Intermittent "short" to case of rotary switch

On a sunny day (Wed, 3 Feb 2016 07:52:01 -0800 (PST)) it happened George Herold wrote in :

Cool, I need to sign in!

Hi Jan... Hmm I screwed up the dropbox link I guess....(grumble.)

Try this,

George

(I'm wondering if I've got some sort of solder whiskers growing.)

oops,

On a sunny day (Wed, 3 Feb 2016 09:08:05 -0800 (PST)) it happened George Herold wrote in :

What does a magnifying glass tell you? I imagine rotating it could scrape off little pieces of metal and the distance to the case s really small.

Once, we had to fix a PLC system in a food processing plant that was close to a coal processing plant. The control PC floppies (at that time big ones) had carbon tracks in it where the read heads moved, sort of a coal residue that was _everywhere_. But I would expect the switches to be sealed.. Guarantee case, but not much help to you. Or moisture? Or manufacturer assembly error with conducting stuff getting in? Try to take a macro picture, if your camera has it, I often do that to find things I cannot even see with a magnifying glass.

Looks like you may have whiskers. Lead was added to plating containing zinc or tin to suppress whisker growth. I wonder if ROHS just bit you?

Hi Jan.. well whatever was in there I cleaned it off, and it's no longer conducting, so now the switch works perfectly. All the pcb's we buy now are Rohs. But we put on the switches and pots and things that can't be washed with tin/lead Kester "44"*. Maybe the tin/ lead does not play well with the Rohs stuff?

George H.

I wonder the same thing... As I just said to Jan, we use tin/lead to solder in the pots and switches... but the pcb's are rohs.

George H.

On a sunny day (Wed, 3 Feb 2016 10:16:30 -0800 (PST)) it happened George Herold wrote in :

I do most with this:

That should be fine, Lead free alone can cause whiskers. I'm thinking your cleaning process is not adequate and that maybe the problem. I've seem piezo buzzer failures because junk was left under the part.

Cheers

I had a case where a set of relay contacts was controlling a logic circuit. The relay was mounted with the contacts facing up and the insulator down. The relay was in the PTT of a 2 meter repeater that clicked 100's of time s a day. After a few years, there was a failure where the logic circuit th ought the contacts were closed even when the relay was off. Careful inspe ction showed there was metal dust across the insulator that came off the co ntacts and fell on the insulator and eventually made a 50 Ohms path which w as enough for the logic circuit to call it closed.

Mark

Thanks Martin, It's hard to see how it's a cleaning/ gunk issue. The flux is cleaned with IPA and a cotton swab. The "short" happened on the other side of the pcb... and beyond that it's not on the pcb but up inside the switch body.. (Which sits about 0.05" above the pcb.) So I don't see how it's whiskers growing the pcb either. I removed the switch from the pcb and the short remained.

I have no idea what it could be... I guess I'll just wait for the third shoe to drop.

George H.

arcing creating a carbon path?

-Lasse

If you look at the vise photo you can see whiskers around the periphery. They seem to be missing in the meter photo.

See the fuzzy stuff (in the photo with the vice) that looks like it is growing from the casing inwards towards the terminals? I think the casing is growing whiskers towards the terminals.

If you want to check now that you have destroyed the external evidence, maybe there is still a chance if you hacksaw it in half so that you can inspect the inside of the casing where you haven't cleaned it yet. There might be more growing inside it, though maybe not as the mechanical stress in the coating may be different there since the external part looks like it has been bent after plating.

I have seen whiskers growing on the casing of a relay out of a WWII tank radio.

I suggest either contacting Grayhill and asking whether they have seen/solved this problem, choosing a different switch, or conformally coating the terminals and the casing near them.

Regards, Chris

No question, those are whiskers. They look similar to an Amphenol TNC jack that caused a production test failure:

Exactly what the whiskers are isn't clear (tin?) This wasn't related to lead-free solder, but it might still have been a RoHS issue.

-- john, KE5FX

Well maybe, (Could be just fuzz.) the meter pic was taken with my shaking hand, the vise pic, with the camera in a tri-pod. I was mostly taking the vise pic for grayhill, to show that there wasn't some big solder blob in there.

George H.

Grin... I sent the same pics to grayhill. I had to call, to get a response. Well it can't happen, and it's some solder blob, or cleaning issue.. They were not interested.

If it's a real problem I'll get another call from a customer.

George H.

Well the switch body looks to be some cheap "pot" metal, almost anything could be in there. The body is grounded and the pins only get +/-15 V. Whiskers would have to grow ~0.050", is that reasonable?

George H.

It might be a few years before whiskers can grow that long.

If you open it up and post a photo of any whiskers that you find inside it on Twitter or something I think they might start to care. As far as I know they are mostly in the business of selling into high reliability markets so they ought to care if their customers percieve their products to be unreliable. For some reason I get more response to any complaints that I make via public twitter.

Does Elma make a compatible alternative switch?

Chris