Author Topic: Advice on setting up a low-mid volume SMT line for <$130k?  (Read 66397 times)

0 Members and 1 Guest are viewing this topic.

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6459
  • Country: nl
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #175 on: July 26, 2017, 02:14:40 pm »
If I use stencil paste for syringe I have to adjust the pressure and vacuum, temperature of the paste is also important, too warm and it starts oozing.

Thinking further the longer lasting paste probably has used less volatile solvents?
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2008
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #176 on: July 27, 2017, 07:16:07 pm »
I agree, but in my position and without experience running these lines I think I should err on the side of having local support.

If things turn out well I will be able to get 2 PnP machines, a 2003 Opal XII and a 2007 Assembleon MG-8, along with several hundred assorted feeders, for ~$80k. I'm trying to figure if this is a good deal or not. It seems marginal although if both machines are in excellent condition and the feeders work, I think it's not so bad. Then again a brand new Samsung SM482 is $70k from Korea, but then I'll spend another $20k+ on feeders alone. It is a faster machine although not quite as flexible as the MG-8. Decisions, decisions...it seems to me I should pay the slight premium to get the support access. If it all goes well I'll be in a much better position in the future to evaluate machine purchases.

It sounds like you're pretty set on getting the Assembleon machines - and there's a LOT to be said for having local support, especially when you are totally new to the PnP world.  However... various machines differ greatly in how easy they are to set up and change over, both in terms of the time it takes to change a feeder out, or to change a reel from one component to another in an existing feeder, or to change the program in the machine, or to program the machine for a new board.  If you are intending to offer assembly services, then the time it takes to set up the machine is going to be one of (if not THE single) biggest time sinks and costs.  Over the years, pretty much everyone I've ever talked to that has run these machines has said MyData machines lead the way in how fast they can be set up and switched between jobs. 

I have no connection to them and don't own one - but I've owned several PnP's over the years and some of them can be a massive PITA to set up or change over the jobs.


As for shipping... unless you are paying for an installation service, I would almost guarantee the terms are going to be FOB, so as soon as it leaves the dock of the seller - you have no guarantee from them.  The shipping company would only cover obvious physical damage upon arrival, and if you have a rigger unloading it, their insurance would cover things like them dropping the machine or boinking it against a doorway as they move it in.  The big concern is always.... you get the machine in position, plug it in, run air to it... turn it on... and... nothing.  Or, you get a whole bunch of error messages indicating some non-trivial problem.  Unless you've paid for installation service and a turn-key setup (where the seller is providing shipping, move-in, installation, set-up, etc), then you're going to be on your own.  Control boards jostle loose in shipping, cables come loose, circuit boards with dodgy connections can fail from the vibration of road trips.  But a shipping companies insurance isn't going to cover a machine that was claimed to work when shipped and doesn't when it arrived and was set up.

Just an FYI.

I am one of the horror stories mentioned above.  I bought a couple of *really* cheap chipshooters from a guy in California once.  He assembled DDR memory boards for Micron and had 2 new Fuji chipshooters coming and needed the old ones gone right away.  He was letting them go for practically free.  I had no time to arrange everything and the long and short of it is that when the machines arrived, they had been transported cross-country on an OPEN truck, without even a tarp.. through a couple of rain storms.  They had been shoved into a box truck at the sellers end, then dragged out with chains at a terminal, and worse.  When they arrived, the sheet metal was bent to hell, there were boxes of parts that had fallen out of the machines shoved inside them, the frames were covered in dents and gouges - just a total mess.  I eventually managed to get one machine working from the two... and luckily the parts on the machines were so valuable (motors, vision system parts, sensors) that I parted them out and got about 20 times back what I paid.  But the shipping company - who I had a written contract stating they were to be shipped in an enclosed air-ride truck - basically told me "Ok, sue us...".  They were in another state.  So be careful.

It's not always the most popular person who gets the job done.
 
The following users thanked this post: ar__systems, SVFeingold

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2008
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #177 on: July 27, 2017, 08:42:41 pm »
Sorry for the late reply, but really want to reply to this one.

This is my experience as well.    We've been assembling in-house for this reason for at least 10 years now.

