Workingman 15 thermal discrepancy

[Delta362]

C12 was off compared to C11 and C17. I replaced those three since they were right next to each other on the board. C16 was also replaced and I'm still at about the same negative DC offset as before.

[The Dude]

OK. Your comment, "R30A/B are both very much on at 22 ohm (not 220 as the schematic mis-shows)" got me thinking. Maybe the whole amp schematic is wrong for this unit and also circuit board designation numbers. See if this schematic better matches your amp.

LA12 and LA15 Power Amp Rev A.pdf

If this better matches your amp, I would check that Q7 & Q8 are matched rather than Q6 & Q7. If this schematic doesn't better match your amp, please ignore this post with my apologies.

[Delta362]

Seems the amp on the bench is a blend between the two schematics, and seems to go along with one of the previous threads on this amp with component values.

Please see attached for voltage readings I've taken with no load attached. Does anything look out of place to cause the DC offset?

[g1]

Q10 B-E of 0.8 V seems high.

[Helmholtz]

I'm seeing voltages at connected points differ by 0.2V.
To avoid such errors transistor voltages should be measured directly between transistor leads.

[Delta362]

I took updated pin to pin measurements. No load. Amp had been on about 20 minutes and stabilized.

One item to note is that Q6 and Q7 are MPSA56 in my amp, the board silkscreen says A56, and I confirmed the traces match the locations for B, C, and E.

Please see attached.

[Helmholtz]

Nothing obviously wrong with transistor voltages.

[Delta362]

Nothing obviously wrong with transistor voltages.

Thanks for confirming the voltages look reasonable. I'm pretty stumped on this.

[g1]

There are some SS amps with symmetry trim pots that affect the offset. Can't think of any off the top of my head, but a schem. for one might suggest which value to tweak.

[Delta362]

Something else I noticed. Voltage across R29 is about 1.8mVDC while R42 is 3.3mVDC. As I add a load and signal, the two stay roughly 2-3mA apart, however, the DC offset becomes less negative as the signal is applied. At full signal and output starting to clip, the offset is down to around -0.120VDC. It almost seems to me that if I can have Q4 conduct a little more, that should pull the offset in the positive direction.

Thanks for the suggestion, g1. I'll keep looking for some examples.

Getting back to Q3, how close does the hFE of Q3 need to match Q9? I don't have, and still haven't found any 2SC4382 in stock. "Due to temporary constrained supply, DigiKey is unable to accept backorders at this time.". Using Octopart, I found authorized stock in Japan, but it's $40 for shipping. Does anyone have another lead? Looks like Quest Components has some, but I'm not sure if they're selling legitimate components.

[x-pro]

how close does the hFE of Q3 need to match Q9?

Matching h21e for the preliminary output stage at Q3, Q9, is important at high output stage currents.
possible replacement of these transistors with:
2SA1667+2SC4381
2SA1859+2SC4883
2SA1859A+2SC4883A
2SA1930+2SC5171

[x-pro]

Voltage across R29 is about 1.8mVDC while R42 is 3.3mVDC.

• Workingman's 12-15-2004 Power and Pre Amp Rev 1.pdf (194.7 KB, 23 views)

There is an explanation for this. Only emitter current Q4 flows through R29.
The collector current of Q10 and emitter current of Q9 flows through R42. Plus also the current through R28 when there is no load at the output.
If the resistances are equal, they will have different voltage drops.

[Delta362]

• Workingman's 12-15-2004 Power and Pre Amp Rev 1.pdf (194.7 KB, 23 views)

There is an explanation for this. Only emitter current Q4 flows through R29.
The collector current of Q10 and emitter current of Q9 flows through R42. Plus also the current through R28 when there is no load at the output.
If the resistances are equal, they will have different voltage drops.

Thank you for that explanation. It makes perfect sense now.

[Delta362]

Matching h21e for the preliminary output stage at Q3, Q9, is important at high output stage currents.
possible replacement of these transistors with:
2SA1667+2SC4381
2SA1859+2SC4883
2SA1859A+2SC4883A
2SA1930+2SC5171

Unfortunately all of these options appear to be obsolete and not available at Digikey. I do eventually see MJF15030/31 as an option, though it has 30MHz vs 15 or 20MHz. Any concern with stability?

Any ideas on which component(s) to potentially look at or tweak to get the DC offset less negative?

[x-pro]

I do eventually see MJF15030/31 as an option, though it has 30MHz vs 15 or 20MHz. Any concern with stability?

Yeah, you can try.