Good batches of AF11x ?

[60 oldjohn]

I have a Masteradio D514 c1963 working perfectly with original AF115, 117. set only needed new battery holder. Just read a post by Fernseh from 20th August 2013, says same radio needed same repairs. No mention of the transistors, mine had the screen lead in place (not cut). I wonder if Masteradio got a possibly short lived GOOD batch of AF11x transistors. BTW Serial no 099063.


John.

[G6Tanuki]

I don't think there were any long-term 'good' AF114/5/6/7/8 transistors - the tin-whisker-growth does seemingly proceed at a variable rate (I can remember replacing AF117/OC170 transistors in the receive crystal-oscillator/multipliers and 10.7MHz IF-strips of Pye Cambridges/Vanguards back in the early-80s - we didn't really know about the whisker-issue then but replacing the transistors got you up and running again).

Pragmatically, I'd say your radio's only still working because of chance!

I recently was given a Bush TR130 [designed by Tom Karen - otherwise famous for styling the Raleigh Chopper bike and the Reliant Scimitar] which didn't work; on investigation the screen-leads had previously been snipped but the IF-amp AF117 had since gone dead. The owner wanted 'originality' so I bought and fitted a NOS AF117 - with the caveat that I viewed this as a sub-optimal solution and it was not guaranteed to last.

[paulsherwin]

They do seem to vary a lot, as with other notorious transistor types. I was looking at an amp the other day, and the advice was to swap out all the 2SC458s because they go noisy and leaky. There didn't seem to be anything wrong with them, but you never know if you've got a good batch or if they're going to fail next week.

[mickm3for]

4x af117 from Langrex NOS out of the 4 one had issues replaced without question so even NOS can be faulty just a matter of time and luck i think as when not if they go faulty Mick

[greg_simons]

Originally Posted by mickm3for

4x af117 from Langrex NOS out of the 4 one had issues replaced without question so even NOS can be faulty just a matter of time and luck i think as when not if they go faulty Mick

I'm sure this is old ground for many here but tin whiskers isn't dependent on use or not, nasa did a lot of research on this problem and found even new unused stock of some semiconductors suffered, the info's still on their website along with highly magnified pictures, makes very interesting reading.
Greg.

[60 oldjohn]

Most makes of radio with AF11x I have bought had issues. Maybe pure luck these two Masteradio sets had good transistors. The D514 seem fairly scarce, I would be interesting to know how many more there are with good original transistors.

[Nuvistor]

Originally Posted by greg_simons

I'm sure this is old ground for many here but tin whiskers isn't dependent on use or not, nasa did a lot of research on this problem and found even new unused stock of some semiconductors suffered, the info's still on their website along with highly magnified pictures, makes very interesting reading.
Greg.

Information much nearer to home.
https://www.vintage-radio.info/whiskers

[llama]

What a pretty transistor radio. If I see one I'll snap it up - obviously to contribute to this thread...
Graham

[MotorBikeLes]

I had a radio like that, not Masteradio, but from the same stable and looked identical (Regentone maybe?). i am afraid I de-gutted it, chopped thge middle out of the PCB, and soldered a bit of print C/W IC out of a little cassette recorder (Sanyo)?
Anyway, it was to be a signal tracer, with a diode probe and a simple probe with an attenuator. The L/S could be switched over to a load res, so that I could tune up a stage without blowing my eardrums out. I sort of modelled it on a Grundig signal tracer we had at a place I worked for.
Les.

[AC/HL]

Sobell S314, so there's likely a GEC version as well.

[60 oldjohn]

Originally Posted by AC/HL

Sobell S314, so there's likely a GEC version as well.

Yes, according to the other thread from August 2013.


Pretty set, Yes not bad for a fiver, unfortunately the plastic is vey brittle all the mounting points for chassis and speaker were crumbling. One other down side is the original Vol knob was missing it is an unusual size 4mm D shaft. The one fitted is one I had in stock, not the correct one. The missing strap is subject of another thread.


John

[60 oldjohn]

Originally Posted by llama

What a pretty transistor radio. If I see one I'll snap it up - obviously to contribute to this thread...
Graham

Thank You Graham, I wonder what the situation is with the GEC and Sobell S314 models regarding the AF11x working in the set.


John.

[suebutcher]

I've never had a dud AF11x in an Australian transistor portable. Again, that might be luck, or perhaps the Australian-made transistors had a slight difference in manufacture that gave them an edge over the British ones. Ditto lockfits, though that may be our lower humidity.

[Maarten]

It would not surprise me at all that some good batches can be found, since Philips did experiment with and change production in general.

@60 oldjohn: Even though I don't know how to decode them, it may be useful to know the batch codes on the ones that are found to be still good.

[Argus25]

"Good batches of AF11x ?"

There are none.