5-6 years ago we had one of those years that we needed to spend some money at the end of the year.   Ended up spending $80K for a complete line + dry box from manncorp, which allowed us to switch from 100% through hole to SMD.   It's been one of the best purchases I've made.   If I was in the market for a second machine, I sure wouldn't count them out (although I would definitely make sure I evaluated the entire market).   Like any purchase there are trade-offs, but for the relatively lower tech boards (0.5mm pitch/0402 and larger), this has worked well for us, and the pick and place machine has been remarkably trouble free.

As we've slowly increased our volume, I continue to look at outsourcing assembly of some of the products.   So far, I haven't been able to find anyone anywhere who can manufacture the volume we do (500-1000pcs) for the cost

This is something I want to address too...

A PnP machine that can realistically place 10,000cph is quite standard these days.  And when I see prices of $0.10/pad for SMD work, even if you assume 2 pads per component, that's $0.20 per placement, or $2,000 per hour.  That is an INSANE price point.  They're trying to pay off the machine in a couple of 40-hour shifts (including generous set-up time).   For something like a CNC machine, an extremely (extremely!) lucrative job would have the machine paid off within perhaps 6 months, but a more realistic timeline is 2-3 years.  For a PnP machine that costs $250k new and can achieve a realistic 10,000cph throughput, if the machine can be running 75% of the time (assuming 25% set-up and idle time), then on a 2,000 hour year (which assumes a standard 40-hour work week and 2 weeks off, which for a machine would be downtime/maintenance/repair time), that would be 1,500 hours of placing time per year.  To pay the machine off within a year, that's $166/hr, or 1.6 cents per component placement.

The prices I have seen are more like 5 to 10 cents per SMD *pad*, and the average # of pads per component necessarily has to be higher than 2, probably even above 3.  So these US-based outfits are looking for a princely sum to run their PnP machines.

Which makes me think they either have massive overhead.  Or they just are not getting a lot of work and are trying to spread the costs over a small number of customers.  Or they maybe have a few established clients for whom costs don't really matter. 

When you consider the cost of a $20k to $75k used PnP machine, the cost equation gets even more unfavorable for the assembly houses.  Some folks may think that they are offering a level of quality control for your money that you don't get by bringing the work in-house, but my experience has generally been the opposite.
It's not always the most popular person who gets the job done.
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #178 on: July 27, 2017, 09:02:24 pm »
Great points about the shipping. I'm not really sure what else to do though. I'm being as attentive as I can to shipping with moisture bagging, skidding, shock mounting, sensors for climate/shock/etc... That and finding a good rigging company in Atlanta. Got a recommendation from a friend that works with heavy
equipment and I'm meeting with them on location Monday.

As long as they are shipped the way they say they will be and unloaded gently, I don't foresee anything catastrophic happening. If something does break well...that's why we left a healthy buffer in the budget for emergencies.

Assembleon recommended to use the shock absorbing feet for the machine during transit, 2 machines so 8 feet total. Only they want $500 for each one. What...it's a metal plate with 2 rubber cylinders bolted to it. Nuts. I could take care of that easily if I had physical access to the machines. We'll see, the seller may have the mounts from prior shipping.

I agree on the insane pricing. This is the experience of many, many startups.

I looked at MyData but I've heard about a lot of issues with those as well, whereas the Assembleons seem to be workhorses. As for setup time, I have no idea. I read through the Assembleon operators manual and it seems reasonable. These machines don't have the optional RFID reader for the feeders, which sounded super handy. That's a shame but I don't know if it's worth delaying another 1-2 weeks to find another machine that does. And even if we do, it isn't going to come with 500+ feeders, offline programming tool, feeder calibration tools, etc...

Also throwing in a CR Tech RTI 6250 AOI system, which probably isn't worth much but it's basically free. May be fun to play with.

 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #179 on: July 27, 2017, 09:36:50 pm »
As for setup time, I have no idea. I read through the Assembleon operators manual and it seems reasonable. These machines don't have the optional RFID reader for the feeders, which sounded super handy. That's a shame but I don't know if it's worth delaying another 1-2 weeks to find another machine that does. And even if we do, it isn't going to come with 500+ feeders, offline programming tool, feeder calibration tools, etc...
RFID for the FEEDER?  What the heck does that accomplish?  You can still load the wrong reel of parts into the feeder, and still make a run of boards that need to be reworked.

