I was printing a TQFP64 (0.5 mm pitch) with a kapton stencil and boy do I get a lot of bridges. Quite a pain...
I was wondering if a metal foil stencil would be that much better, and if so where a good place to get them front is (I found plenty on google but would like opinions).
The kapton lines are not all uniform for the TQFN but they all are certainly separate. When I print they run together certainly during the soak (hot plate) they run together. Perhaps I am just not doing everything perfectly. I am using NXG1 paste. Perhaps I just need more practice.
I have never worked directly with stencils myself. I handle the schematic and PCB design and then let my clients decide who to hire for the fab steps.
One client actually bought a pick-and-place and tried doing this in-house. I have to say from their experience that getting the solder paste right takes a lot of trial and error. They were using metal stencils that cost around $90, maybe less. They had lots of problems until they learned to use less solder paste. I think the key is to use very little if you have problems with bridges. Then again, I'm talking purely about second-hand information.
My client eventually gave up on my boards with TQFP64 and had them outsourced - the build quality went up when they let someone with more experience handle the job. Now, my client just uses their pick-and-place for simple button, switch, and LED boards - not for the fine-pitched stuff. Since they like customizing LED colors and other things, they still benefit from doing the small, custom runs in-house. But the main board is handled by the pros.
Hummm... Well I could contract it out but I feel like that is quite a large amount more. I would say this was more an issue of the paste getting to where it should not be than quantity. The first one I did certainly I used to much, but with a hard sqeegy only the stencil thickness is left. I have a feeling perhaps the kapton moved under that much force even though it is taped down it is a very flexible tape.
I hope someone else chimes in, because I really only have a few tidbits of info that I've picked up here and there.
One thing I heard is that the metal stencils do get stretched as soon as they're used, but I have the impression that they're mounted into a solid frame - not just taped on.
One client actually had me create a "diff" of the stencil between two revisions, so they could rework a stencil for a earlier PCB revision into a stencil for a later PCB revision (we added more bypass and filtering caps). Anyway, they were only able to do this because the metal stencil had never been used. Even one use distorts the stencil. I have no idea how it can remain precise under those conditions, but I've never worked with the process directly.
Good luck, and keep us posted. It's probably worth your effort for the cost savings, but you might still have to work through a few iterations before you get it right.
I don't have any experience with using metal stencils. However, I've used both mylar and Kapton stencils. When I first switched to using Kapton stencils (made by Ryan Ohara, www.ohararp.com (http://www.ohararp.com)) I had some trouble with solder bridging on smaller pitch components. Ryan uses 3.5 mil Kapton film which is somewhat thicker than the mylar I was originally using. So I attributed the bridging to the extra thickness and the fact that I wasn't being careful enough to squeegee off the excess solder paste. It takes an amazingly thin layer of solder paste to do the job and once I was more careful to cleanly squeegee the stencil, my solder bridge problems pretty much went away. You don't need to press down hard but just make sure you make a nice clean swipe with your squeegee. I like to use a piece of flexible plastic, like a fairly thin, flexible credit card, as a squeegee.
In my experience, the solder paste slumps quite a bit during the soak phase, as the flux activates. At that point, things look pretty bad but, as long as you have a solder mask on your board and have applied the correct amount of solder paste, the solder ends up contained on the pads once it melts.
the pink Bondo squeegees work well. when the paste is cold it is harder to get even coating squeeze a little out to warm up to room temp before spreading.
[quote author="sdixon"]
When I first switched to using Kapton stencils (made by Ryan Ohara, www.ohararp.com (http://www.ohararp.com)) I had some trouble with solder bridging on smaller pitch components
[/quote]
I am using his kapton as well. I was using a plastic sparkle knife but I will try a credit card next time, as that would require less force.
@sqkybeaver the paste was at room temp. This is a lead free paste don't know if that makes a difference, perhaps because it doesn't wet as nicely it might bridge more...
I am curious how you guy's lift the stencil off the PCB. I had it taped to 1 side of an L in 3 places and then the corner of the other edge with one tape. I stenciled, removed the one corner tape then lifted the stencil from the angle.
I think it is possible shaking or flexing while lifting is the reason the paste ran together. Though I do agree if everything wets nicely then the paste should just reflow and wick up the right parts.
PB free is always a little more challenging, it is sometimes easier to clean up bridging after the initial re-flow. search youtube for solder bridge repair.
Invest in a good no clean(low/no particulate) liquid flux.
How long has the paste been open?
The lead free paste has different heat requirements for re-flow, check the manufacturer recommendations, sometimes it is an easy fix that was overlooked(done it many times myself).
