.

[khs]

I use a https://blundell.co.uk/product/manual-screen-printer-foil-type/ for unmounted stencils and we have made thousands of boards using it in the past few years.

I made a shoort video of our first go with it on . It wasnt very cheap but it is well made.

Nice printer. 

[rx8pilot]

I use a https://blundell.co.uk/product/manual-screen-printer-foil-type/ for unmounted stencils and we have made thousands of boards using it in the past few years.

Have you tried larger panelized setups? Anything fine pitched?

[briandorey]

Have you tried larger panelized setups? Anything fine pitched?

I havent used it on larger panelized  boards but have used it on a single 20cm x 15cm pcb with down to 0.4 pitch QNF packages.

[DerekG]

Nice printer.

Just a kind pointer. You will get the paste right to the edges of the pads if you screed in all 4 directions

[girts]

Have you tried larger panelized setups? Anything fine pitched?

Panelized / double sided boards boards / fine pitch - no problems, excellent.
Must use some additional supports to avoid curving and bending, standard set of supports is not enough for larger or flexible boards.

Just a kind pointer. You will get the paste right to the edges of the pads if you screed in all 4 directions

There is a simplier solution - use "rounded rectangle" instead of "square" for pads.

Just a kind pointer. You will get the paste right to the edges of the pads if you screed in all 4 directions
 

[khs]

If you have a bad temper - 0201 IS a problem

Maybe. But take a look at the temperature:

To pick the 0201 you need an absolute accuracy of about dx = 0.1 mm
The dimension of your pnp is about  L = 2 m
The temperature coefficient of steel is about K = 12e-6 / K

So we can calculate the maximum temperature change for a 0.1 mm shift.

dx = K * dt * L
dx / (L * K) = dt = 4.1K

So placing 0201 may require a temperature stabilized room.
(Just my two cents..)

Some more info:

www.circuitsassembly.com/cms/images/stories/pdf/0305/smtai02_motorola.pdf

[rx8pilot]

My machine and almost any commercial system use glass linear encoders. They are very stable and .1mm is no problem at all. The Quad machines use steppers, but the linear encoders don't really care of a step is missed. (missed steps are likely very rare, only occurring if something is very wrong).

This old machine is probably in the range of .01mm repeatability. Assuming the ball screws and every other motion component is dialed into perfection. Obviously, the modern gear will beat that even when running at 10x-20x the speed.

For 0201 and 01005 parts - the nozzle and feeder performance need to be amazing. The nozzle for 0402 is about $40 - the nozzle intended for 0201 and 01005 is $300 each.

[mrpackethead]

If you have a bad temper - 0201 IS a problem

Maybe. But take a look at the temperature:

To pick the 0201 you need an absolute accuracy of about dx = 0.1 mm
The dimension of your pnp is about  L = 2 m
The temperature coefficient of steel is about K = 12e-6 / K

So we can calculate the maximum temperature change for a 0.1 mm shift.

dx = K * dt * L
dx / (L * K) = dt = 4.1K

So placing 0201 may require a temperature stabilized room.
(Just my two cents..)

Some more info:

www.circuitsassembly.com/cms/images/stories/pdf/0305/smtai02_motorola.pdf

One of the reasons to use fidudicals on boards is so that you dynamically scale things every time a board is loaded.     Yes, there is some shift in your machine, but in reality its not a huge issue.

[AgileE]

Manual printers can be pretty good if you take your time. I use a small USB microscope thingy on a PC to help with alignment when smaller parts are used. It's an APS printer for framed stencils and I still use even though the original APS P&P machine (LEV40) is long gone. I replaced the APS P&P with a Europlacer XPII as even with a manual printer as long as the boards were panalised I could printer faster than the P&P could do it's job. Still can with a lot of boards even with the faster P&P machine.

I do mostly my own boards, but all so other peoples, it certainly teaches you good design for manufacture.

[flasonsmts]

I want to adjust some idea of the Automatic solder paste printer, An automatic solder paste printer don't means it's high speed stencil printer, Automatic solder paste printer is just a high accuracy SMT printer, that it is automatic adjusting the stencil position by high clearance camera and computer. The print speed of this kind Automatic solder paste printer is just about 2 times printing per minute, a manual or semi auto solder paste printer is faster than this. When will need to use automatic solder paste printer, when the components is too small like 0201 and 01005 chips, this kind solder pad is hard to identify by human eyes, and manual or semi auto printer maybe slightly changing the print position by shaking etc. This will need Automatic SMT stencil printer. how to got the high output, it's simple, use 2 or many printer.

[rx8pilot]

I feel like a manual printer with a USB camera used for alignment would be right for low-volume / in-house printer usage. The full auto printers are generally huge and complicated, although probably do well with tiny apertures and critical alignment.