Well, as I mentioned before, I whipped up a really rough program that created the placement file from the CA/CAM software's placement file in a couple days.  Now, when I do a new board, it is a matter of assigning feeder locations for each part and running the program.  It takes about 30 minutes, tops to do that, even for a complex board.  Setting up  different feeder setup (I do not have carts for my machine) is a bit of an issue, but rarely more than one hour.  The vision system may be more complicated, but once you have created the vision library, you just call out each vision template for specific parts, and only have to do anything special when a new footprint is used for the first time.  You also have to set the conveyor rails and the push-in and push-up pins, that takes 5 minutes, max.

For panels, you need to know the offset between board instances and the location of the bad marks.

So, the total changeover for a first time run of a board is maybe 2-3 hours, including the CAD/CAM prep.  To changeover to re-run a board you have already set up, it is rarely more than an hour.
Sometimes it takes me longer to order the parts online than to do the actual setup on the machine!

Jon
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13695
  • Country: gb
    • Mike's Electric Stuff
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #180 on: July 27, 2017, 09:49:41 pm »
As for setup time, I have no idea. I read through the Assembleon operators manual and it seems reasonable. These machines don't have the optional RFID reader for the feeders, which sounded super handy. That's a shame but I don't know if it's worth delaying another 1-2 weeks to find another machine that does. And even if we do, it isn't going to come with 500+ feeders, offline programming tool, feeder calibration tools, etc...
RFID for the FEEDER?  What the heck does that accomplish?  You can still load the wrong reel of parts into the feeder, and still make a run of boards that need to be reworked.

But once you know the right parts are there, you can swap feeders around for different jobs easily, assuming you have enough feeders that you don't have to unload them to free them up for new jobs. 
Larger operations can also automate some of the process, e.g. by scanning a barcode on the reel at a loading station that also reads the RFID tag.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #181 on: July 27, 2017, 10:41:11 pm »

But once you know the right parts are there, you can swap feeders around for different jobs easily, assuming you have enough feeders that you don't have to unload them to free them up for new jobs. 
Larger operations can also automate some of the process, e.g. by scanning a barcode on the reel at a loading station that also reads the RFID tag.
Hmmm, well, I just don't have enough feeders to do anything like that.  AND, I wouldn't have a place to STORE all those feeders if I DID have them.  Maybe in some huge contract assembly shop with a bunch of guys keeping several lines running 24/7 this could be a big deal.  But, in a smaller shop, just being careful and double-checking each time you put a reel on a feeder or a feeder on the machine will prevent any mishaps.



Jon
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #182 on: July 27, 2017, 10:46:58 pm »

Also throwing in a CR Tech RTI 6250 AOI system, which probably isn't worth much but it's basically free. May be fun to play with.
[/quote]
Is this for paste inspection or parts inspection post reflow?  I can see a paste inspection system, I find defects when I stencil print my boards, and it would be a lot worse with a big panel. 
With a P&P, wrong parts is pretty unlikely.  I occasionally find a part that went on sideways, but it is pretty rare.  I think I've now had my 3rd tombstone after making over 1000 boards.  So, that is rather a rarity.  If the AOI can pick up soldering defects, that would be good, but just to locate wrong parts, I think that you'd turn the thing off after the first couple days of it never finding anything.

Jon
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #183 on: July 28, 2017, 12:18:59 am »
Supposedly it can find solder defects, tombstones, etc...

http://www.firstplacemachinery.com/firstplace_CRTechnology_RTI6520.html

Still looking for a datasheet. This is a little bigger than I expected it to be...not sure where I'm going to put the thing. Maybe I can part it out to some interested engineers nearby, if nothing else.

For the RFID readers, yes it's not strictly necessary...would be a cool feature. Getting several hundred feeders so I shouldn't be running out for a while...if I can just figure out where to put them.
 

Offline janekm

  • Supporter
  • ****
  • Posts: 515
  • Country: gb
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #184 on: July 28, 2017, 01:55:37 am »
(snip)
A PnP machine that can realistically place 10,000cph is quite standard these days.  And when I see prices of $0.10/pad for SMD work, even if you assume 2 pads per component, that's $0.20 per placement, or $2,000 per hour.  That is an INSANE price point. 
(snip)

Which makes me think they either have massive overhead.  Or they just are not getting a lot of work and are trying to spread the costs over a small number of customers.  Or they maybe have a few established clients for whom costs don't really matter. 