I've tried the toaster oven re-flow, for a batch of 500 (2x6 pallet) many were unevenly heated, 90% needed to have bridges fixed, i ended up rebuilding the toaster to get an even heat. worked a little better but still needed to clean a lot of boards.
[quote author="drwho9437"]
I am curious how you guy's lift the stencil off the PCB. I had it taped to 1 side of an L in 3 places and then the corner of the other edge with one tape. I stenciled, removed the one corner tape then lifted the stencil from the angle.
I think it is possible shaking or flexing while lifting is the reason the paste ran together. Though I do agree if everything wets nicely then the paste should just reflow and wick up the right parts.
[/quote]
For most of the small boards I do, I secure the board in between two L shaped pieces of plastic the same thickness as the PCB. Then I align the stencil and tape it to the plastic on the top edge so that it is hinged. If you are careful, you can put a spot of solder paste at the top of the stencil and squeegee it on. Then lift the stencil from the bottom in one motion so that you don't smear the paste on the board. This probably wouldn't work for larger boards or full panels, where you might want to set up some sort of frame and registration system. But it works for what I do.
[quote author="sqkybeaver"]
How long has the paste been open?
[/quote]
I have only had the paste like a month, it has a shelf life of 6+ months as I recall. I am using a lab hotplate rather than an oven. Reflowing of smaller areas happens first.
@sdixon, that sounds essentially like what I did actually, save for how I secured it in the corner so the PCB could not move... I used a plastic knife (food) to spread the paste. I used shims (Al foil folded) to level the board verse the stencil L.
I just found these on dealextreme:
http://www.dealextreme.com/p/multi-purp ... -25mm-6714 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-cell-phone-diy-0-25mm-6714)
http://www.dealextreme.com/p/multi-purp ... -25mm-6712 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-sony-ericsson-series-0-25mm-6712)
http://www.dealextreme.com/p/multi-purp ... -25mm-6715 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-samsung-series-0-25mm-6715)
http://www.dealextreme.com/p/multi-purp ... 3-v3x-6703 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-motorola-v3-v3x-6703)
http://www.dealextreme.com/p/multi-purp ... e-diy-6704 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-cell-phone-diy-6704)
http://www.dealextreme.com/p/multi-purp ... e-diy-6705 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-cell-phone-diy-6705)
http://www.dealextreme.com/p/multi-purp ... e-diy-6700 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-cell-phone-diy-6700)
http://www.dealextreme.com/p/multi-purp ... e-diy-6701 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-cell-phone-diy-6701)
http://www.dealextreme.com/p/multi-purp ... -25mm-6708 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-for-nokia-series-0-25mm-6708)
http://www.dealextreme.com/p/multi-purp ... -25mm-8674 (http://www.dealextreme.com/p/multi-purpose-ic-soldering-film-0-25mm-8674)
no experience whatsoever, but I saw these, and the price is very very cheap (2-3$). So you could alway try if it is usable or not
There is a 'nack' to stencil printing.
The thickness of the stencil, and the pad dimensions are what is critical, bridges are due to far too much solder paste being laid down, possibly 50% too much solder paste.
DEK has a manual, if you read it you will be enlightened.
The 'k' is silent.
It sounds like you are using too much paste. You may also has some slump in the paste. Is the paste 100% on pad, or is it off the pad after print. If it is off pad after print, then you have paste reflowing without a metalic surface to attach to. It will bridge.
I actually have found that while it may be too much paste for the TQFN it is too little for the bigger parts... I am getting better at it with some flux I can unbridge it okay with a bit of work...
The critical action on stencils is NOT to force paste through the mask, also there is a need for a rubber squeegee with leading edge at a 45 degree angle to the stencil.
Finally you need a 'surf' wave, this is where the solder-paste rolls over on the leading edge (just like a 'surf' wave), this ensures that a 'clean' surface of solder paste contacts the solder pads through the mask, it is also why you need the rubber squeegee, such a wave is difficult to produce with a metal or plastic blade. Another issues with plastic, it wears faster and you can end-up with fragments of micro-plastics in your solder paste.
Yes bridges can be removed, but why bother with extra work when a decent screening technique can produce a perfect board every time.
Interesting tips, I don't know how I could get a reasonable amount of paste without quite a put of pressure. Saw a video of Sparkfun production and they used a metal Spackle knife. I tired the rubber squeegee first, I had the best results with a plastic putty knife. I feel like the slots in the stencil are just too wide really the paste slumps enough and there is too much overall... The problem would be fixed if they were 30% smaller because ever other place on the board needs *more* paste. I have a lot of issues with pins pulling on high force headers because I used so little paste, it means I have to check the boards a lot to make sure they are perfect.
Then the solder-paste is the wrong mix, when a blade is at about 45 degrees, it should barrel roll.
