Talking AF117 etc

[vinrads]

I decided to test a load of AF 117 etc first testing them on my transistor tester if ok checked for leakage to the screen ,the leaky ones were subjected to a quick blast of 150 volts with c.b.e connected together ,then re tested most survived this treatment .Is this a long term cure?.
Time to try another method I heated the case, the build up pressure forced the cap off destroying the junction ,next clamped the case in the vice a piece of card was used so as not to conduct all the heat away heated the case, slowly removed the case success, the junction was intact.
Question do I re fill the case with silicone grease or just leave it.
Strange thing some transistors read NPN on my tester ,but worked ok on my test rig a Bush radio.
The yellow ones a bit of heat shrink sleeving , are the ones I saved with a shot of 150 volts. Mick.

[Techman]

The secret is if you're going to connect the cbe leads together and blast them, is to make sure there's absolutely zero resistance between them, in other words just clamping them with a crock clip is dodgy. However, in theory, the whiskers will eventually return. I just leave them alone and and disconnect the screens and I've never had a problem with this.

This subject came up in a recent thread (as it always does) and I said (as always) just disconnect/de-solder and bend back/snip etc. all the screens and it was argued that the whiskers could eventually short the junction, which theoretically they could, but in practice doesn't ever happen. Since that thread I've dug out an old Decca TP radio, which I must have disconnected all the AF screens some 15 or even 20 years ago when I dug it out of the loft and it wasn't working. I used this radio as a kitchen radio for several years and then retired it for another one. I dug it out again the other day and stuck a PP9 in it and it still worked fine. I've been using it on and off since and it's on now and I'm listening to it as I type this post, so I stand my posts regarding just disconnecting the screen leads and doing nothing more than that - the proof is in the pudding, as they say!

Edit: Removing the innards and filling with something else is interesting and something I've read about before, but never tried, so well worth experimenting, I would have thought.

[paulsherwin]

It's certainly true that you can usually get away with it for many years. The whiskers will continue to grow though, as they're produced by the metallurgy of the case. It's difficult to think of zapping, screen cutting etc. as anything other than a short term bodge. Sometimes bodging is acceptable of course

[Techman]

In theory you'd think that there'd be no reason for the whiskers to in some cases actually short across the junction, but for some reason (unknown) this doesn't ever seem to happen. Why this doesn't happen would be interesting to know.

[Techman]

There's another strange thing in that not all AF11x transistors fail. I happen to have a Sobell S314 standing on the side, next but one to the Decca (which I've just switched off) and I'm now listening to the Sobell on radio 4 long wave (while it lasts) and I've just checked inside and I note that it has three of the AF11x type transistors and none of the screens have been cut and the innards seem to be totally original - this radio would date to the mid 60s, but the Decca would date a bit earlier, I think.

[60 oldjohn]

The AF11x "Made in Holland" seamed not to be affected by tin whisker, I don't know if this reputation will last.

John.

My post crossed with the above, Would your Sobell happen to have Dutch transistors?

[Techman]

Originally Posted by 60 oldjohn

My post crossed with the above, Would your Sobell happen to have Dutch transistors?

Yes John, I've still got the back off and looking with the magnifying glass I can read on one "Mullard AF115 Made In Holland 2+P", The other that I can partially read is an AF117 and seems to also be made in Holland, the third I can't see its markings on without bending it about (which I'm not going to do), but expect it to be the same.

[60 oldjohn]

Originally Posted by Techman

I can't see its markings on without bending it about (which I'm not going to do), but expect it to be the same.

No that would be a crime to kill a perfect AF11x by breaking the wires off!

John.

[Vintage Engr]

Originally Posted by Techman

Yes John, I've still got the back off and looking with the magnifying glass I can read on one "Mullard AF115 Made In Holland 2+P", The other that I can partially read is an AF117 and seems to also be made in Holland, the third I can't see its markings on without bending it about (which I'm not going to do), but expect it to be the same.

That's made my day, as I have quite a few AF made-in- Holland series still in their original packing.

David.

[Silicon]

I was thinking of painting the inside of the metal case with a thick layer of cellulose paint.

If the original goo inside the case is silicone grease, it would have to be cleaned very thoroughly.

[G6Tanuki]

If you really want to bother with this, I would look more at lobbing the removed transistor caps into something that aggressively dissolves Zinc, so removing the problem rather than trying to cover over it.

But, pragmatically, why bother?? I have a bunch of PNP Silicon transistors that work just fine in place of the nasty AF11x types.

[JohnBG8JMB]

'I have a bunch of PNP Silicon transistors that work just fine in place of the nasty AF11x types.'

What type numbers, please? I have affected radios to fix...

John G8JMB

[Cobaltblue]

Originally Posted by G6Tanuki

But, pragmatically, why bother?? I have a bunch of PNP Silicon transistors that work just fine in place of the nasty AF11x types.

I understand the sentiment however these transistors were an important developmental phase in their time. As the IF stages no longer needed the neutralisation of the OC45's the designs became simpler and cheaper.

As the AF116 was in the Philips Electronic engineer EE8 my earliest hands on with new components I for one just like to see them around

Cheers

Mike T

[paulsherwin]

Originally Posted by JohnBG8JMB

'I have a bunch of PNP Silicon transistors that work just fine in place of the nasty AF11x types.'

What type numbers, please? I have affected radios to fix...

John G8JMB

As discussed in many other threads, a wide variety of Si types will work, particularly in MW/LW sets. They aren't a drop-in replacement though, and you will need to experiment.

You can buy 50 2N3906s from Chinese eBay sellers for a pound or two.

[Maarten]

My problem with blasting is that it doesn't destroy any whiskers that are still a fraction of a millimeter short of reaching other conductive parts. Something that tapping or removing the case and heating would solve better.

I don't think "Made in Holland" transistors in general are exempt, though I suppose it's possible they experimented with the exact tin alloy so some batches may be more prone than others (unfortunately I do not yet understand the date/batch/type coding systems used before ~1971). In general, every TO-1 or TO-7 encapsulated Philips transistor is likely to have some dendrites - until we can know for sure or by statistics that some are definitely exempt.

[jhalphen]

Hi to All,

ex-forum member Argus25 (Dr Holden) proposed this solution to whiskers & cleaning in Feb 2018 :

https://www.vintage-radio.net/forum/...1&postcount=15

Best Regards
jhalphen
Paris/France

[G.Castle]

I've done the zap treatment several times, the first was on an Ever Ready Sky Master maybe twenty years ago. I still have the set, and it still works.

Since then I have done others successfully.

I don't have any evidence as to why, but although I've come across several sets that use 'Made in Holland' devices, none have been faulty: I have found the same with transistors marked with Bel branding.

I was once told that the plating on the British produced transistors was pure tin, and that it was possible that the others were some type of alloy possibly with lead, as in solder? Though that is a purely anecdotal comment.

What I have observed is that metal encapsulation T01, 7, 18 etc., used in hi-rel scenarios were often specified to have gold flashing on the interior of the cans at a minimum.

The whiskers have been a known quantity in metallurgical and electrical industries for a long while.

[paulsherwin]

The Dutch ones do have a better reputation in this respect, but there seems to be a lot of variation generally. Not all the UK made AF117s have developed shorts (yet).

Does anyone know where the 'Bel' ones were made?

[G6Tanuki]

I had an idea.

If you can get the cap off a failed AF11x/OC170/171 type transistor, apart from pickling it in something that would aggressively remove the tin-plating [and probably any printed logos etc. on the outside, which matter to some people] I wonder if there would be any mileage in giving the cap a good clean inside to remove any of the silicone-gel stuff, then arranging the cap open-end-upwards in some sort of holder, with an electrode fitting into it.

Then fill with some kind of electrolyte, and apply a suitably-polarised DC supply to electrolytically remove the tin coating from inside the cap.

I guess there could be issues if the electrolytic stripping went too far and started to attack the Zinc in the brass of the cap. [The cap is brass, isn't it?]

Then clean it thoroughly inside, add some silicone grease, and re-fit.

I would of course add grease; though most people think of the AF11x/OC17x type transistors as low-power devices without needing to dissipate any heat, in times-past I built a 27MHz crystal-controlled radio-control transmiter whose output-stage used four OC170 in Class-C push-pull-parallel (common base, RF drive fed to the emitters, load connected to the collectors), 15V 'HT', putting out around 200 Milliwatts!

[wireless_john]

Interesting to read about the BEL transistors. I have just taken one out of a Decca TP99 which was originally fitted with three AF117. One of the two remaining AF117 was faulty and the BEL one was OK. I replaced them all with AF127.