(snip)


Ball-park local pricing in China in volume is 0.0015USD per pad. So yes those prices you've been getting quoted are insane, by two orders of magnitude! Of course you have to add on setup and testing fees on top of that but still...
Of course the local Chinese pricing is not accessible directly for someone from the US (you need people on the ground to handle the day-to-day communications) but you should be able to find something in the middle.

I believe a big reason why the US-based (and also Europe-based) outfits tend to be so expensive is because they have been pushed into the very high-end market by market forces. If you need a big volume of cheap boards you know where to go. So what I've seen from the UK operators I've dealt with for example is a very high expectation of "get it right first time" and the additional communications and setup costs that go with that. Together with various certifications and other practices expected from high-end (military and medical) customers that can afford to pay their prices.
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #185 on: July 28, 2017, 02:39:01 am »
The problem is that so far, in my experience, they DON'T get it right the first time. In the time it took for Advanced Circuits to assemble ~200 boards for a client - which were then shipped with a missing part (which was both on the BOM and ordered), I could have assembled them all by hand. And AC is NOT cheap.

If you're a startup or small business, you are kind of screwed unless you're very adept at navigating all this. $0.0015/per pad might apply for high volume but I've never seen anything remotely near that for low-moderate volumes (few hundred to several thousand). Even in China. Those quantities are not high enough to get real volume benefits (you're just starting to get there) nor are they low enough to do in-house easily.

While I grant that $0.01/per pad might be crazy compared to the lowest cost Chinese PCBA for an order of 500k, don't you think that would be attractive to a lot of small businesses in the US?
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2008
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #186 on: July 28, 2017, 07:22:37 am »
Great points about the shipping. I'm not really sure what else to do though. I'm being as attentive as I can to shipping with moisture bagging, skidding, shock mounting, sensors for climate/shock/etc... That and finding a good rigging company in Atlanta. Got a recommendation from a friend that works with heavy
equipment and I'm meeting with them on location Monday.

As long as they are shipped the way they say they will be and unloaded gently, I don't foresee anything catastrophic happening. If something does break well...that's why we left a healthy buffer in the budget for emergencies.

Assembleon recommended to use the shock absorbing feet for the machine during transit, 2 machines so 8 feet total. Only they want $500 for each one. What...it's a metal plate with 2 rubber cylinders bolted to it. Nuts. I could take care of that easily if I had physical access to the machines. We'll see, the seller may have the mounts from prior shipping.

I agree on the insane pricing. This is the experience of many, many startups.

I looked at MyData but I've heard about a lot of issues with those as well, whereas the Assembleons seem to be workhorses. As for setup time, I have no idea. I read through the Assembleon operators manual and it seems reasonable. These machines don't have the optional RFID reader for the feeders, which sounded super handy. That's a shame but I don't know if it's worth delaying another 1-2 weeks to find another machine that does. And even if we do, it isn't going to come with 500+ feeders, offline programming tool, feeder calibration tools, etc...

Also throwing in a CR Tech RTI 6250 AOI system, which probably isn't worth much but it's basically free. May be fun to play with.

I didn't mean to suggest you were doing anything wrong on shipping - just that some of your comments such as "I won't sign off on anything until the machine is on my floor and I verify it works" (paraphrasing) are unrealistic, IMO.  The seller's responsibility will end when it leaves their dock, the shipper's will only be for damage in transit, and the rigger too.  As long as you know the machine is good as-sold and arrives in good physical, I wouldn't worry too much.  It's common to have issues and mostly they are due to cables that have jostled loose or driver boards that have worked loose from back plane connectors.

As for protecting the machines, $500 per vibration isolation foot is utter robbery, and I would be hard pressed not to be downright angry at someone quoting me that price!  Go to McMaster.com and look up vibration damping pads, vibration control levelers, and adjustable air springs.  The last of those can get expensive, especially from McMaster but if you look up Enidine Air Springs on eBay, you can get them cheaper.  The anti-vibration feet are more for when you get the machine in your shop, not for transport.  Air springs are the ones you want for transport.  You could also buy a pallet or make a skid with the air springs and vibration damping mat pre-attached to it, and ship that out to the seller - wouldn't cost too much, and just have them load the machines on that and strap them down.  Then you will have peace of mind.  And for the kind of $$ you're talking about, it may be worth a plane trip out there to see the machines in-person and be there while they are prepped for shipment. 

