Advice on setting up a low-mid volume SMT line for <$130k?

[janekm]

(snip)
What would be professional about taking someone's stencil design and arbitrarily changing it? If a manufacturer doesn't exactly follow the CAD data I give them, I'll wouldn't use them again. Its good if they query apparent anomolies, but if they just make changes as a standard practice its a recipe for dismal failure. They have no unambiguous way to know what I want or need.

It is very common for stencil manufacturers to adjust apertures (I have experienced this both in the UK and in China). Usually it's for the best as they have experience and software that implements common rules of thumb (for example large pads need to be "window-framed" to avoid the squeegee getting stuck), so if you don't want them to I would recommend putting a clear note on the paste layer that you want it to be cut "as is".

The same arguably applies to PCBM...

[ovnr]

It is very common for stencil manufacturers to adjust apertures (I have experienced this both in the UK and in China). Usually it's for the best as they have experience and software that implements common rules of thumb (for example large pads need to be "window-framed" to avoid the squeegee getting stuck), so if you don't want them to I would recommend putting a clear note on the paste layer that you want it to be cut "as is".

The same arguably applies to PCBM...

No, it really isn't. The only common modification - to both stencils and PCBs - is adding identification data like order numbers. Any manufacturer otherwise modifying a design will not enjoy a lot of repeat customers.

[Kjelt]

Well dived into the datasheet that referred to the design in specifications that was very vague to say the least:

The suggested paste mask layout for the NINA-B1 series modules is to follow the copper mask layout as described in NINA-B1 series Data Sheet [2]. These are recommendations only and not specifications. The exact mask geometries, distances, and stencil thicknesses must be adapted to the specific production processes of the customer.

So they suggest to follow the copper pad for the stencil but you actually have to experiment with the stencil thickness and surface area to adapt to your specific production process, nice disclaimer  .
Thanks for all the advices I will reduce the amount of paste with 25% on the pads manually and experiment further 

Sorry for the small hi-jacking of this thread, please continue on topic.

[jmelson]

I used to agree... however, I spent a long time and a lot of frustration chasing a problem with ATTINY chips in SOIC-8 packages.
<snip>
I've never had any issues with any other parts, ever.  And it was only ATTINY chips from one specific fab plant... so I always felt it was an Atmel problem, but I can't prove it.

Well, I think you have proved it.  So, they may have had excessive porosity in the encapsulation, or some other situation that caused moisture to invade the package.  But, of course, you are left holding the bag.

Jon

[jmelson]

Yes, the trick with anything leadless is to make the paste aperture WAY less than the pad size.  When the leadless part pulls down against the board, it squeezes the solder out, causing shorts.  I think there are articles about how to figure the factor of aperture reduction.
Jon

Thanks, I hoped that the pro pcb and stencil factory would have taken care of this, but I guess they just take the pad size as the stencil size, so I have to make two pcb layouts one for the pcb and one for the stencil, does not make much sense 

No, this is all left to the PCB designer, unless a contract manufacturer also generates the stencil, then they would know to reduce the apertures to fit their process.

Jon

[jmelson]

Well dived into the datasheet that referred to the design in specifications that was very vague to say the least:

The suggested paste mask layout for the NINA-B1 series modules is to follow the copper mask layout as described in NINA-B1 series Data Sheet [2]. These are recommendations only and not specifications. The exact mask geometries, distances, and stencil thicknesses must be adapted to the specific production processes of the customer.

So they suggest to follow the copper pad for the stencil but you actually have to experiment with the stencil thickness and surface area to adapt to your specific production process, nice disclaimer  .
Thanks for all the advices I will reduce the amount of paste with 25% on the pads manually and experiment further 

0805 and even 0603 passives can be done at 100% aperture size with little trouble, but you will get some displaced solder beads.  SOIC can also generally be done without reduction.  SSOP and other chips with .65mm lead pitch need a reduction to maybe 75% pad area, and .5mm lead pitch require reducing the apertures to about 50% or something like that.  These numbers are for my process using .003" (approx .075mm) stencil thickness.  You may have to adapt this for commercial stencils.  I etch my own stencils out of brass shim stock, the etching process likely makes them come out larger than drawn.  (Hmmm, someday I really ought to measure the apertures and find out how much they grow!)