My super low-end setup uses a photographers loupe that I put directly in the stencil and use a handheld LED light to get the alignment. In the end, it gets a great alignment but at the expense of being a total pain in the butt. Wishing I had about 2 days that I could dedicate to designing and machining an alignment system based on a camera with a monitor.

I'll just add that to my 2-mile long lab wish list........

[jedas]

I've purchased TVM920, manual stencil printer and T960 oven one year ago. Now it is the time, when it almost payed off. We've assembled about 3k boards with it. PNP behaviour is unpredictable to say the least. I've tried to cure it with openPNP software, didn't helped much. It needs lots of baby siting and manual component realigment on the PCB quite often. But if you can live with it, it can be quite productive. We were able to do 300+ boards per day. I'm considering to upgrade to Kayo 6 head system in a year or so. Few shops around has purchased those, so I'll be able to get honest reviews about it.

I wanted to ask your opinion about semi automatic paste printer from Kayo (or any other vendor):
http://en.kayosmt.com/prod/detail/40.html
They've quoted it for 2390 USD, plus shipping. It looks quite not expensive, considering there are manual printers for 2-4k in the market. Would be such printer good fit for few hunder pcb's per day operation? Maybe it's not precise enough and would be total waste of money? I'm looking a way to improve paste printing quality for small footprints, and maybe speed up process a bit. Thank you for opinions.

[rx8pilot]

They've quoted it for 2390 USD, plus shipping. It looks quite not expensive, considering there are manual printers for 2-4k in the market. Would be such printer good fit for few hunder pcb's per day operation? Maybe it's not precise enough and would be total waste of money? I'm looking a way to improve paste printing quality for small footprints, and maybe speed up process a bit. Thank you for opinions.

That is a lot of hardware for the money. Of course that says nothing of what it may be like to use it. Visually, it looks pretty serious but it is the details you cannot see that will determine whether you love it or want to drop it into the bottom of the ocean.

[tautech]

Went to see a client today with a cute little PnP setup in his office. (thanks Mark)
It can do 300mm² PCBs and is self contained not needing any external air or vacuum supply.
NeoDen brand. ~$7k NZD

IIRC 32 feeders.
Issues, really only the SMD card working loose after periods of operation on the table that could be more rigid.

Upgrades from 'as supplied":
Keyboard, mouse and monitor although it can be driven and managed from the touch screen.

Overview


Platen


OS


We discussed the 'getting up to speed' issues and the common traps for novices and most have already been mentioned in this thread, eg: collisions, feeder tape, alignment and components left out for hand placement prior to reflow. SMD electrolytics where hand placed.
He had his own componentry management system/library that mainly grew from issues managing the Z axis variations of even very similar components.
These pics demonstrate his component management style.

Footprint values


Footprint list.

[Reckless]

This is not my pick and place lifestyle, but I am drooling for just a moment.....




Now....back to our regularly scheduled programming.....

I bought that machine for $4500 (mine is dual gantry so has 2 of those heads for twice the speed) but I had to replace a $1000 servo and clean a virus.  Not running it yet as I need 60 nozzles and some more feeders.

I just saw this thread.  It took me ~6 months to get up and going.  Like others I contemplated chinese machines, eventually I bought a used universal instruments genesis dual gantry 7 nozzle head linear motor machine off ebay for $4000 (cost new $300k).  When it arrived one of the axis wouldnt origin due to encoder being off axis and safety switch being broken, I spent 1 month solving.  Then the unit wouldnt read my fidicuals due to having 5mm edge instead of 3mm.  Lost one month trying to learn how to pop coordinates eventually redid CAD data to be more readable.  Lost 1-2 months over bad nozzles/feeders, also learning how to configure component/feeder/board data.  Had to purchase $6k of gold feeders and $500 worth nozzles.  I bought a used DEK automatic screen printer for $450 (cost new $250k) and an essemtec 3 zone oven with edge rail for $800 (new $30k).  Got some smema conveyors for $200.

[rx8pilot]

I bought that machine for $4500 but I had to replace a servo.  Not running it yet as I need 60 nozzles and some more feeders.

Holy crap!!
I don't think I have the stomach for the maintenance on that beast. My little single nozzle machine is already a handful - either expensive or time-consuming or both depending on the maintenance needed.

[Reckless]

Surprisingly ultra low maintenance according to everyone who has one.  The spindles need to have dust removed, universal recommends once a month but some people never do it, others once a year.  I think thats why people love universal for its super reliability.  Like you, I drooled heavily over that machine (gc-30, I got the gc-60).  My other machine GI-14D flex mounter is its sister (single gantry gi-07 shown in above video) often sold together for cell phone manufacturing back in the day. 

I have heard a few people describe it as having no maintenance machine.  One guy had to replace a $400 spindle in 5 years without doing proper preventive maintenance. 