On the machine itself... are you buying direct from the seller?  or is this ground through a broker or is it being re-sold through the Assembleon service folks that are going to support you with the training?  I am not sure if or how much you haggled on the price of the machine, but I've bought and sold quite a few machines and my experience is there's usually a 10-20% increase in what they're asking vs what they will take.  That's especially true for PnP's IMO.  They may seem sort of hard to find from a buyer standpoint, but just imagine it from the seller standpoint.  It's exceedingly difficult to find someone willing to shell out tens of thousands for a used machine, that's especially true for private sellers.

500 feeders will be nice... the only thing you can really count on is that there's no way to know all the "what ifs?" until you take the plunge and go through the trial by fire.  So I wouldn't worry TOO much about buying the absolutely perfect machine.  Until you have some experience under your belt, you won't know what things matter to you a lot (and not at all) in a PnP and in an assembly line.  You'll know that next time around... so this machine will hopefully get you going and if it takes your business to Phase 2, you'll be well informed then to buy the perfect machine.
It's not always the most popular person who gets the job done.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13695
  • Country: gb
    • Mike's Electric Stuff
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #187 on: July 28, 2017, 08:29:03 am »

Ball-park local pricing in China in volume is 0.0015USD per pad.
At what sort of volume ?

I've never understood the reasoning for "per pad" pricing - it takes no more time to place an 8-pin resistor network than a single resistor
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 4067
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #188 on: July 28, 2017, 11:49:58 am »
Placing large pad count components is slower due to the part mass?
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #189 on: July 28, 2017, 02:25:26 pm »
Corporate666: I have spoken directly with the seller but we're doing it through a broker, they're getting all the equipment together for us under one package so that we can finance more easily. Visiting the equipment on rigging day is not such a bad idea.

I'm a little concerned about the price I was given for those foot bumpers...of course there's no way I'm paying that but I hope it's not indicative of the rest of their spare part costs. Luckily my father moves heavy equipment (microscopes) from time to time and may have a bunch of those auto-leveling air supports I can borrow.

The PnP package is: ~85k-90k, still working it out. That includes:

~600 feeders
Assembleon MG-8 (2007)
Assembleon Opal Xii (2004)

All this stuff "for free:"
Off-line programming/teaching tool
Feeder calibration tool
A couple of tray feeders
Ultrasonic cleaner
Ultrasonic vapor degreaser
A bunch of carts/racks
CR Tech RTI 6250 AOI

Seems like a decent package if it all works, but it's definitely getting into serious money for us.

This is also not including the rigging/bagging/shipping which will probably be another $3-4k per machine. What's a reasonable amount to pay for moisture-barrier bagging?
 

Offline ar__systems

  • Frequent Contributor
  • **
  • Posts: 516
  • Country: ca
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #190 on: July 28, 2017, 04:16:11 pm »

Exactly what the feeder calibration tool does?
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #191 on: July 28, 2017, 04:28:54 pm »
I've never understood the reasoning for "per pad" pricing - it takes no more time to place an 8-pin resistor network than a single resistor

Me neither.....

I don't take outside work, but intuitively, I would charge by part placement and break that down into 2 categories - precision and non-precision. The precision parts are anything that needs up facing vision alignment and non-precision covers all other parts that can be aligned on the fly.

Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2008
  • Country: us
  • Remember, you are unique, just like everybody else
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #192 on: July 28, 2017, 05:39:30 pm »
Corporate666: I have spoken directly with the seller but we're doing it through a broker, they're getting all the equipment together for us under one package so that we can finance more easily. Visiting the equipment on rigging day is not such a bad idea.

I'm a little concerned about the price I was given for those foot bumpers...of course there's no way I'm paying that but I hope it's not indicative of the rest of their spare part costs. Luckily my father moves heavy equipment (microscopes) from time to time and may have a bunch of those auto-leveling air supports I can borrow.