Jon

[Corporate666]

Well, I think you have proved it.  So, they may have had excessive porosity in the encapsulation, or some other situation that caused moisture to invade the package.  But, of course, you are left holding the bag.

Jon

True, however it wasn't a single run of chips... this was an ongoing problem over a period of years with chips purchased from Digikey, Mouser and Arrow.  Atmel were made aware of the issue and it continued to happen for years afterwards.  I took the time and expense to collect tons of failed chips and record the storage, usage and reflow parameters and then sent samples of each batch of failed chips to their HQ (was it in Denmark?  I forget) at my expense via FedEx for them to decap and analyze.  I had to hound them to do so - and eventually they basically just gave me a very curt reply of "bad process control, the chips have popcorned due to moisture - we are Atmel - a huge company.  Do you REALLY think we are producing bad batches of chips?  You are the only person to report this.  It's obviously you that is doing something wrong". 

Anyway, my point is that when we used the chips immediately from the package, we didn't have problems - so it definitely seemed to be popcorning to blame, and I agree with you it seems like a process control issue at that fab plant, but I was lucky that I was manufacturing my own product and I was able to test each product during manufacture and replace the $0.40 microchip before bad parts were shipped out.  But if I was manufacturing parts on a contract basis where I had no way to gauge those failures, I could have been in a huge bind with my customers who would be furious with up to 50% of their boards not working and a multi-billion dollar manufacturer on the other side blaming me, the assembler

I'm not suggesting SVFeingold do anything on the basis of this story - just getting my little Atmel rant in and telling my popcorning story.  It seems to be the unicorn of PCB assembly... we've all heard about it but nobody ever really experiences it and it seems mythical almost 

[forrestc]

Anyway, my point is that when we used the chips immediately from the package, we didn't have problems - so it definitely seemed to be popcorning to blame, and I agree with you it seems like a process control issue at that fab plant....

It wouldn't surprise me to find out that for whatever reason those parts were msl3 instead of 2....

Although it seems odd that any soic-8 package is that sensitive - msl2 maybe.

[Smallsmt]

Well, I think you have proved it.  So, they may have had excessive porosity in the encapsulation, or some other situation that caused moisture to invade the package.  But, of course, you are left holding the bag.

Jon

True, however it wasn't a single run of chips... this was an ongoing problem over a period of years with chips purchased from Digikey, Mouser and Arrow.  Atmel were made aware of the issue and it continued to happen for years afterwards.  I took the time and expense to collect tons of failed chips and record the storage, usage and reflow parameters and then sent samples of each batch of failed chips to their HQ (was it in Denmark?  I forget) at my expense via FedEx for them to decap and analyze.  I had to hound them to do so - and eventually they basically just gave me a very curt reply of "bad process control, the chips have popcorned due to moisture - we are Atmel - a huge company.  Do you REALLY think we are producing bad batches of chips?  You are the only person to report this.  It's obviously you that is doing something wrong". 

Anyway, my point is that when we used the chips immediately from the package, we didn't have problems - so it definitely seemed to be popcorning to blame, and I agree with you it seems like a process control issue at that fab plant, but I was lucky that I was manufacturing my own product and I was able to test each product during manufacture and replace the $0.40 microchip before bad parts were shipped out.  But if I was manufacturing parts on a contract basis where I had no way to gauge those failures, I could have been in a huge bind with my customers who would be furious with up to 50% of their boards not working and a multi-billion dollar manufacturer on the other side blaming me, the assembler

I'm not suggesting SVFeingold do anything on the basis of this story - just getting my little Atmel rant in and telling my popcorning story.  It seems to be the unicorn of PCB assembly... we've all heard about it but nobody ever really experiences it and it seems mythical almost 

I used Atmel chips like you but never received this problems even if the chips where stored more than 6 Months!
Our storage humidity is below 40% so maybe your environment has high humidity?

