Testing an AF117 transistor
 

The tin whisker problem of AF117 and similar transistors is well-documented. Out of interest I decided to test a couple of AF117s in my parts box. I am uncertain if they are new or used. Attached are the results from testing one on my component tester, which identifies it as a PNP transistor. I did not find any shorts between B, C and E and the shield on additional testing. Could it be assumed that this is a 'good' transistor (although could succumb to tin whisker disease anytime in the future ?)

Re: Testing an AF117 transistor
 

For a indeterminate length of time it may be absolutely fine.

There must be a few that don't succumb to the whiskers, but how would you know which, when, if they go down?

I've cooked a few with the iron and they seem to work, but for how long?
It has been established by NASA and others that the whisker growth is a continuous process so they all should fail repeatedly.
I have uncanned a couple, that works but its a pfaff and a half.

Re: Testing an AF117 transistor
 

just rebuilt 2 Roberts R500 where they were all still worked fine so left in

Re: Testing an AF117 transistor
 

Those readings look fine. 318mV Vbe is right for germanium and is used internally to power its telepathy circuit which it uses to decide when the exact most inconvenient moment is for a whisker to complete its short circuit.

David

Re: Testing an AF117 transistor
 

The transistor junctions tend to short to the can rather than each other. In practice I find that most measure fine on a component tester as this doesn’t connect to the shield wire which is connected to the can. A simple test with an ohmmeter between the shield wire and each other wire should reveal any shorts. I zap any NOS ones before use regardless of whether they have shorts.

Regards

Graham

Re: Testing an AF117 transistor
 

The tested AF117 worked fine as a replacement in a Stella radio I bought from the NVCF last year. It was silent before replacing the suspect AF117, which was producing a brief crackle in the speaker while tapping the transistor with a screwdriver.

The desoldered AF117 has a resistance of 640 ohm between collector and case. I might zap it at some stage or even de-case in an attempt to clean away its hairs.

Knowing that AF117s have a mind of their own, I installed sockets in case I need to replace the AF117 again.

Thanks everyone for the comments and helping out.

Re: Testing an AF117 transistor
 

I never have to replace AF117 type transistors, I either disconnect the screen or snip it off.

The possibility of an across the junction short is very unlikely, but is possible in about 0.01% of cases if you're unlucky.

Re: Testing an AF117 transistor
 

Alternatively change them for AF127. Direct replacement, reliable but different lead-out configuration.

Re: Testing an AF117 transistor
 

I did once use GT322Bs for another repair job, but have run out of them. They're reported to be reliable too.

Re: Testing an AF117 transistor
 

As I have posted before I use resin filled Tesla OC169 types (but now hard to get) or 2N2084's or AF178's (langrex have). One issue though, is it is nice to have longer lead wires than AF127 etc.

Perhaps one overlooked transistor, great for the job of replacing AF11x whisker affected types, is the beautiful if not wonderful Japanese 2SA234. Its transition frequency is close to the AF11X at around 60MHz and it has about the same collector capacitance around 3pF. Also, the transistor is easy to get from ebay or Langrex:

https://www.langrex.co.uk/products/2...ansistor-x1pc/


Over the years I have become increasingly impressed with Japanese Germanium transistors. Other examples include low noise audio types that the Japanese pioneered for audio tape recorder front ends, they led the world in battery operated portable tape decks. While UK/Europe had just the AC107 and across the Atlantic RCA just had the 2N2613, the Japanese had about 8 or more types all equally good as those two.The first time I encountered the Japanese ones was in the front end of an American made 8 track cartridge player and was astonished at the low noise.

So there are many good Japanese transistors out there for substitute parts in vintage radios, you just need to be turned on to them and aware of them.

Re: Testing an AF117 transistor
 

I keep an old set where the AF117 mixer oscillator is socketed,as this is quite a demanding test for functionality of germanium devices. Most unusual fault was an AF117 which was able to function as oscillator, but not mixer in a HMV set - replacing the device cured it.

Re: Testing an AF117 transistor
 

It seems more logical to me that a short across is only half as likely to occur than a single short to the can (as it requires two shorts to the can). Drilling a hole and rinsing could very well be the most reliabe method to get some more life out of them, followed by heating the can with a soldering iron. Disconnecting the screen would be the least reliable method, followed by zapping and tapping.

Re: Testing an AF117 transistor
 

AFxx types can be 'repaired' with a bit of work, to open them up I made a special ring bit for my Weller gun, (see pic) when heated this melts the solder
holding the can and the transistor assembly can be pulled out. (most times)
some are a bit reluctant. then I clean out the inside of the can with a cotton bud and IPA then coat it with Servisol plastic seal 60. This should stop new
whiskers from growing from the tin plating.
on some of them I replaced the silicone grease, on others I just removed it.
the can may be replaced by sitting the transistor in a bench vice with the legs
down in the gap between the jaws and tapping it back on with a wooden block.
I came by a big box of these transistors in a radio club sale, so have plenty to play with. most of them were leaky.
I have half a dozen set aside from about 5 years ago and I have just checked them out, they are still ok.
All this is described in more detail in an article I had in RB no 158 about vintage transistor radios.

Mike G3Zii, that was my previous ID on this forum but I had a computer
breakdown and could not get back in. so re-registered.

Re: Testing an AF117 transistor
 

Yes, it is interesting that. I found the same thing, the occasional transistor the top was reluctant to separate. In the end I gave up on applying heat because I detected some changes on before and after tests that led me to believe the junction had been affected. Though obviously they would have soldered these at the factory without damage to the junction in the first place. So I went to the method of drilling a small hole and doing a washout, shown in the photos on this post:

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

After a while though I concluded that the performance and reliability of the AF178 & others couldn't be beaten, so none of my radios have any of the repaired AF117's in them anymore, even though they would work ok.