The PnP package is: ~85k-90k, still working it out. That includes:

~600 feeders
Assembleon MG-8 (2007)
Assembleon Opal Xii (2004)

All this stuff "for free:"
Off-line programming/teaching tool
Feeder calibration tool
A couple of tray feeders
Ultrasonic cleaner
Ultrasonic vapor degreaser
A bunch of carts/racks
CR Tech RTI 6250 AOI

Seems like a decent package if it all works, but it's definitely getting into serious money for us.

This is also not including the rigging/bagging/shipping which will probably be another $3-4k per machine. What's a reasonable amount to pay for moisture-barrier bagging?

I haven't paid a specific extra charge for moisture bagging, so I can't tell you the upcharge on that.  I would say, if at all possible, definitely go there for the prep of the machines for loading.  Just consider what will happen if the machines arrive and are damaged.  For the sake of $500 (round trip flight + hotel room + Lyft rides), it's a tiny price to pay to protect your $90k investment.  The shipper will have nowhere near the incentive you have to make sure the machines are well prepared for shipment.  $3-4k per machine isn't unreasonable from door-to-door.  I usually pay $1k to get a CNC machine unloaded from a truck and placed in my shop... those guys have heavy equipment and expensive insurance and have to haul their stuff to every job, although the price for the second machine should be fractionally more than for the first one.

If you're concerned about parts costs - call up Assembleon and ask for a price quote for a nozzle or something.  Or ask the seller to give you a list of what parts have been replaced on the machines in the past couple of years.  Have you called the local (to the seller) Assembleon service center and asked about the specific machines?  They surely must know them and should be able to provide you with a service history or something.  And since you and the seller are in contact... hopefully you're keeping the broker's fees reasonable... hope he/she isn't pocketing too much on the deal.  Financing is dead simple... I have a few folks I've worked with if you want some contacts, financing up to $250k is basically just a personal credit check and a FICO of 700+ away. 

RE: all the "free stuff" that goes with the machine, it's certainly good to get that stuff, although I remember when I sold my Fuji machine I "gave" all that stuff to the buyer as well.  He was so happy.  I am overly honest and I told him all this stuff was largely specific to that machine/line and useless to anyone else.  It took up a lot of space and if he didn't want it, then it was going in the dumpster or on the curb with a 'free scrap metal' listing shortly thereafter :) 

It's not always the most popular person who gets the job done.
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #193 on: July 28, 2017, 06:43:28 pm »
Placing large pad count components is slower due to the part mass?
No, it may be that small passives are placed blind, but anything with more than 2 leads has to use vision, and that slows the machine down.  Or, maybe the larger parts have more mis-picks or something.
Possibly, this is an added factor for manual programming.  But, I can't imagine many people manually program a P&P.  I have never done it, even my VERY FIRST board was programmed from the CAD/CAM data.

Jon
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #194 on: July 28, 2017, 06:44:20 pm »
If I have to throw it away then no big loss, provided it's actually being thrown in without an upcharge. The AOI I can bring to work and let the engineers tear it down, lots of cool stuff in there they'd be interested in.

And yes you're absolutely right. My advice to people doing their first production run in Shenzhen is always "Why aren't you already on a plane to Shenzhen?" Stands to reason the same advice applies here. :)

The service tech that serviced these machines is actually 30 minutes from me (Yamaha). He flies up to Ontario to service them. According to him there's been nothing untoward aside from regular maintenance intervals, mostly head re-builds to replace/lube o-rings and seals as preventative care.

As I mentioned there are actually two service centers 30 minutes away from me. One is K&S (Assembleon) and the other is Trans-Tec (Yamaha). Both can service this machine since they were both made by Yamaha and sold by them under one name and by Assembleon under another.

My assumption, with the term "dedicated air-ride truck," is that equipment is loaded on the truck at the seller, and it does not leave the truck until it arrives at my location in Atlanta. I.e. it's not going through any terminals, warehouses, etc... Strictly point A to point B. Is that a safe assumption or no?
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #195 on: July 28, 2017, 06:47:27 pm »
Placing large pad count components is slower due to the part mass?