[jmelson]

I bought that machine for $4500 (mine is dual gantry so has 2 of those heads for twice the speed) but I had to replace a $1000 servo and clean a virus.  Not running it yet as I need 60 nozzles and some more feeders.

Wow, that is insane.  I do boards with a wide mix of parts, from 0603 passives up to 20mm square FPGAs, and a lot of SOIC-16 and similar types.  Not sure the Universal handles such a wide variety of sizes.  The big wheel with the vision camera at the top is REALLY smart, though!

Well, I have a Philips CSM84, it does about 3000 PPH in real usage, and that is enough to keep me busy.  It is built like a tank, I've been running it for close to 12 years now in very intermittent use, with only a few problems.

Jon

[Reckless]

That machine is a chip shooter.  It's sister machine is a flex mounter.  Universal was most famous in the industry for making the best flex mounters back in the 90s.  When all the japanese units could only handle 3-5mm tall components they were able to handle 15-20mm.  I need speed in my application so I prefer the turret machine gun style head.  The GC-60 does 57,000 cph which I feel is too slow.  We are debating whether to chain the GC-60D with GI-14D to get 86,000 cph.  My tallest component is 3mm.  I have been trying to figure out a way to have radial components fed through a feeder (Universal makes one called the Radon but only works in new machines and they are pricey). 

[HHaase]

Ahhh, the Universal Lightning.   My previous employer had one of them, and I was in charge of maintenance on it.   Really a great setup and as you said, comparatively trouble free.   Feel free to tap me as a resource.   Compared to the Fuji CP6/IP3 lines, people hated working on the Universal line,  they said it was boring because nothing ever happened.  It just spit boards out all day long.   The OVEN had more trouble than the placement machine.

I'll give a longer post on my own P&P life in a bit.   But my TVM802A is placing away behind me, and it may not be loud but still disturbs my thought process when I'm internetting.

-Hans

[SMTech]

I think would be better if you could show your work for the high qualified specialists. I think you would get much more recommendations from them than here 

Ahh for a thumbs down button.

[PixieDust]

[GUESSING] I have a couple of hundred hours total fiddling, fixing, and learning on top of the effort to just setup and run the system.

You would be surprised how much less time it actually took you learning about the machine. I track my time and I'm usually 2 to 4 times out in terms of estimated time something took and reality, i.e. whatever felt like an hour was usually just 15 mins. My experience working with people on various projects and they were usually 2-4 times out on how long something took as well.

I'm at the very beginning of my electronics journey and all I can say is that there is no chance that I'll buy a Chinese machine, no way, that is if I get a PnP machine at all. I'm still poking away at many things and still have many unknowns. If I decide to go the PnP route, I think I'll build a custom made thing, even though it will probably take around a year to make I'm still confident that the end result will be much better than buying Chinese.

[rx8pilot]

I think I'll build a custom made thing, even though it will probably take around a year to make I'm still confident that the end result will be much better than buying Chinese.

This is one of my favorite DIY pick and place stories....endless lessons to be learned in the dark art of PCB assembly.
https://www.youtube.com/c/StephenTheRobot

[jmelson]

I think I'll build a custom made thing, even though it will probably take around a year to make I'm still confident that the end result will be much better than buying Chinese.

Well, I have only used name-brand "professional" machines, bought used.  I have had 2 machines, a Philips CSM84, made by Yamaha, similar to their CM84, and a Quad QSA30A, made by Samsung on their basic CP30, but with Quad electronic feeders and component vision cameras.
If you have never run a P&P machine, you will not appreciate the amount of attention to error recovery required!  My QSA30A inspects each part for X, Y and height dimensions, and discards parts that don't measure right, as that probably means the part was picked up sideways on the nozzle.  Even my old Philips machine would try to pick parts 4 times before stopping and asking for manual recovery.  The QSA will try 3 times, and then give up on that part, but will remember that that particular part type has not been placed.  Once it finishes the board, you can reach in and clear the feeder jam or whatever the issue was, and hit start, and it will place those skipped parts.  Pretty neat!  Also, there is a LOT of setup testing help built into the system.  When setting up the fiducials, there is a "scan test" button, and it will go the the location of each fiducial and then image it and tell you the offset from the nominal location, and wait a second for you to view how it located the centroid.  There is also a test pick button, that allows you to pick up a part and it will run the vison tests and tell you what the part size was and how far off from the center of the nozzle.  Also, you can use the down-looking camera to check and adjust the pickup position of every part, to improve pick-up reliability.  All of these software features probably took a while to develop.

Jon

[mikeselectricstuff]

The QSA will try 3 times, and then give up on that part, but will remember that that particular part type has not been placed.  Once it finishes the board, you can reach in and clear the feeder jam or whatever the issue was, and hit start, and it will place those skipped parts. 

