Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Accuracy Requirements for New Technology

#33591

Accuracy Requirements for New Technology | 7 April, 2005

I am having a common problem, justifying capitol expenses. We are currently using 0402's and BGA's and CGA's with 1MM centers. Our current equipment is horrible, we have a hand operated screen printer machine (No vision system) and we are placing all parts by hand. The twist is we build one to maybe 4 boards of each program a year. We have 0201's on the horizon along with micro BGA's. Our parts are extremely expense too. Any ideas on how to justify the need, even in low low volume, for at least a good screen printer and a placement machine? What is human accuracy good for? I have recently joined the company and I am at a loss on how they are getting by with this. I need proof of accuracy needs for this technology.

reply »

CN

#33603

Accuracy Requirements for New Technology | 8 April, 2005

Sounds like it would by tough to justify, thought of oursourcing? if the volume is really that low I bet it would be cheaper to find an outside source to build them for you. As far as hand placment accuracy If you are hand placing BGA's or even worse CSP's my personal opinion is that should never be done. do you have x-ray or rework equipment? if no x-ray how would you ever verify placement and if you had x-ray but no rework how would you remove and replace. However if your volume is very low and you do have high end rework with camera system for alignment you could use it to place BGA's and CSP's.

reply »

#33604

Accuracy Requirements for New Technology | 8 April, 2005

Take a look at this equipment:

http://www.essemtec.com/e/102.php

http://www.essemtec.com/e/145.php

This stuff is fairly short money.

reply »

Grant

#33605

Accuracy Requirements for New Technology | 8 April, 2005

Hi,

What are you doing for reflow. Sounds like a cheap MYDATA TP machine second hand might do the trick, but don't know if it could do 0201 though. You can get a fairly ok manual stencil printer that might do the trick. We had one that had a linear baring down the side to keep the squeegee straight, and we added a metal blade. It worked ok, but we needed to use a 5 thou stencil to get reliable paste release.

Good luck, and I agree with some of the other guys, it might just be worth outsourcing it. However with a good microscope, you might be ok if your volume is so low. We place BGA's manually all the time, and they self center, and generally work fine.

Regards,

Grant

reply »

#33607

Accuracy Requirements for New Technology | 8 April, 2005

Does the product that you assemble meet your customers' requirements?

reply »

#33610

Accuracy Requirements for New Technology | 8 April, 2005

We are using a low end Air vac system to place the BGA's. We have X ray we can use. Our CGA's as well use this system. They are one off parts that are above the 20K a piece range. So I have been tasked with making sure all of these procedures are the best they can be and I am having to eye ball the placement. The next hurdle is the length of time for placement. We are running around 8 to 14 hours to populate a board. This is far outside the parameters of the flux in the paste. I have been in the business for ten years and have never had to justify capitol like this before. I have looked at human error rates and have seen numbers from 66% to 80%. Our printer by the way doesn't even have a blade mounted to the frame. You hold it in your hand! I guess what I am looking for is justification of expense for future technology. I have something generic whipped up on human error rate. Then I was asked for what kind of skew and shift tolerances do these parts have. The best I can come up with is 10% of the smallest lead dimension. So an 0402 would have a placement tolerance of .002"? Does this sound right? I wanted to find printed proof of my theory though. Any ideas where to look for this? The outsourcing issue, that can't be done to NASA requirements. Traceability alone would send shivers up there spines at a board shop. We solder to the J standard but there is alot more to space flight requirements.

reply »

Grant

#33617

Accuracy Requirements for New Technology | 8 April, 2005

Hi,

It seems weird that an automated pick and place machine at a board shop would not be up to NASA standards, but placing the components by hand would be. However most SMT will self center a bit, so if your close enough, you should be ok.

You really need some kind of machine I think. It's important to look at each part of the process and see if anything is not working. For example, your printer might be primitive, but if it's doing the print ok for the low volume then your ok.

However to me it seems your placement is the problem, and it's taking too long. We originally built our prototypes by hand, and know how slow it is. You might need to pick up some kind of pick and place machine, at least to place the passive components, which might be the slow part.

You might also want to get Vapor Phase for a cheap oven. We used the fluid in a paster steamer originally, and it worked great for a few units. You won't overheat the board, and it's oxygen free, so that should be up to NASA requirements.

Good luck, and some people on the forum might be able to suggest some low cost machines that are ok for low volume. I have used MYDATA which would be ok, but there might be better choices. That seems like your main issue, and it's the time it's taking to place all the components.

Regards,

Grant

reply »

Base

#33632

Accuracy Requirements for New Technology | 11 April, 2005

You might want to take a look at some of the Assembl�on/Yamaha stuff. I believe they have a machine that has a couple of dispense heads and a couple of placement heads, so you can get your solder-paste and component-placement all done in one machine. I'm not sure they still offer it though...

reply »

Grant

#33633

Accuracy Requirements for New Technology | 11 April, 2005

Hi,

Yes, I think that's what he needs. Something that does fine pitch placement, but could also do faster chip shooting if required. Just a fine pitch placer might be too slow, but if he's only doing proto's then it might be ok, as time would be be a big deal.

Regards,

Grant

reply »


Rob

#33634

Accuracy Requirements for New Technology | 11 April, 2005

You'll struggle to find low end machines that accurately and reliably place 0201's (despite some of the manufacturers claims). You'll also obviously need to place the larger components on the same platform, accurately which will really cost you.

You really should consider subbing out, and to someone who has a good screen printer & 3D paste inspection - there is no way you can guarantee accurate volumetric deposition of paste for the components you are talking of using a hand held blade.

reply »

Grant

#33637

Accuracy Requirements for New Technology | 11 April, 2005

Hi,

Yes, it's going to be hard. I was thinking a high end fine pitch placer could place the fine pitch accurately, and then slowly do the passives. That might do the trick, but unless he's prepared to spend at least half a million, it's going to just end up costing more in the long run. It's the 0201 requirement thats a real killer here.

Grant

reply »


Rob

#33638

Accuracy Requirements for New Technology | 11 April, 2005

There's not many ways of doing it on the cheap except...

Buy a second hand universal GSM1 with VME 630 vision, 0.5 mil per pixel camera & some expensive ceramic nozzles for the 0201's. This will be quite slow & the software quite awkward but looking at the requirements it doesn't need to be quick.

Should cost about $30000 including feeders. Now buy a decent 2nd user printer (MPM AP or DEK 265) and an offline solder paste height checker. (Approx $35000 for the pair).

Add in training & servicing, air & instsallation and it's going to be close on $85,000

So still cheaper to sub out.

reply »

#33646

Accuracy Requirements for New Technology | 11 April, 2005

GSM1 won't do 0201's you would have to buy a new GSM for this. You should contract out since it sounds like you $$$ situation is not good.

reply »

#33647

Accuracy Requirements for New Technology | 12 April, 2005

Hi In regards to the screen printer you might like to check our Reprints R29-V semi automatic. For low to medium volume it would be perfect for your needs and being a 29" machine and able to print down to 01005 pads with ease. It has a digital zoom facility as standard which would assist not just for the very fine pitch areas but also for any post print inspection you might need to make.

Its certainly worth considering and would last a life time.

Check it out here - http://www.reprintservices.co.uk rgds Ian LG

reply »

#33648

Accuracy Requirements for New Technology | 12 April, 2005

Hi In regards to the screen printer you might like to check our Reprints R29-V semi automatic. For low to medium volume it would be perfect for your needs and being a 29" machine and able to print down to 01005 pads with ease. It has a digital zoom facility as standard which would assist not just for the very fine pitch areas but also for any post print inspection you might need to make.

Its certainly worth considering and would last a life time.

Check it out here - http://www.reprintservices.co.uk rgds Ian LG

reply »


Rob

#33649

Accuracy Requirements for New Technology | 12 April, 2005

Hi SMTUser,

I've never personally tried it, but one of my customers builds hybrids & low volume very complex boards & gets very good results using GSM1's placing 0201's using the aforementioned set up.

reply »

Base

#33651

Accuracy Requirements for New Technology | 12 April, 2005

You mentioned you only do 1-4 boards per program per year, and that you're doing 8-14hrs per board. Does that mean you're only doing a couple of boards per week in total? In other words: consider the amount of running time you expect to get on the machine and then consider the investment you have to make. Your cost per running hour may exceed the cost of labour by far. I haven't done the math yet since I don't know your boards, but expect a low-cost high-tech machine to place 3kCph, so I think populating an '8hr'-board would take the machine an hour or less.

But apart from that: looks like your main motivation is not an economic one, but technology/quality driven. Management is usually only about the $$$, (or at least mostly but then again: that's their job), so you will have to crunch the numbers to make it financially attractive. Probably they have done this already or they are expecting you to do so. Ask them for a budget that would be acceptable for them and then search the market for what you can get.

Also: have you considered the amount of additional business you could gain if you would expand your capacity?

Good luck!

reply »

SMT Services - STI Electronics

Manufacturing Software