Could be a simplified way to pass reel handling/labor costs to the customer. Stands to reason that more pads means either more BOM items (requires additional feeders or even an additional machine for the overflow) or higher pin counts (can fit fewer parts per reel). A reel might be able to hold 60k resistors but only 2k QFN packages. Or a JEDEC tray for large pin count parts which may only hold 50 parts. More labor involved.

That's what makes sense to me anyway, doesn't necessarily mean it's the real reason.
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #196 on: July 28, 2017, 06:49:32 pm »

~600 feeders

Hmmm, my suggestion is you select the 100-150 best feeders, taking plenty of the larger size ones (12, 16, 24mm) and then sell the rest.  From earlier discussion, you are not going to have room for them, unless there is some other storage area available.  Having a fair number of feeders here (Just barely enough, actually, for our most complex board) I can't IMAGINE what dealing with 600 feeders will be like.  It will DEFINITELY fill your entire 500 sq. ft. room.  They are also fairly fragile, so you can't stack them to the ceiling.

Jon
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2758
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #197 on: July 28, 2017, 06:57:01 pm »

Exactly what the feeder calibration tool does?
The feeder pick-up locations are generally pre-programmed into the machine.  It can be taught the pick-up location for one specific feeder on the rail, and then all other feeders are computed from that one.
So, if you work that way, each feeder should bring the part to the same location, relative to the dowel pin holes that the feeder mounts to.  I have a home-made tool that mostly came from the guy I bought my machine from, I added a bridge over the feeder with a cross-hair to view the component pocket.  Sometimes dirt or dings cause the feeder to tilt, causing the pocket to end up left/right of the expected position.  Or, the sprocket and ratchet assembly drifts a bit, and the pocket is forward or backward of nominal.  The calibration tool allows you to see this error and adjust things until the pocket is properly centred.

Jon
 
The following users thanked this post: ar__systems

Offline Pseudobyte

  • Frequent Contributor
  • **
  • Posts: 283
  • Country: us
  • Embedded Systems Engineer / PCB Designer
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #198 on: July 28, 2017, 07:21:12 pm »
In my company we love our AOI machines, fantastic! However just know that they will only ever be as good as the person who runs it. These guys were amazing, http://www.aoisystems.com/Scanner-Based-In-Line-Desktop-AOI.html they will come to your business if you would like and teach you the process around doing AOI right. We couldn't be happier after we bought our first one. Very easy to program due to the component database that they maintain. They are basically big high resolution flatbed scanners that were inverted, nothing special, but very good software. 9/10 would recommend to a friend.
“They Don’t Think It Be Like It Is, But It Do”
 

Offline SVFeingoldTopic starter

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: us
Re: Advice on setting up a low-mid volume SMT line for <$130k?
« Reply #199 on: July 28, 2017, 07:27:53 pm »

~600 feeders

Hmmm, my suggestion is you select the 100-150 best feeders, taking plenty of the larger size ones (12, 16, 24mm) and then sell the rest.  From earlier discussion, you are not going to have room for them, unless there is some other storage area available.  Having a fair number of feeders here (Just barely enough, actually, for our most complex board) I can't IMAGINE what dealing with 600 feeders will be like.  It will DEFINITELY fill your entire 500 sq. ft. room.  They are also fairly fragile, so you can't stack them to the ceiling.

Jon

The space has been upgraded a bit to 700 sq. ft. That will be ~1000 sq. ft. once the room next door opens up and then I can take that wall down. So it's a little more room to breathe. The feeders should come with carts/racks.

What sorts of other precautions would you all recommend as far as room preparation? So far I have this sequence:

- Install double doors
- Run electrical service (will be done by building owner), 100A or 200A dedicated breaker.
- Clean the hell out of every surface in the room, floors, walls (brick), ceilings
- Coat the floor with static-dissipative paint ($200/gallon x 4 gallons)
- Install HEPA air filtration unit to keep the dust down
- Install de-humidifier
- Install air conditioner
- Sticky floor pads like this for people walking in.
THEN start moving equipment into the room. From the machine specs I will need like a ~5HP air compressor. Not sure this can go outside so if not it will need to be something relatively quiet like a Kaeser. Another several $k right there. Then an air dryer (harbor freight) and filtration.
« Last Edit: July 28, 2017, 07:29:49 pm by SVFeingold »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf