Yacht-Boy diode puzzle

[G.Castle]

Originally Posted by paulsherwin

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

Very true, however they also used Elko brand which I've found don't age well, particularly the moulded versions, bin on sight, is the best plan.

[Boulevardier]

Originally Posted by 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.

Thanks Simon - a good summary of the steps I'd planned out in my head. The speaker-coupling capacitor feels like the most obvious and easiest first check, then the other caps and transistors. I'll post on progress. DC-coupled amplifiers are a bit of a nightmare for their intense connectivity! There's a high risk that you find the faulty components but don't find what caused them to fail - so it all just happens again!. I'll have to put a careful current-limit on the power supply.
Mike

[Boulevardier]

Paul, the electrolytics in this set are the darker blue ones rather than the lighter "powder blue" types that are often seen.

[Boulevardier]

I have now tested the speaker coupling capacitor (measured OK, but considering its age I replaced it anyway. The strange diode tested OK on the simple diode test, so I’ve left it in for now.
I’ve removed the AC187K and 188K outout pair and tested them (while doing so I discovered a very corroded soldered joint on the former, and reflowing this didn’t improve sound much, but did bring Iq down to a slightly more respectable 14mA.
With my Chinese tester (not so reliable on Ge junctions?), I got the following results:

AC 187K (NPN T9)
ICEO – 1.2mA
hfe – 143 (I assume “B” stands for hfe)
Vf – 90mV
ICES – 0.09mA


AC188K (PNP T10)
ICEO – 0.16mA
Hfe – 43
Vf – 143mV
ICES – 0.19mAV

The hfe's above are not typos, and they’re supposwed to be a matched pair. So it looks as though, although neither is totally shot, there’s a problem with one or both of them. I assume that’s enough to cause the low volume and distortion I’m hearing. Am I right?

Mike

[G.Castle]

An emitter follower PP stage normally doesn't require the transistors to be particularly well matched.

[agardiner]

I still think the problem is related to the quiescent current. It should be around 5.5mA with no volume. Distortion is the common problem when this is incorrect.

[Herald1360]

Depending on how "simple" the transistor checker is, the high hFE for the AC187K could be an artefact of its higher leakage current.

Does the pair check out OK for any leakage from junctions to case (whiskers)?

[mhennessy]

Quiescent current has been mentioned a few times now - I've been holding off saying anything in case it became clear subsequently, but as it hasn't been specifically pointed out, I'll say this now (but please forgive me it it's stating the obvious).

The 5.5mA figure quoted is across the test link. In my experience, a lot of people quote the total current taken by the radio when they report quiescent current. The OP hasn't clarified which it is.

The earlier high value was a bit alarming, but if it was the whole radio, with the dial light on, then it might have been fine. Likewise, 14mA for the whole radio might be OK. But if we really are measuring through the test link and the current won't go below 14mA, then we have further investigations to do.

Again, sorry if that's stating the obvious, but as I say, it's not apparent to me which it is.

Because of the difficulties of measuring Ge transistors, I always recommend testing by substitution. For a quick test, you don't have to use exact matches. Indeed, you could get away with a silicon pair (e.g. BC327/337). Obviously the bias will be different, but if you get normal working, albeit with crossover distortion and not enough quiescent current (via the test link), then that will tell you that one or both of the original transistors is playing up.

[Boulevardier]

Originally Posted by G.Castle

An emitter follower PP stage normally doesn't require the transistors to be particularly well matched

Strange that Grundig firmly specify a matched-pair for the OPs then. Perhaps over-engineering?

Originally Posted by Herald1360

Depending on how "simple" the transistor checker is, the high h[/I][I]FE for the AC187K could be an artefact of its higher leakage current

It’s an ATMEGA tester. Perhaps I should have said “cheap” rather than “simple”!

Originally Posted by Herald1360

Does the pair check out OK for any leakage from junctions to case (whiskers)?

Thanks, I hadn’t thought of that one, but yes, the OP pair check OK with no case-electrode shorts.

Originally Posted by agardiner

I still think the problem is related to the quiescent current. It should be around 5.5mA with no volume. Distortion is the common problem when this is incorrect.

Seems sensible to pursue that line of thought. I’ll test the other two transistors (common-or-garden types), and perhaps test/replace the electrolytics (only a small handful). Maybe the resistors too.

Rapidly getting to the point where I’ll have tested/replaced every component in the amp circuit! I did wonder about posting quiescent DC voltages, but, in a DC-coupled amplifier, I assume that just one fault anywhere would make most of them fairly meaningless.

Thanks all. Will report back, Mike.

[Boulevardier]

I hadn't seen Mark's comment when I just posted. Thanks Mark - I was measuring Iq properly I think by breaking the test-link on the pcb and thus measuring current at that point through the AC188K's emitter with volume control at zero. So yes, the 14mA Iq is certainly something that needs investigating.


I'll plod on with your suggestions and previous ones.


Mike

[Boulevardier]

Sorry, should have read "through the AC188K's collector".

[G.Castle]

Originally Posted by Boulevardier

“An emitter follower PP stage normally doesn't require the transistors to be particularly well matched.”
Strange that Grundig firmly specify a matched-pair for the OPs then. Perhaps over-engineering?

Grundig did usually build decent gear, but I think the close matching of transistors in push pull output stages like that was a mindset from the days of common emitter, transformer coupled circuits that also provided voltage gain, even then it was only perhaps to within 10%

Good luck with your repair.

[radiomobile]

If the fault is low volume and distortion instead of distortion at low volume then the actual AC187/88 quiescent current measurement could be a red herring. If the Iq is too low there will be distortion at low volumes and if too high the output transistors will run in to class A and get hotter (and in extreme cases lead to thermal runaway), but in neither case will this cause severe loss of volume. More significant is the mid-point voltage which I have not seen any reference to. If this is incorrect then the fault lies in the driver/preamp stages. As one contributor stated these d-c coupled audio stages can be tricky.

[Boulevardier]

Originally Posted by radiomobile

More significant is the mid-point voltage which I have not seen any reference to. If this is incorrect then the fault lies in the driver/preamp stages. As one contributor stated these d-c coupled audio stages can be tricky.

OK, I've put the OP transistors back in and measured the mid-point voltage (at OP pair emitters). It is 3.36V (with 7.5V power supply, Iq preset set to minimum 14mA). Increasing Iq with preset to its 52mA max takes mid-point V down to 2.2V. Setting the preset to give correct mid-point V of 3.25 produces an Iq of 47mA.



Mike

[agardiner]

Hmm, way too high. I will take another look at the schematic, but probably not until tomorrow now.

I have had the preset itself cause issues, but unlikely in your case as you can adjust it.

[ms660]

33 posts and no model number that I can see in order to see the full schematic?

I would have thought the mid-point voltage should be 3.75 Volts if the B+ is 7.5 Volts.

Lawrence.

[Boulevardier]

Originally Posted by ms660

33 posts and no model number that I can see in order to see the full schematic?

I would have thought the mid-point voltage should be 3.75 Volts if the B+ is 7.5 Volts.

Lawrence.

Sorry, it's a Yacht-Boy 208

Sorry also for my hasty arithmetic - yes mid-point should be 3.75V. The maximum I can get it to by adjusting the bias is 3.36V - that gives the Iq of 14mA.

Incidentally, on initial switch-on the Iq shoots up to something well over 1 Amp before settling back down to 14mA (fired the cut-out on my Avo 8 on 100mA and 1A ranges). And that's worrying since I've set the PS's current limit to 50mA!

Mike

[Boulevardier]

PS I did attach an extract showing the audio circuitry to post #1. I gave the Electrotanya version as it's the only one I found with good enough resolution for blowing up to legibility.

[roffe]

Grundig specifies the total current drawn at 7.5V to 22mA(no sound and FM or AM).
If we deduct 5.5mA for the output transistors and 7mA for the driver transistor we then arrive at 9.5mA for the rest of the radio.
Sounds reasonable to me.

[Boulevardier]

Originally Posted by roffe

Grundig specifies the total current drawn at 7.5V to 22mA(no sound and FM or AM).
If we deduct 5.5mA for the output transistors and 7mA for the driver transistor we then arrive at 9.5mA for the rest of the radio.
Sounds reasonable to me.

Yes, but the quiescent current I was measuring at 14mA was the current through the removable link on the pcb, and Grundig specify this as 5.5mA. So 14mA at that point is nearly 300% of the specified value. Have I misunderstood what you're saying?



Mike