Peavey Windsor - Hum and White Noise Hissing

[g1]

if it is an issue, I will look to see if I can find the SKA-3919 or equivalent transistor.

Sorry, I mis-read and thought you were dealing with hiss/noise at send output. If it's just hum it is likely a ground loop as Helmholtz stated above.

[The Dude]

I'll add: If you have insane amounts of extra gain somewhere, you would very likely have additional noise/hum along with that gain. Fixing one problem may well fix the other.

[TomCarlos]

Update - Perhaps I should "undo" my prior claim of loud. Everything in a shredder amp is loud to me.

Wanting to focus on the output section first...

I am inserting a 900hz signal into the Return jack. I am monitoring Pin 6 of the Phase Inverter, looking for clipping. That happens around 21vrms. I then look at Pin 2 of the Phase Inverter. That voltage reads 413vrms. So overall, the PI is yielding a gain of 50 (give or take).

The output only yields approx 9.vvrms with a non-symetrical waveform. The specs boast a 100 watt output. I installed new EL34 tubes and biased to 65%. The attached photo is of the output with the signal turned way down!

The output tubes have similar dc readings....

Pins 2/7: Heaters 6.5 vac
Pin 1: -41.4 vdc
Pin 3: 439 vdc
Pin 4: 428 vdc
Pin 5: -40.5 vdc
Pin 6: N/C
Pin 8: 0.015 vdc

I have a set of KT77s that I can drop in and rebias. (The amp came to me with the KT77's). But before I go there, does the above look right?

Thanks,

[Helmholtz]

I am inserting a 900hz signal into the Return jack. I am monitoring Pin 6 of the Phase Inverter, looking for clipping. That happens around 21vrms. I then look at Pin 2 of the Phase Inverter. That voltage reads 413vrms. So overall, the PI is yielding a gain of 50 (give or take).

Pin 6 is the plate output of the PI. 21Vrms corresponds to around 60Vpp with a pure sine.
That's too low. Available clean PI peak-to-peak output should be at least twice the bias voltage.

Pin 2 doesn't belong to the PI. Rather it's the grid input of the gain stage before the PI.
413Vrms is impossible here. Do you mean 413mV instead?

Please use capitals for electrical units like V, A, W, Hz.

[TomCarlos]

Sorry.... yes I meant to type 413mV.

It looks like you're implying that I have an issue in and around V3, as I am not seeing the maximum Vpp expected before clipping. So I will look there.

But even with less that max Vpp out of the PI, I am still wondering why I see what I do on the output.

Thank you.

[Helmholtz]

Your cathodyne (aka "split load") PI V3b has 2 opposite phase outputs: Plate (pin 6) and cathode (pin 8).
If you don't see pure sine signals there, the amps's output will not be a pure sine.
(A cathodyne PI has a gain below 1)

Supply, plate and cathode DC voltages (without signal) tell if the tube is working as it should.

What is the power tube grid bias DCV?

http://valvewizard.co.uk/cathodyne.html

[TomCarlos]

The EL34s are a matched set. The voltages on the output tubes are as follows:

The output tubes have similar dc readings....

Pins 2/7: Heaters 6.5 vac
Pin 1: -41.4 Vdc
Pin 3: 439 Vdc
Pin 4: 428 Vdc
Pin 5: -40.5 Vdc
Pin 6: N/C
Pin 8: 0.015 Vdc

Using my bias probe, I measure 38.5ma. So that puts it just shy of the 70% range.

[Helmholtz]

Ok, I wasn't clear enough.

What are the voltages at the bias test point P29, at B++ (P11/P12) and V3 pins 1,3,6 and 8?

[TomCarlos]

Helmholtz....

Thanks for the link to the Valve Wizard. It's helpful to compare that info to a real life circuit.

P29: -41.22 Vdc

B+, P10: 430.5 Vdc

B++, P11/P12: 368 Vdc

V3, Pin 1: 167 Vdc

V3, Pin 3: 1.36 Vdc

V3, Pin 6: 269 Vdc

V3, Pin 8: 97.5 Vdc

I should add, I rechecked the AC signal on V3b to compare for symmetry.

Pin 7: 20Vrms

Pin 6: 18.3 Vrms

Pin 8: 20 Vrms

[Helmholtz]

So the PI should be able to deliver clean 29Vrms ( 82Vpp) at either output.
I'll look at DC voltages tomorrow.

[g1]

Does the master have any effect when going in through FX return jack?

[TomCarlos]

G1 - The Master Volume does not effect the output when using the Fx Return.

I measured the ac swing on V3b up and until one side would begin clipping. Pin 6 begins to clip on the top half of the wave at 26.91 Vrms. Using the same input level, Pin 8 is at 29.26 Vrms (a clean signal).

Just for the hell-of-it, I disconnected the Output Transformer and measured the dc resistance.

Primary Brown/Red: 30.3 Ohms
Primary Blue/Red: 24.1 Ohms

Secondary Black/Orange .3 Ohm
Secondary Black/White .2 Ohm
Secondary Black/Green .1 Ohm

[TomCarlos]

Ok... part of the mystery solved…. Turn the Texture control to the “Class A/B Full Power” position!!

This being my first EL34 amp, perhaps I need to confirm the signal path to the output tubes. But... I took some ac measurements at the grid (Pin 5) of the output tubes.

I am still testing using the Return input. (I believe there is a separate issue at the input).

The first image below shows the output waveform just before clipping. You can see the voltage on the scope we get 17.46 Vrms @ 4 ohm load. There appears to be a slight crossover distortion. The other voltages are as follows:

V3b, Pin 6, 24.85 Vrms
V3b, Pin 8, 23.85 Vrms

V6, Pin 5, 23.6 Vrms
V4, Pin 5, 23.6 Vrms
V5, Pin 5, 24.6 Vrms
V7, Pin 5, 24.6 Vrms

Back in Post #10, Helmholtz mentions the signal path to include C15. I overlooked that. Yep, the Texture control is part of the output circuit. So if you set the Texture control to halfway, you get the second image (at 12.69 Vrms). So it is a tad more than 50% power but the waveform looks asymmetrical. Anyway, that is what happens when you mess with that control.

So I am not at a full 100 watts but I'll take 74 watts rms. But I doubt that spec as it did not specify the distortion level. For the power amp, I see that the output transformer may not be perfectly balanced and this is what could be leading to the different ac swings feeding V4/V6 vs V5/V7.

Anything else before I turn my attention to the preamp section?

[Helmholtz]

V3 DC voltages look perfect.

The TEXTURE control lowers the grid drive of V4 and V6, thus causing artificial imbalance.
Result is different harmonics texture (added harmonics) at the cost of lower (clean) output power.
If not set to full power, PI signals will be different because of NFB action.

No reason to suspect the OT. If there was a problem the effects would be much more dramatic.

I tend to think that the power amp is ok.

[g1]

I tend to think that the power amp is ok.

Agree, I wonder if V3 pin1 is already getting asymetrical when pin6 does.
Also whether output can get up to around 56Vp-p non-clean.