Yacht-Boy diode puzzle

[Boulevardier]

I am working on the audio section of a YB - low volume, distortion, quiscent current around 50-100mA, etc. I have checked the quiescent-setting pre-set, and although it looks battered it seems to work OK.

Currently investigating the OP pair's base-bias diode. This is shown on the Grundig (Electrotanya) schematic (see attachment) and in the parts list only as "ECO-3393" (the other diodes are specified by conventional nomenclature). The Trader schematic is a bit more expansive with "Double sil: diode".

I don't know why Grundig have defined it like this. Can anyone throw any light on what it is? Just a straight silicon diode?

Thanks.
Mike

[joebog1]

It could be a dual or double diode. Voltage drop twice as much in other words. Also temperature coefficient will be much less. There were a lot of Japanese dual diodes produced.

Joe

[joebog1]

Its easy enough to measure, just lift one end and use a digital meter.

j.b.

[Radio Wrangler]

If the output transistors are silicon, two Si diodes in series would under ideal circumstances provide quiescent current control.

But the real world isn't ideal and this circuit will give poor stability over temperature. Shunting the bias diodes with thermistors is the proverbial "Trying to polish ****" Without some stable emitter resistors it turns into a battle of semiconductor tempcos. These techniques sometimes were used to work from very low battery voltages but even then the excuse was a bit thin.

David

[joebog1]

Output transistors are germanium. AC187 AC 188 pair. Was very popular back in the 1960's and early 70's, but you know that David. I built a few from the Mullard handbook. They do sound nice when they work however!!.

Joe

[G.Castle]

Would an amplified diode connected transistor be better instead of trying to find an exact replacement the bias diode if faulty? Just to confirm the bias network faulty could it be shorted to tie the output transistor bases to see if the quiescent current goes down? If not the output pair may be overly leaky. (Edit. Comments made without seeing the actual circuit, phone browser wouldn't allow download of the OPs PDF), Grundig often used a thermistor in the bias network bolted to the heatsink which could go O/C.

I have spent hours in the past trying to get thermal tracking into line in PA stages. The three legged fuse will never replace the fire bottle in my book

[joebog1]

There IS the traditional 100 ohm thermistor in between the bases of the output pair.
( If memory serves me: little round carbon disk with a brown dob of paint thereon )

It's still " I think " a double diode. The Japanese cases of these diodes actually have the diagram of two diodes in series. Unless I am very much mistaken
YES I am very much mistaken. Its a long time since I looked at German CCT diagrams.

Joe

[Herald1360]

Mullard/Philips used to make the BZY88C1V3 (and -C0V7) at the bottom end of the 400mW voltage regulator diode range. The -C1V3 was a double diode not a zener.

[Boulevardier]

Thanks for replies so far. I have to be out for most of today, but will return to the fray this evening.
Mike

[paulsherwin]

Just 2 1N4148s in series should be OK as a replacement.

Test the transistors though. AC187/8s are now starting to develop tin whisker problems just like AF117s.

[mhennessy]

Whatever voltage is developed across the diode, it is reduced by the potential divider formed by the preset resistor and the thermistor in parallel with R54.

The 100R thermistor in parallel with 560 comes out at around 85R. If we assume the designer had the mid-point of the 500R preset in mind when choosing component values, then that has a division ratio of a quarter...

Assuming we need between 0.4 and 0.6V between the two bases of the two Ge output transistors, that suggests the diode voltage should be between 1.6V and 2.4V, which suggests the diode voltage will be somewhere in that range.

Of course, the range of suitable voltages will be quite a bit wider than that. For example, setting the preset to 125R rather than 250R means that the theoretical range of voltages across the diode is 1V to 1.5V, and that's bang-on for a pair of Si diodes...

So I could well believe that a pair of silicon diode junctions would be inside that package, and I'd be quite happy to replace it for a pair of 1N4148s or similar if necessary. But as has been said, it's easy enough to measure the forward voltage drop of the existing part, and if it's in the 1.2-1.4V region, then it's probably fine...

[agardiner]

Another vote for 2 1N4148s in series, although my preferred diode here is a BA314 which is a stabiliser diode.

Your quiscent current is far too high though. It should be 5.5mA.

[roffe]

Then why do I measure 0.72V across the diode and 0.34V across the bases on a similar Grundig Concert Boy 209?
I've also checked several Grundig circuit diagrams,and all of them specify ca 0.7V across the diode.

(the driver transistor current is the same 7mA in all cases so the voltage across the diode should be about the same)

[mhennessy]

Yes, I could well believe that. With the preset adjusted to something like 85 ohms, that could give the voltages you report.

In other words, the diode could be double or single - without measuring a known-good one, it's impossible to know which it is. Others were speculating that it was a special double diode, and I simply ran the numbers to confirm that was possible. But your measurements seem to confirm that it is a single Si diode. It would be nice to know why they used the exact diode they did though.

And it would also be nice to know why they used a 500R preset when 200R would be better - it's always nice to not be at one end of the range of travel. But then again, they put the pre-set in the wrong place, whereby a failure (or just rise in value - not uncommon) will cause the Iq to shoot up. If I was building this today, I'd swap the fixed resistor with the preset, for what that's worth...

[MotorBikeLes]

I was under the impression that this was likely to be two diodes (Ge) in series with another across the pair in opposition. I may be wrong.
Les.

[Boulevardier]

Just got back home. Thanks for all the comments. I'll check the voltage drop across that diode. Like MB Les, I was also wondering whether it would be Ge or Si. The Trader sheet seems to be saying it's Si.


Just to clarify - I have no particular reason so far to think that the diode is faulty - I was just preparing myself for that possibility. I'll also need to check o/p transistors, o/p capacitor, etc, etc. and electrode voltages.


Mike

[Cruisin Marine]

Change that speaker coupling cap asap, had many go leaky in Grundig's, change all relevant electrolytic's while you are at it if you want reliability and decent audio.

[Boulevardier]

Thanks CM - sound advice!
Mike

[Hybrid tellies]

Hi Mike I have heard that electrolytic capacitors could be a problem
on Grundigs old and new. So I would be inclined to check any around the output stage first. If you still have high quiescent current then I suspect output transistors, bias diodes and if DC coupled the driver transistor as well.

[paulsherwin]

Electrolytics are normally the Philips axial ones with the transparent light blue plastic sleeves, which aren't generally troublesome.