I'm amazed that this isn't a standard feature on all P&P's - it's such an obvious way to do it - the whole point of a P&P is that it will do as much as possible of the job by itself without babysitting.

[SMTech]

The QSA will try 3 times, and then give up on that part, but will remember that that particular part type has not been placed.  Once it finishes the board, you can reach in and clear the feeder jam or whatever the issue was, and hit start, and it will place those skipped parts. 

I'm amazed that this isn't a standard feature on all P&P's - it's such an obvious way to do it - the whole point of a P&P is that it will do as much as possible of the job by itself without babysitting.

Right, but it should also be smart enough to know it shouldn't place something that will then get in the way when this other part is placed out-of-order but this is often called collision avoidance and charged as an optional extra.

[jmelson]

The QSA will try 3 times, and then give up on that part, but will remember that that particular part type has not been placed.  Once it finishes the board, you can reach in and clear the feeder jam or whatever the issue was, and hit start, and it will place those skipped parts. 

I'm amazed that this isn't a standard feature on all P&P's - it's such an obvious way to do it - the whole point of a P&P is that it will do as much as possible of the job by itself without babysitting.

Well, you have to realize I started on a 1995 Philips CSM84, and then upgraded to a Quad QSA30A, made in about 1999 or so.
The feeders on the Philips were really bad, the cover tape pull was VERY weak, and I had to choke up on springs to try to get more pull.  So, I had to babysit the machine all the time.  The Quad has feeders that work a LOT better, so I can generally walk away and do something else.
Jon

[jmelson]

Right, but it should also be smart enough to know it shouldn't place something that will then get in the way when this other part is placed out-of-order but this is often called collision avoidance and charged as an optional extra.

None of the machines I've worked on (only 2) have this kind of sophistication.  Collision avoidance is a VERY complex issue, it VASTLY complicates the motion planning.  It is easier to just raise the nozzle higher.
Jon

[mikeselectricstuff]

The QSA will try 3 times, and then give up on that part, but will remember that that particular part type has not been placed.  Once it finishes the board, you can reach in and clear the feeder jam or whatever the issue was, and hit start, and it will place those skipped parts. 

I'm amazed that this isn't a standard feature on all P&P's - it's such an obvious way to do it - the whole point of a P&P is that it will do as much as possible of the job by itself without babysitting.

Right, but it should also be smart enough to know it shouldn't place something that will then get in the way when this other part is placed out-of-order but this is often called collision avoidance and charged as an optional extra.

That's easy enough if you can specify the retry behaviour - collision avoidance isn't going to be needed that often unless you have tall parts and a limited nozzle height range, or a very dense board with some tall parts,  so on the occasions it is, you should be able to tell it to halt on retry fail.

"Place as much as you can then tell me about any issues when you're done" should certainly be the default behaviour. Unfortunately many of the Chinese manufacturers don't seem to understand this

[rx8pilot]

My Quad has the Windows software which simply stops when the retry limit is reached. Sometimes this is good, but mostly I wish it would just continue and provide me a list of errors at the end.

If I am doing 0805/0603 I can generally walk away. Once I get into 0402 and 0201, I have to stay pretty close and expect to tweak feeders regularly. Over time, I have identified the better feeders in my inventory and those are the ones that get used for the small parts.

At the moment....none of this matters since I cannot even buy the parts I need for my existing products or the new designs. 

[48X24X48X]

"Place as much as you can then tell me about any issues when you're done" should certainly be the default behaviour. Unfortunately many of the Chinese manufacturers don't seem to understand this

On my Chinese HWGC machine, you can set the retry attempt count to as many as you want or no retry at all. At least in their latest software version I can now mute the buzzer alarm (it's annoying) when an error happens and it gives you options to retry, skip this time around, skip entirely for that particular feeder or abort the entire job completely. I also did suggest what you mentioned here to them. I think over time, they might just add these features, hopefully.

[mikeselectricstuff]

"Place as much as you can then tell me about any issues when you're done" should certainly be the default behaviour. Unfortunately many of the Chinese manufacturers don't seem to understand this

On my Chinese HWGC machine, you can set the retry attempt count to as many as you want or no retry at all. At least in their latest software version I can now mute the buzzer alarm (it's annoying) when an error happens and it gives you options to retry, skip this time around, skip entirely for that particular feeder or abort the entire job completely. I also did suggest what you mentioned here to them. I think over time, they might just add these features, hopefully.

I wouldn't bet on it - having it place as much as possible then fix up afterwards will usually mean a fundamental change to the way the data is handled, i.e. it needs to have a "placed" marker for every part instead of just remembering how far through the file it got. It's really something that needs to be designed in from the start

[48X24X48X]

In the current version, it already has logging on failed placement including a screen capture of the vision with all the parameters. I think it's more on whether they want or don't want to do it. But I have seen the software progress so much over a span of 2 years, so I still have a tiny hope of that.