Tin whiskers in transistors other than the AF11x series.

[ukcol]

Originally Posted by pwdrive

Apparently tin whiskers can develop in the AC176/AC128 types.

http://www.markhennessy.co.uk/articl...ransistors.htm

Out of interest tin whisker faults can sometimes be induced by a sharp impulse/static voltage (probing with a meter sometimes), there was a recent case on the forum somewhere of this involving a valved Ekco radio where switching a light on induced the problem.

I've come across tin whiskers in WW2 military receivers.

Not saying it is tin whiskers....Remote diagnostics can be fun and games sometimes

Lawrence.

Originally Posted by ukcol

Hi Lawrence

Yes Mark Hennessy reports that TO1 transistors can suffer from tin whiskers:

"The final thing to say about these transistors is that they can suffer from the tin whisker problem. This is contrary to the "accepted wisdom" found on radio forums, and people are often surprised when I report it, but yes, it happens. Unlike the AF11x types, it's by no means certain that they will develop tin whiskers, but don't rule it out. Of course, there is no screen lead, so it doesn't necessarily cause quite the same problems that the AF11x types do - unless the transistor is mounted on the chassis for heat sinking purposes."

It will be interesting to know if any other people support this view from experience.

If so it will course probably be a good idea to make it the subject of a separate thread.

Originally Posted by mhennessy

While I've not been counting carefully, I'd estimate that I've found about 2 dozen instances. But then, I fix a *lot* of transistor radios

Without searching too hard, I could point to a thread on another forum where 2 other people mentioned coming across them. Believe me, it happens.

Hi Mark,
I'm not saying it doesn’t happen; it’s just that I need convincing by evidence from at least a few sources.

Since tin whiskers are not a problem of the transistor itself, but one of the packaging, I consider that it is highly likely that it may occur in other transistors. Indeed the problem is not confined to transistors; it has been reported in tuning gangs, potentiometers and even chassis components.

As far as I know the mechanism of the formation of tin whiskers is not fully understand but if other transistors were encapsulated in the same material and if some other (possibly unknown) factors were also present, then there is every chance it will happen.

It is easy to be convinced about the problem in the AF11x series because it happens in very large numbers and has been reported from multiple sources with the hard evidence of photographs and not just a fault symptom based argument.

I would like to see reports from a number of sources of tin whiskers in other transistors, with the reasons for suspecting tin whiskers as the cause of symptoms. I would be particularly convinced by photographic evidence as I have been in the case of the AF11x series.

[mhennessy]

Well, it's certainly good journalistic practice to "second-source" your information. It took me no time at all to type "AC128 tin whiskers" into Google:

Posts #7 and #8: http://golbornevintageradio.co.uk/fo....php?pid=32609

http://www.diyaudio.com/forums/solid...r-madness.html

http://www.antiqueradios.com/forums/...6123&view=next

All from page 1 of the search results. I didn't bother looking at the next pages, but then it's not me that needs convincing

Heck, Paul mentions it on this very website: http://www.vintage-radio.com/repair-...or-faults.html

I have never taken apart a TO1 transistor in an attempt to see the whiskers. As far as I'm concerned, there's no need to destroy an otherwise perfectly viable transistor - they can be "fixed" in exactly the same way as a TO7 package. Works well.

However, you can sometimes see whiskers on the outside of TO1 packages, just like you occasionally might with TO7 devices.

Like TO7 types, they can be intermittent, and they can respond to a sharp tap.

Unlike the infamous TO7 types, the true incidence of whiskery TO1 devices is likely to be higher than we realise because they only show up when the device is mounted on a heat sink. Or, when the whiskers are so bad that contact with two electrodes has occurred. Or, when the package comes into accidental contact with an adjacent conducting object that is connected to the circuit in some form...

[McMurdo]

HWMBO has a vintage Ekco portable which I suspect has tin whiskers in the output pair. If you turn it up beyond a whisper the sound goes off and the output pair start to heat up. If you tap them, it will recover for a few minutes then go off again.

[paulsherwin]

It's probably just a matter of time with any transistor with the same alloy used for the case. Problems with AF11x types only started to emerge after about 30 years, and they were particularly troublesome because the case was earthed by the screen wire.

I have no doubt that we'll see more and more tin whisker problems in AC126s, AC128s, AC176s etc in the future. I don't know if BC108 style devices will be affected but I wouldn't be surprised.

Ironically early glass OC44/45/71 types may live forever.

[Nuvistor]

Hi Paul,
I don't think it took 30 years for the AF 11x problems to surface. I cannot confirm it was tin whiskers in the 70's, but they certainly shorted to be the case, I changed many for that fault in that decade.
New transistors were easily obtained so they were replaced rather than any investigation in the workshop.
As for the AC128 type, failures were usually shorts to all terminals, no idea if whiskers were involved with the failure.
Frank

[turretslug]

Wasn't the gunk (silicon grease?) in some packages said to assist propogation of whiskers, in the manner of a crystal garden?