[Jeroen3]

Chip vendors rarely tell someone who does not order an entire fab worth of business for weeks that they have done something wrong. To get a confession, you have to be Bosch (automotive) or Apple.

But, when you get failures you might have:
- Missing or damaged bond-wires.
- Empty packages.
- Solder ball oxidation.
- Wrong lot shipped.
- Wrong die packaged.
- Bad plastics.

The moisture is a problem in permanent moist climates. Like Taiwan.

ESD damage during processing will not show until the products are in the field.

[Mr. Scram]

I've looked at a lot of equipment, and this is what I've settled on so far as a preliminary setup:

• Philips Assembleon Topaz Xii (2005) - $38,800
• MPM Accuflex Stencil Printer - $24,800
• Heller 1706EXL Reflow Oven - $16,000
• Mirtec MV-7L AOI Inline Inspection System - $25,000
• Additional Assembleon Feeders - up to $4,000
• HEPA Air Filtration System - $1,500
• PCB Conveyors - up to $4,500
• Ultrasonic Cleaner for Stencils - $2,000
• Sil-Air Air Compressor - $4,000
• Esimated Freight Cost - $10,000

Total: $130,600

Even though I see I am rather late: did you calculate in the preparation of the building, setting up the line and the various bits and bobs? In my experience, people tend to focus on the big ticket items. Often, the smaller costs and items add up to a significant number. Depending on the sector, this number can be roughly equal to the big ticket items. You will need to install the devices, appropriate power, air lines, possibly prepare the floor, look into safety requirements such as emergency stops. Accessories and supporting equipment are generally not hugely expensive, but you need a lot of them.

There are a lot of things that come into play besides the big ticket items. Did you factor those in?

[SVFeingold]

Yes I've been collecting all of that stuff. I haven't abandoned this thread we're just getting the financing worked out so there hasn't been much to report.

For auxiliary stuff I'm trying to plan for <30k. This includes:

- Air compressor
- Air dryer
- Hoses/fittings/filters
- Electrical service (this will cost ~$3k, have been actively talking to the building electrician); includes a dedicated sub-panel with space for 30-40 breakers, the several breakers I'll actually need and conduit/outlets into the room.
- Floor prep - aside from cleaning it and ESD, not sure there's much else to do.
- Air filtration system (probably just going to get one of those big ceiling-mount HEPA filter units from eBay)
- Rigging; $305/hr, figure 2 full days as a reasonable max
- Dip solder pot or some such
- Large workbench w/ shelving - not a priority and not super expensive anyway
- New soldering system to tackle those damn ground planes; also not a priority
- PCB pre-heater; not a priority either and I'll get by fine with a $200 cheapie
- Shelves/racks/containers
- PCB racks
- Bitchin' speakers

I'm getting a nice machine table from work that they're tossing so that will be useful for putting some of my current equipment on.

Safety-wise, the e-stops are on the equipment. For a whole-room e-stop, not sure if there's much benefit at present. I'll be the only one in there for a while anyway.

[Pseudobyte]

- Floor prep - aside from cleaning it and ESD, not sure there's much else to do.

Please do not put conductive epoxy on your floor, if it is done poorly the end result will be a floor that chips away and looks gross. I prefer the polished concrete with a conductive wax finish, such as: http://desco.descoindustries.com/DescoCatalog/Floor-Care-Maintenance/Floor-Finishes/Statguard-Dissipative-Floor-Finish/10512/#.WYoVC4TyvDc.

[SVFeingold]

Interesting, definitely open to that. It's cheaper than the option I was considering for sure.

[Pseudobyte]

Yes, we had a conductive epoxy floor for the longest time. I got so tired of looking at it that, I took the debit card to home depot and bought $300 worth of kleanstrip and went to town scrapping the whole floor. Now there are just patches of epoxy hidden underneath our machines.