The barrel falls down over the aperture then the leading edge of the blade spreads and levels it out, the thickness of the mask and the size of the aperture dictate the 'load'.
Many people think that you just stick a few holes in a metal plate and it becomes a mask, it is not that simple
Even the size, shape of the component leads dictate how much wastage there is, sometimes if you look at a 'pro' board you will see small solder balls, this is incorrect application of paste.
you can adjust the 'holes' , by adding sticky tape on the other side, it is a pain in the ass, but faster than re-working.
Good results are possible, I setup a private company to do SMT about 25 years ago, when at that time they were still using 1206 (nice and easy.), but I have see just as good results from a 2-3 person production line in China on 0805 or smaller all using hand stenciled boards and a couple of CD cases to hold the boards in position whilst they were pasted. total cost of kit < $30us
I have no issue with 0805 or even 0603 just fine pitch stuff. I don't know what you mean by mix, my paste is just in a jar, I spoon some out and let it warm up then apply. (NX01 Kester).
I would say the load dictated by the stencil for the fine pitch is too much, and too little for big stuff, so I should use a thicker mask with smaller apertures everywhere but the big stuff. That would load everything correctly.
Kester is an 'excellent' brand worth the extra money, I tried many different brands, but with Kester we were seeing a 15% increase in positive results, less rework, less tomb stoning, excellent bond strength.
If you look at the catalogue, they have different 'mixes' particle size and viscosity.
SMT is easy, but its just getting the right mix and experience. once that is done you are fairly well set for life.
I can support the points hardcore says.
Bridging means most of the time too much solder paste (if your stencil was placed accurate).
a simple solution is to use a thinner stencil material. but as brian wrotes, other components need more paste...
the solution you handle this is to use a normal thick stencil (127 um or 150 um) and reduce the pad size in the stencil for the fine pitch components.
professional PCB software is able to handle this stuff. you can define the solder mask pad size in the component library differently than the copper pad size etc. (it can even have a different shape than rectangular, which is also often used now with lead free solder paste). the pro's do more tricks to stencils, for example really big pads are splitted into smaller ones ("rasterized") to get better paste print results.
also CAM software used prepare the data for the machine can adjust the pad size. if you order your stencil you can often ask for a general pad size reduction for all pads on the board -> go for it, this helps a lot!
we reduced our pads normally by 0.5 um (for a rectangular every side is pushed 0.5 um closer to the center) for a 0.127 um stencil. we did this in our CAD but also our stencil manufacturer has this option in his online order form.
I used to use DEK in the UK, you send them the copper or silkscreens and they would deal with it for you, since they are experts in screening they knew exactly what changes to make.
Out here in China, they make the stencil to what you give them, but interestingly enough they hate doing anything other than metal stencils.
Obviously because they are easier & faster as regards stretching the material, you can be less of an expert stencil maker, since distortion is removed with a metal stencil, unless you do something really stupid.
As I said, with the kester you get a nice clean print and even after 20 minutes it does not slump, some of the other crap has so much fluid movement, that it starts to spread off the pad after a minute or so, and kester has a nice smell. :-)
Something about it makes me feel a bit queezy after about an hour, but yes the smell is interesting, kind of a sweet smell.
That is because they removed the lead.
All growing children need lead, didn't do me any harm.
Mine is lead free. But it is the organic flux in the paste no doubt that smells. Compared to the solvents in my cleanroom I do not worry about it... Hazard: (100) piff you could probably put it on toast and not die. [Kids do not eat solder paste please]
I saw a link from the forum to my site and I thought I would chime in to see if I could help. Anyways I have found fine pitch tqfp to be the bane of my business. It will be tough to avoid bridging with a non-metal stencil but it can be done. I usually resize all the pads to give the customer a good soldering experience. I am not sure if your vendor did this for you, but I certainly do it for the TQFP parts. Definitely use a plastic type squeegee. I cringe at the SFE example with the metal on kapton technique. Old hotel key cards are perfect.
Another option is to actually put a thin strip int he stencil that is very narrow instead of individual pads. By knowing that the soldering is going to bridge and hopefully wick you can actually prevent bridging with this technique. I'll often employ this with 100+ pin TQFP parts.
Hope that helps!
Ryan O'Hara
FWIW you cut my kapton one but who knows if I did the paste layer right, it is a good tip on the thin line. I was thinking of just shrinking the pads some but it I think I was already near the min line size (corners were rounded).
When I used metal stencil it was a success even for the first time. When I tried mylar stencil, I had bridges problem. As I can see the size of the hole for pad is bigger than the ones on the metal stencil (maybe the laser cut makes it bigger) so I think I should reduce the size of hole for Mylar stencil.