[paulsherwin]

Originally Posted by Nuvistor

Hi Paul,
I don't think it took 30 years for the AF 11x problems to surface. I cannot confirm it was tin whiskers in the 70's, but they certainly shorted to be the case, I changed many for that fault in that decade.
New transistors were easily obtained so they were replaced rather than any investigation in the workshop.
As for the AC128 type, failures were usually shorts to all terminals, no idea if whiskers were involved with the failure.
Frank

Tin whiskers won't have been the fault in the early years, as they take time to grow. I remember the AF11x series was always seen as a fragile transistor which wouldn't stand any abuse. When Mullard/Philips introduced it as the OC170 it was pretty cutting edge stuff.

[ukcol]

Originally Posted by mhennessy

Well, it's certainly good journalistic practice to "second-source" your information. It took me no time at all to type "AC128 tin whiskers" into Google:...............

Hi Mark

I am moving towards being won over by the anecdotal evidence ; If anyone has any photographic evidence I will promise to shut up.

[Refugee]

I once did a little one hour technology search on the web with the aim of finding out if nonotube solar panels can be made to produce power from gamma rays and tin whiskers came up in many other things but not AF117s to start with.
The most interesting of these was an IT centre that had some new faster data cables installed.
Over the following months power supplies started blowing up in equipment from multiple manufacturers and obviously a lot of money was spent on power monitoring over an extensive period.
Tin whiskers were eventually found to be the cause and they had been quietly growing on the galvanized steel cable ducts and the works had disturbed them. The air conditioning stirred them up and made them air borne so that they got drawn into computers and other IT equipment where they settled in the power supplies.
You get an awful lot more tin whiskers on galvanized steel than you do in transistors and I have found them and taken a photo in my very own workshop.

[mhennessy]

Computer flooring is another big source of whisker problems - I think that might be detailed on the NASA site.

I think I can beat your whisker farm there, Ref. Check out mine - this is from a Sanyo cassette deck fitted to a Hacker music centre

[kalee20]

Ah, but is that a crop of metallic whiskers, or is it corrosion /oxidation of the metal surface which has led to growth of spindly crystals of the oxide/hydroxide in a humid atmosphere?

If the latter, it's nothing like as serious as the growth of the crystals will be non-conductive.

[mhennessy]

Definitely conductive

[Herald1360]

Won't they be Zinc whiskers anyway on galvanized steel sheet?

[Refugee]

You would need to know the composition of the original plating if you were to work out what they are.
The growth mechanism is much of a muchness for both zinc and tin and the problems they cause are pretty well the same too.
Those wooly ones on the cassette deck would be closer to what caused the power supply failures in the IT centre.
The ones in my Marconi signal generator look thick enough that they would be unlikely to get sucked into a fan cooled bit of equipment.
I decided that the risk was low enough for me to let them well alone as they are not long enough to reach between the connections in the equipment and too heavy to get blown about.
In the IT centre they had been disturbed and that is what triggered the problems.

[Maarten]

An old repairman told me once that in the 1970's he tested metal case ACxxx transistors for defects by heating the case with a soldering iron and see if the fault cleared.

Of course this might help getting the transistor to work in case of more than 1 failure mode, but I see it as a weak anecdotal evidence that tin whiskers might have already been causing trouble back then.

[ukcol]

I believe I may have had a case of tin whiskers in an AC128.

Recently I changed the AF117s in a Bush TR130 with G322Bs and the results were very pleasing. However, I turned it on again recently and the sound was of reduced volume and grossly distorted.

Use of an audio generator and an oscilloscope proved that the audio amplifier was responsible with the scope showing a substantially rectified sine wave pointing to an output stage problem. A diode check on the output transistors revealed one to have an emitter-base short circuit. After removing the suspect I repeated the test at the print where the short was and sure enough the short had gone. However when I repeated the diode test on the AC128 I had removed the short circuit had gone!

Is it possible that the heat from the soldering iron had melted a tin whisker? Has anyone else had this experience?

A new AC128 restored the TR130 to normal health.

[paulsherwin]

That seems perfectly plausible.

I now suspect most tin whisker faults are not diagnosed as such in these transistors, as a fault will only become apparent when two connections have developed tin whiskers to the case, and superficial testing will indicate what looks like a bad junction.

[Leon Crampin]

It's entirely possible that due to the much lower circuit resistances in output stages, the available current levels in the c-e branch are such as to burn open any formative whiskers. This would explain why failures of AC128 and similar devices are seldom seen (or apparent).

This would not apply so readily to small signal devices as the emitter resistor will limit the current - hence the AF117 syndrome.

I would also discount pnp (AC176) failures, as the polarity is reversed which, I suggest might inhibit rather than promote whisker growth.

Leon.

[ukcol]

I believe the mechanism of tin whisker grown in transistors is little understood (if at all). How the evidence from the fact that whiskers grow in transistors that have never been used (new old stock ones) suggests that there is little or no electrical factor involved in there growth. So I would guess that polarity doesn't come into it.

[greg_simons]

It's one of the strangest phenomenons I know of, NASA has published quite a lot of stuff about this, it even occurs in new unused semiconductors for satellites and such like.
Greg.