[jmelson]

Yes I've been collecting all of that stuff. I haven't abandoned this thread we're just getting the financing worked out so there hasn't been much to report.


- Air filtration system (probably just going to get one of those big ceiling-mount HEPA filter units from eBay)

If you don't have a good exhaust system for the oven, the flux fumes will clog up a HEPA filter pretty quickly.  If you DO have a good exhaust system, it will suck all the air out of the room, bringing in dirty air from outside.  I don't have a lot of experience with this, but it seems like a problem that is hard to solve.

Jon

[Kjelt]

If you don't have a good exhaust system for the oven, the flux fumes will clog up a HEPA filter pretty quickly.  If you DO have a good exhaust system, it will suck all the air out of the room, bringing in dirty air from outside.  I don't have a lot of experience with this, but it seems like a problem that is hard to
Jon

No expert either but you would put the filter on the air input not the output unless you have environmental rules that need to be obeyed.

[jmelson]

No expert either but you would put the filter on the air input not the output unless you have environmental rules that need to be obeyed.

Oh, then the HEPA filter has to supply more air to the room than the exhaust is pulling out.  That could get to be a fair amount of air, quickly filling the HEPA filters.

Jon

[SVFeingold]

It would have to be a series of pre-filters rather than just a HEPA filter, as you are correct that it would get clogged extremely quickly. It's not at all out of the question to cut a hole high up on the wall, into the hallway, that would have a pre-filter screen and baffles for noise. That would provide the inlet air. The exhaust would need to filtered as well since it will come out over a walkway. Basically to get all of the noxious stuff out, but that should be a relatively straightforward filtering task by comparison. With appropriate filters and pre-filters I can't imagine needing to clean or replace them more than once a month, and less often if I'm not running at full tilt.

Inlet air directly from the outdoors is not ideal for a few reasons. One is that there is a dirt path that goes by the nearest outdoor access, not a great place to get air. I'll be changing pre-filters often. In addition it's been raining quite a bit and I don't want to pull very humid air directly into the room.

The HEPA filter would only be re-circulating air inside the room. I imagine after the intial period which will be naturally more dusty, the exhaust fan would cause air to be pulled through the inlet, which would pre-filter the air coming into the room to lessen the load on the HEPA filter. Aside from that there will just be the building shedding, hallway air, and whatever people carry in.

[jmelson]

The HEPA filter would only be re-circulating air inside the room. I imagine after the intial period which will be naturally more dusty, the exhaust fan would cause air to be pulled through the inlet, which would pre-filter the air coming into the room to lessen the load on the HEPA filter. Aside from that there will just be the building shedding, hallway air, and whatever people carry in.

creating slight positive pressure in the room through a good filter will keep the room much cleaner.  (How much cleaner depends on how dirty the air is outside your room.)  But, that is always a good idea.

Jon

[SVFeingold]

That is a good idea indeed. So in that case the HEPA/pre-filter would be only on the air inlet, but I imagine that will require some sort of pressure-regulating air outlet, or perhaps just a smaller opening. I feel like those filters will get clogged faster pulling air into the room constantly vs. re-circulating air that's already been pre-filtered. 

I'm sure some clean-room tech is laughing at me right now trying to figure out this stuff that's been solved for decades. :p

[mrpackethead]

So, how is the project coming along..

Interestingly i've largely done what your planning earlier this year.   Set up a SMT line.. but i did it for $60k.   

NB - Dont' sell the feeders ( unless you want to sell them to me ), you never have enough!
 

[SVFeingold]

It's moving along, sorry for the lack of updates! Just now I'm actually securing the space. We've been dealing with getting the financing set up which is always a hoot.

I do think I will wake up many nights and think "holy shit I could have saved $100k and achieved almost the same thing." But...we'll see! MPH Can you tell more about your line and your experiences?

If all goes well we may be up and running in a couple of weeks, best case scenario. We planned to already have the equipment here by now but obviously our plans had other plans...

