AF116 in the nude

[vinrads]

I just thought I would have a play with an old bush tr82d all three af's were duff whisker problem , I have a load of silicone bc479's =bc556 .
To check the gain I disabled the osc and fed an if sig in adjusted the scope so as to see the difference easily, then fitted two bc479's , there was a loss in gain which was to be expected , so I adjusted the base bias on the second IF , there is a 22k to the poss rail by using a pre set pot I found that a 22k across it brought the gain up to more than before , very good so far after some experimenting I left the first IF as was , any more gain sent it unstable. not very good as the ocs though fitted an AF126 .

Just for fun cut open a duff new stock AF116 cleaned it out , tested it on my super duper tester the best thing since sliced bread , it even tells you the pin out , very useful , fitted it in the radio see it working in the nude , all good fun , if you want to restore a duff AF117 etc I found cutting the case whilst gripping it in a small vice make the cut near the bottom about one mil up quite easy, it looks like the top will fit back on then a blob of solder should do the job . Mick.

[Craig Sawyers]

Wow. It is no wonder that early transistors were expensive when you look at the labour-intensive internal construction.

[greg_simons]

Remember when this first started becoming a problem, curiosity getting the better of me I opened an af117 to try and see what exactly the fault was, hey presto the short to ground was now gone and the transistor worked fine, little did I realize you need a very powerful microscope to see the whisker strands, i'm sure everyone knows this but there was a pile of stuff on the nasa website about this problem and how it afflicted early space electronics.
Greg.

[Graham G3ZVT]

Originally Posted by vinrads

I just thought I would have a play with an old bush tr82d all three af's were duff whisker problem , I have a load of silicone bc479's =bc556 .

It's ironic that germanium transistors are the ones filled with silicone.

I'm (gently) pulling your leg of course.

[vinrads]

Hi Graham I did notice but to late to correct it . Mick.

[flyingtech55]

Were these early transistors really hand assembled?

Tim

[Nymrod121]

Originally Posted by vinrads

I just thought I would have a play with an old bush tr82d all three af's were duff whisker problem , I have a load of silicone bc479's =bc556 .
To check the gain I disabled the osc and fed an if sig in adjusted the scope so as to see the difference easily, then fitted two bc479's , there was a loss in gain which was to be expected , so I adjusted the base bias on the second IF , there is a 22k to the poss rail by using a pre set pot I found that a 22k across it brought the gain up to more than before , very good so far after some experimenting I left the first IF as was , any more gain sent it unstable. not very good as the ocs though fitted an AF126 .

Just for fun cut open a duff new stock AF116 cleaned it out , tested it on my super duper tester the best thing since sliced bread , it even tells you the pin out , very useful , fitted it in the radio see it working in the nude , all good fun , if you want to restore a duff AF117 etc I found cutting the case whilst gripping it in a small vice make the cut near the bottom about one mil up quite easy, it looks like the top will fit back on then a blob of solder should do the job . Mick.

Just spotted this ... loved it. Nice one, Mick; there's hope for my various Roberts radios yet
Guy

[kalee20]

it will be interesting to see how long it works! I believe atmospheric moisture reacts with germanium.

Raytheon tried making plastic-encapsulated Ge transistors, but found they didn't have a massively long life due to the plastic letting moisture in.

[Nymrod121]

Doh! - now he tells me. I've been cleaning some out & just put the Karcher away

[Argus25]

Originally Posted by rambo1152

It's ironic that germanium transistors are the ones filled with silicone.

Ha Ha, I like that.

As noted if a silicon transistor is used it is important to increase the bias voltage a little to allow for the higher base-emitter voltage, or the gain will be lower. The easiest way is to tack a resistor across the base to negative supply resistor. Otherwise the only other way is to increase the base to ground resistor which is more awkward. Overall, it is probably better to replace with a non whisker affected germanium RF type.

[Argus25]

Originally Posted by kalee20

Raytheon tried making plastic-encapsulated Ge transistors, but found they didn't have a massively long life due to the plastic letting moisture in.

Yes and that problem seemed a non issue for silicon. I have a 1970's vintage Hughes Corporation LED calculator watch, they have bare logic IC's with the silicon junctions, hundreds of them, exposed to air....and they still work!