Eden WT800 (MY '06) low and distorted output on left amp

[tubby]

Problem is a low and distorted output just on the left channel power amp. Pls find schematic here:

So far I've checked:

- V+ and V- supplies: ok, ~80V
- Input Signal to the power amp section: OK on both channels

Power amp output on the scope with sine input looks good for the right channel, left channel shows a form distortion, see below.

Any experts can point me in the right direction?
Next step for me would be recapping the elytics on the power amp board and replacing the driver transistors judging by what I've read here so far about these amps.
Thanks!


eden_wt800c_sch.pdf

[Jon Snell]

You have checked the bias is around 7mV, without signal or load?
Check R19 & R22.

[J M Fahey]

Problem is a low and distorted output just on the left channel power amp. Pls find schematic here:

So far I've checked:

- V+ and V- supplies: ok, ~80V
- Input Signal to the power amp section: OK on both channels

Power amp output on the scope with sine input looks good for the right channel, left channel shows a form distortion, see below.

Any experts can point me in the right direction?
Next step for me would be recapping the elytics on the power amp board and replacing the driver transistors judging by what I've read here so far about these amps.
Thanks!


[ATTACH]n961891[/ATTACH]

1) is the amp driving a load?
What happens if you unload it, same waveform?

2) what RMS signal level do we have there?
Apparently very low, a few hundredths mV; wht happens if you rise it to real world levels, say 10-20V RMS?

Can that channel reach clipping?

Please show that waveform.

IF that funky waveform appears at very low levels only, it might just be caused by underbiasing, a VERY nonlinear area, guess JonSnell suspects something of the kind.

3) signal might come already distorted from preamp; also only non linear elements there are those Mute Fets.

Inject signal to get some 10-20 V RMS at speaker out and scope points 1 - 2 - 3 - 4 - 5

Show waveforms if not clean sinewaves:




Mute fets "should" be either ON (so shorting signal)or OFF (so doing nothing) -but- insufficient Mute voltage might partially bias them into behaving as resistors.
Quite non linear ones.

[tubby]

Thanks Jon & Juan!
I´ll do the recommended checks & measurements and post the results here.
Give me a few days...

[Helmholtz]

Don't forget that you have a working right channel. So you can directly compare voltages/signals.

[tubby]

You have checked the bias is around 7mV, without signal or load?
Check R19 & R22.

Bias measured 5mV on the good channel and 4mV on the bad channel. Adjusted both to 7mV. Cold, no load, no signal.

[tubby]

1) is the amp driving a load?
What happens if you unload it, same waveform?

2) what RMS signal level do we have there?
Apparently very low, a few hundredths mV; wht happens if you rise it to real world levels, say 10-20V RMS?

Can that channel reach clipping?

Please show that waveform.

IF that funky waveform appears at very low levels only, it might just be caused by underbiasing, a VERY nonlinear area, guess JonSnell suspects something of the kind.

3) signal might come already distorted from preamp; also only non linear elements there are those Mute Fets.

Inject signal to get some 10-20 V RMS at speaker out and scope points 1 - 2 - 3 - 4 - 5

Show waveforms if not clean sinewaves:




Mute fets "should" be either ON (so shorting signal)or OFF (so doing nothing) -but- insufficient Mute voltage might partially bias them into behaving as resistors.
Quite non linear ones.

1) Amp was driving an 8R load in that picture.
2) As mentioned before, that is not a practical signal level. Channel can reach clipping but does so too early. Pls continue below
3) I injected a signal directly on the power amp board. Find below scope pics and measurements with annotations below each pic.


Scope screen grab at the output of GOOD channel with 16R load.
53 Vrms at the onset of clipping is almost 200W and looks good.


Scope screen grab at the output of BAD channel with NO load.
Signal injection at point 1, level adjusted to reach ~15Vrms at output.


Scope screen grab at the output of BAD channel with 16R load.
Same signal and level injected as above at point 1.
~11.5 Vrms at output


Scope screen grab at the output of BAD channel with 16R load.
Same signal and level injected as above at point 2.
~12.2 Vrms at output


Scope screen grab at the output of BAD channel with 16R load.
Same signal and level injected as above at point 3.
~9.3 Vrms at output


Scope screen grab at the output of BAD channel with 16R load.
Same signal and level injected as above at point 4.
~11.2 Vrms at output


Scope screen grab at the output of BAD channel with 16R load.
Same signal and level injected as above at point 5.
~17 Vrms at output

In addition I measured for DC on the outputs (no load, no signal):
4mV on GOOD channel
760mV on BAD channel

[Helmholtz]

I think you misunderstood what JMF meant.

Question is if clean signal reaches e.g. point 4 - or if it's distorted/clipped by one of the mute Fets.
So inject signal at point 1 and scope the other points indicated in the picture.

[Jon Snell]

Bias is near enough to be fine.
If you remove Q1 & Q2 does the clipping stop?

[Helmholtz]

[
760mV on BAD channel

It should be possible to lower output DCV by adjusting the offset trimmer VR1.
But that can wait until the distortion issue is fixed.

[tubby]

I think you misunderstood what JMF meant.

Question is if clean signal reaches e.g. point 4 - or if it's distorted/clipped by one of the mute Fets.
So inject signal at point 1 and scope the other points indicated in the picture.

You're right, I obviously didn't read JMF's post correctly. My bad...

So, now I've injected a signal at P1 and scoped P2 to P5.

Result: I get a good clean sine wave at P2 - P4.
If I'm not mistaken that means the muting FETs work as they're supposed to, correct?

Scope pic at P5 looks like this:


If scope the output into an 8R load, the onset of clipping at the output starts at a drive voltage of ~50mV rms which gives an output of 736mV rms.
The waveform looks like this:


I can't get the output to go a higher than 10Vrms into 8R, regardless of drive voltage at P1. At that point the output waveform is heavily clipped at top and bottom portions of the curve.

[Helmholtz]

Ok, so it's not the Fets.
No surprise the signal at P5 looks like the output, as it's the NFB signal being a sample of the output.

Check C5 and C6 and measure transistor DC voltages (especially base to emitter) and compare with good channel.

[tubby]

Ok, so it's not the Fets.
No surprise the signal at P5 looks like the output, as it's the NFB signal being a sample of the output.

Check C5 and C6 and measure transistor DC voltages (especially base to emitter) and compare with good channel.

I have replaced C5 und C6. No change.

When I compare the tranny voltages to the good channel, the voltages are all off on the bad channel (except where -VCC and +VCC are concerned)
starting with slight differences at Q5 and Q6 and bigger differences at Q10 and Q11 (e.g. emitters on good side at 5,5V resp. -5,5V, emitters on bad side at 2,5V resp. -0,5V).

Are there general recommendations/rules in how to proceed in such a case? I'm having a hard time telling what's actually faulty and what isn't since everything is thrown off by each other (unfortunately not an AC coupled tube amp ).

[Helmholtz]

Are there general recommendations/rules in how to proceed in such a case? I'm having a hard time telling what's actually faulty and what isn't since everything is thrown off by each other (unfortunately not an AC coupled tube amp ).

I don't know a simple/standard method for troubleshooting SS power amps. But others having more experience with that stuff might.

I would measure circuit/transistor voltages and try to infer bad components from off voltages.
Also, it's always useful to directly measure base to emitter voltages. If not between 0.45V and 0.7V the transistor is not working properly.

You might fill in voltages to a copy of the schematic and post it here.

[g1]

Try using the diode check function of your meter with the unit turned off. Compare transistor readings between good and bad amps, looking for any major differences.