There are "better" batches which develop whisker disease later in the time line.

None can be regarded as a long term reliable proposition. This is why, in every radio I own, I have replaced them with the AF178, or the 2N2084.

[60 oldjohn]

Originally Posted by Maarten

It would not surprise me at all that some good batches can be found, since Philips did experiment with and change production in general.

@60 oldjohn: Even though I don't know how to decode them, it may be useful to know the batch codes on the ones that are found to be still good.

I once read, maybe on this forum that the Dutch AF11x were more reliable than the British made ones. The transistor in my set with AF115 wrote on the top is from Great Britain, The two with clear tops are Made in Holland. Batch number 2 + P. If we could find out what the P stands for could we cure the world's Tin Whisker problem ?


Sue, I wonder if it is your climate or are the transistors Dutch made, to have them working 55 years or so later.


John.

[Maarten]

Thanks!

I think the reliability difference between Dutch and British is more a lockfit characteristic as lockfits seem to fail less over here (as well as in Australia apparently) while TO1 and TO7 cased transistors such as the AF11x types seem to fail just as often in the UK as in The Netherlands.

The tin dendrite problem is quite wide spread, so finding some that are still working justifies taking a note of it to see if there's a pattern. If those types were ever made using a better tin alloy, this would not have been during the bulk of the European production anyway, since they fail so often and consistently.

On a side note, this may have come up before, but does anyone know whether the Matsushita equivalents used to form dendrites as well?

Unfortunately, it's not always clear where a transistor was made, even if it says so in print on the thing itself. Also, I don't know how to read pre-1971 batch codes but hope to find out one day (if anyone knows, let me know!). I will make note of this one in any case, to build up the data to extract this from.

[Maarten]

Originally Posted by paulsherwin

They do seem to vary a lot, as with other notorious transistor types. I was looking at an amp the other day, and the advice was to swap out all the 2SC458s because they go noisy and leaky. There didn't seem to be anything wrong with them, but you never know if you've got a good batch or if they're going to fail next week.

If the legs are black and sometimes even with visible dendrites, they will fail. The jury is still out on whether they can be fixed by just scrubbing them clean on the outside, but I know of a repairman who does this when original looks are important and hasn't had repairs bouncing back yet. Silver migration might be involved (as well as tin dendrites?) on those types of transistors.

[Argus25]

Originally Posted by Maarten

The tin dendrite problem is quite wide spread, so finding some that are still working justifies taking a note of it to see if there's a pattern.

I have found a pattern and it is this:

When I started evaluating the AF11x series transistor over 20 years ago, initially I though I had been unlucky with the ones I had that had failed. So I imported large stocks of them from the UK.

I built a test jig to evaluate them over the frequency range they would be in use in my Eddystone EC-10 radios, up to 30MHz.

I culled out about 20% of them on initial testing with whisker related issues. I assumed the whisker thing initially was bad luck and the good ones would remain good... that was a big mistake.

As a number of the "good ones" failed as the years past, I started cutting them open for forensic inspection (just like I do with failed electrolytics and learnt the horror stories that reside within) and noticed that the only difference between the transistors that had failed and the ones that had not, was not the presence of whiskers, all had them that I opened up, to one extent or another. It was the length of the whiskers, in the ones that were working, they simply had not reached the conductors or the junction area.

It was after that that I developed the methods to flush out the grease & whiskers (without heating the transistor). I had tried un-soldering the cap, but before and after tests in my jig indicated the transistor was degraded by the heat.

But in the end the testing in the jig showed that the gain and the noise was superior with the AF178 anyway, so I moved to those instead.

Sure, you may find some of these AF11x transistors that are working, but they are all circling the drain and from my perspective are expected to fail sooner if not later.

It doesn't really matter in a transistor radio though, it is not as if it is life support equipment, you can wait for them to fail before replacing them.

[Maarten]

I think we can all agree that in case of tin whiskers, no good ones will remain in the end and that most samples we encounter sport whiskers to some degree, regardless whether they were Dutch made or British made (I don't know exactly which factories produced them).

The only thing that I cannot guarantee is that Philips never experimented. Around 1962/1963 the first cases of tin whiskers in OC170/OC171 may have already popped up which would inevitably trigger the kwallab (quality department) to start experimenting. Maybe the outcome of this experimenting was "just don't bother and release the AF12x series" but the outcome could also have been to adopt some production lines to other processes (too little, too late evidently). Also, I don't know what role Matsushita played as (the only?) second source manufacturer for those types. They may have combined the Philips license with their own establised TO1/TO7 manufacturing processes. I've never seen other types and brands of Japanese transistors suffering from whiskers (but they could have).

So I can guarantee that if you spot a random AF11x, it should be assumed to be filled with whiskers, but I cannot guarantee that there aren't whiskerless AF11x at all.