Acetoxy- curing silicone used on circuit boards. 2

[btrueblood]

So, sometime before the holidays our company received a batch of new circuit boards, to the tune of several hundred dollar per board. These have several large electrolytic caps on them, which the board shop glued to the board using silicone rtv. Not until a couple of days ago did I realize this, when I got a couple of boards out of their bins, removed from the anti-static bags and got hit with a snoot full of vinegar. I was told one should never use the acetic-acid kind of silicone for electronics, due to the possibility of corrosion occurring, but only the non-corrosive types of silicone should be used. We are lucky that it has been quite cold here recently, and the humidity in the office and production areas is very low (so, maybe less likely to form aqueous acetic acid from the vapor fume that condenses on the IC's and such.

Any thoughts on how to determine if any damage has occurred, and other corrective actions that could be taken with these boards? I've had our inventory guy take them all out of the bags and set them out in an un-used office, so that the fumes can dissipate.

 

[itsmoked]

WOW. What a complete bollix that was.

I'd use a stereo microscope at about 30~40 magnification and look at the leads. Things will probably be OK. If you see residue then consider cleaning the boards again once no smell is detected.

You should write "methanol RTV only" into your assembly instructions.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[btrueblood]

Thanks Keith, would you expect to see white or green corrosion deposits? I'll take a look with what we have here.

Pretty surprising, that the board shop that built these claims that this is their "standard potting compound" and they use it "all the time".

 

[MacGyverS2000]

If that's their claim, I bet they don't do much military hardware, then...

When I do a run of boards, it's typically around $1k-3k/board in quantities of 10. You tend not to make too many mistakes when that's the price tag... or at least HOPE you don't make too many mistakes.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[3DDave]

One of the funniest things about the methanol RTV is the caution that it is harmful if ingested in the uncured state.

Here's one NASA take on the difference

https://code541.gsfc.nasa.gov/Uploads_materials_tips_PDFs/TIP 044R.pdf

 

[MikeHalloran]

In order to ensure that the cure goes to completion,
I'd air the boards out in a room with constant airflow.
This time of year, I'd also put a humidifier in there.

... and I'd start qualifying another board supplier.




Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

Man Mac, that is some highfalutin boards you mess with. Must be covered with overpriced FPGAs.

I'm with you Mike. Silicone needs water out of the air to cure. (Both kinds)

btrue; I cannot say what color because IC leads are made out of several different metals and alloys, and are also plated with different metals. I've seen mostly white, some clear, and some greenish corrosions, well actually black too. I would look more for any discoloration on leads. Look for any 3D stuff - which would be really bad.

Your board house is populated with idiots if you haven't guessed this yet. :/

Keith Cress
kcress -

http://www.flaminsystems.com

 

[btrueblood]

Luckily, the humidity has come up in our offices over the last few days, owing to torrential rains hitting our area, so that's roughly sorted. As I said, the boards were being stored at fairly low temperature and humidity levels over the preceeding holidays, which probably slowed the cure and production rate of the acetic acid vapors...so maybe we managed to dodge a bullet. We have pretty good air exchange/airflow in the office spaces in question. Odor has dissipated quite a bit overnight, will check again today, but will probably wait a full week before doing the inspections as you suggested Keith.

Mike, you assume we qualified the first supplier...

 

[MacGyverS2000]

Keith, it's mainly the boards themselves... 3mil trace/space, 4mil via holes with sub-11mil annular rings, numerous layers, etc. That price generally includes parts/assembly, but to be fair, the parts are typically only in the $10-$100 range. Quick turn / tight tolerance makes for puckered keesters.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[ScottyUK]

Are there any locations such as under IC's where the corrosive vapour could be trapped in an air pocket and sealed in by the RTV? Those may be the places to be concerned about because it won't disperse over time.

 

[mcgyvr]

Maybe I missed it.. Was there anything in the data you sent to the board house mentioning anything about adhesive/capacitor securement,etc??
If not that alone is enough to reject the whole lot.. and thats what I would do..
Nothing like finding out a year or so later that what you though was not an issue is now a million dollar recall situation..

 

[btrueblood]

mcgyvr,

A colleague from our sister company is our board designer/EE. His drawings have a note that specifies the caps be glued to the boards with RTV; but no, he did not specifically request an adhesive that is non-corrosive, I guess he ASSumed that it would be one of the methanol types. Our prototypes came in with the caps just floating loose, and we all thought it wouldn't be prudent to ship such across the globe, so we updated the print for the first beta run. I'm just the dumb mechie who noticed the problem and squawked, and is trying to figure out how to remedy. There are no entrapment areas near ICs or other devices, just lines of RTV along sides of some big-ish electrolytics standing by themselves off on one side of the board, and no real entrapment (other than bubbles in the beads of sealant themselves) there either. Still...I'm planning to follow Keith's advice and examine these after they have had a chance to fully cure, and may take some other steps as well, such as a localized wash with a pH measurement of the rinse water (there are some hand soldered parts that I don't think would stand immersion).

 

[Woodyjumps]

This issue comes up with large DC motors as well. The ducting that delivers cooling air to the motor, including the brushes, is often unwittingly sealed with basic "whatever is on hand" sealants, and they can have an adverse effect on the brushes.

 

[OperaHouse]

We built a machine that did a signature analysis on new pump noise. Some of these pumps were also shipped to Korea and there were numerous failures. Factory firmly believed it was poor manufacturing practices in Korea. A pallet of pumps destined for Korea was diverted to us for final testing of the machine. 80% of pumps failed our test. Opened one up and found rust inside. The pumps were placed in an unsealed bag and foamed in place. Out gassing of foam was destroying the pumps.