Most recently I've been thinking about all of the ancillary stuff. In addition to the air filtration I'd like to get a nitrogen bottle to see if nitrogen soldering is really so great - manual soldering, not reflow. I have a JBC rep coming to my primary job next week so hopefully she'll bring a nitrogen generator and I can try it out firsthand. I actually just ordered a new JBC handpiece and a bunch of tips at work, and coming from a Weller WD1 + WP80 I have to say I'm just a little bit underwhelmed at the JBC. Even the large tips on the T245 are about hopeless for 8-layer ground planes but maybe I had too lofty of expectations. The tips also don't wet nearly as well as my trusty Weller.

What else...the air compressor setup is still an open question. A flammable liquid storage cabinet is going to be a prudent investment. Workbench, shelving, etc... Lots of prep work.

I've been also thinking about PCB cleaning. Right now I do it in a very tedious manual way, with a lab hotplate, flux remover, and then an IPA rinse. Takes about forever. After finding out how cheap DI water is I'm planning to run a water line to the lab to then run through a DI water generator/filter system so I can have it on tap when needed for cleaning. No drain in the room but there is a shop sink in the hall.

Speaking of cleaning, if this helps anyone I found Techspray E-line Universal Cleaner to be far superior to MG Chem 413B. Cheaper too. 

Curious what others do for PCB cleaning of larger batches. PCB washing systems are loony tunes expensive.

Last step is to find a couch to cram in there so I can hang out with girlfriend and my little dog, cancelling out the benefit of all that expensive air filtration. :p

[mrpackethead]

I do think I will wake up many nights and think "holy shit I could have saved $100k and achieved almost the same thing." But...we'll see! MPH Can you tell more about your line and your experiences?

I've got two Yamaha YV100-ii machines.  I was able to get them at auction when a company went into receivership in Australia.   They were really cheap.   Was a 'as-is/where-is' purchase.   But we lucked in, and both machines where in good order.   I've replaced nozzles and some sensor..   For reflow i'm using vapour phase, which is slow but awesome.    I got a decent air-compressor, to run them..    I was able to lease basically a concrete shell.. Painted / sealed the walls and ceiling,  and painted the floor with ESD paint.    Air lines /power are in conduit just screwd up to the wall.      Storage is simple but effectivel $100 shelving units from the local hardware warehouse.
Have installed a Aircon unit, because it gets too cold.. No air filtration, yet, but its pretty clean..   I have a manual stencil printer (cheap unit) and got a $3500 semi automatic unit from Torch in china..  The manual is great for short runs, its quicker to setup for one.. the auto is good for longer runs,  it takes longer to setup but requires less work each time.. think the cut over woudl be about 20 panels.     

I've been aquriing more feeders ( ex china, clones ) as need for projects..   and i seem to buy bits and peices of tools.. But baiscally it seems that the rule is not to buy anything untill you absoltely need it.    Its amazing what you can do with out it.   Bought ESD mats for our workbenches, and cut it to shape.   made my own bonding points. 

>I've been also thinking about PCB cleaning. Right now I do it in a very tedious manual way, with a lab hotplate, flux remover, and then an IPA rinse. Takes about forever.

Big bonus with the vapour phase is that the boards come out really clean.. the flux ends up in the bottom of the tank!

. 
>Curious what others do for PCB cleaning of larger batches. PCB washing systems are loony tunes expensive.

Im comign to the conclusion that there is a LOT of Myth and BS spun about PCBAssembly and what you need and must have.    With some good thinking you really can get away with out all the 'flash' gear.   My stencil cleaner cost me about $300. i bought some Ultraonic transducers on aliexpress, and had a tank made from stainless steel for example.  Its really not that tricky.


Last step is to find a couch to cram in there so I can hang out with girlfriend and my little dog, cancelling out the benefit of all that expensive air filtration. :p
[/quote]

[Pseudobyte]

I've got two Yamaha YV100-ii machines.

You should pick some of these up for your FV feeders http://feederfinger.com/product/yamaha-fv-feeder/ would make your setup time much lower. We wanted to buy these for our company, but they don't make one for our assembleon feeders.