Fuzz pedal (blocking distortion) makes one EL34 shows early signs of red plating in 50 watt

[Gtr0]

Hi Friends.

I understand that blocking distortion can cause red plating - though I have never experienced the possibility until recently while using a Fuzz pedal.

I just finished a Marshall 1987 type circuit. I biased it a bit cooler than 70% so that the hotter tube is running at about 33mA.

Plates, both at 460VDC

v4 CT to pin 3 at 33.4Ω
v5 CT to pin 3 at 33.9Ω

voltage drop at v4 -0.80VDC
voltage drop at v5 -1.0VDC

v4 @ 23mA
v5 @ 33mA

bias voltage is set at -42VDC

Yes, I know - the tubes are grossly mismatched. I have three pairs of tubes I use as "beater" tubes in situations such as this - first startup etc. This was the better of the three sets unfortunately. This time I just didn't happen to have a good balanced pair on hand - just ordered a set of TAD EL34B-STR - should be here on Wednesday. I think at that time I will discard the three sets.

Anyway, here is the scenario... Amp sounds very good. I play this type of circuit with many controls on max, bass on 0. Everything up to this point is great. Once I introduce a fuzz pedal or a uni-vibe (in this case a Fulltone 70 for Fuzz and a Fulltone MDV-3 Deja Vibe) v4 slowly starts to show signs that eventually it will red plate... there is a very slight glow that begins to take place. I assume if I play through it long enough, it will go into full blown red plating.

And v4 is the "cooler" of the two. v5 shows no sign of glow whatsoever. So I am curious to better understand this. Can someone give me some details?

What are the chances this is happening in other amps that I have that I am not seeing because the other chassis are all installed in their boxes?

Just FYI, when the tubes are better matched within 2 mA or less, I set the bias to be more in the 50% to 60% range.

Thank you all in advance.

[Jon Snell]

A pair of matched valves will sort out your issue.
"Blocking distortion" is actually when the output valve is running into Class AB2 and drawing grid current. Most commonly found in Class C power transmitters.
Increasing the drive capability with a buffer can sometimes stop the distortion but still the anodes will glow if worked too hard.

[g1]

Swap position of V4 and V5. Does the fault follow the tube or stay with the socket (position)?

[Gtr0]

Swap position of V4 and V5. Does the fault follow the tube or stay with the socket (position)?

I have not tried since the "playing" part of this build started... but when I was biasing and getting the way off bias readings, I swapped then and the readings followed the tube. I will try this in the early afternoon.

Another thing I have noticed, one guitar in particular (a 2020 Fender Parallel Universe strat) that is very bass heavy (more so than my Les Pauls) also causes this to start red plating...

[Gtr0]

well, there is some not so good news... the red plate stays with the socket. that would mean an imbalance in the circuit? but last I checked bias was the same on both sides as well as plate voltage.

how can I proceed?

thank you!

[Jon Snell]

If you place a 10k 1/4W resistor as grid stoppers that may stop the parasitic oscillation, if there is any oscillation causing overheating. What have you got on your oscilloscope?
Could be layout if it is a home build.
Marshall used to bias almost cold 20 - 30%.

[Gtr0]

Thank you for your help, it is greatly appreciated... I am still learning and therefore I appreciate the direction.

Here are some pics of my scope's display...

RIGHT side (v5) grid...


LEFT side (v4) grid...


Left (v4) plate....


Right (v5) plate...


And the circuit image...


I hope I didn't mix those up... I have a tendency to do so sometimes ...

[Gtr0]

Had to write in another reply box, that started to become unwieldy... and for some reason it shows more attachments then there should be

So it looks like some blocking into the grid on the side that is red plating... I will try and trace back ward to see where it starts... the channel volume is about 5. I couldn't get it to red plate using a 500Hz tone... turned it up to about 8 on the volume.

Thanks!

[Gtr0]

I was scoping and using my DMM in the NFB circuit... on the tranny side of the 100k NFB resistor, I get an AC sine (according t the o'scope), but on the PI side of the 100k resistor I do not. The resistor reads 99k ohm. Is that a normal occurrence? I would think not.

here is the schem I used (1st one of the set)...
https://www.thetubestore.com/lib/the...-Schematic.pdf

[pdf64]

The mark space ratio at the power tubes looks a bit uneven. That can be a characteristic of the way that the preceding stages overdrive, or maybe originating with the input signal (the way the fuzz operates).
What are the waveforms at the input, and going into the phase splitter?
What does the HT Vdc sag down to at the plate (OT CT) and g2 (after the choke) nodes when the amp is being overdriven?
A potential issue being that modern PTs can be too stiff, insufficient degree inbuilt protecting resistance carried over from tube rectifier era designs.

[Gtr0]

The mark space ratio at the power tubes looks a bit uneven. That can be a characteristic of the way that the preceding stages overdrive, or maybe originating with the input signal (the way the fuzz operates).
What are the waveforms at the input, and going into the phase splitter?
What does the HT Vdc sag down to at the plate (OT CT) and g2 (after the choke) nodes when the amp is being overdriven?
A potential issue being that modern PTs can be too stiff, insufficient degree inbuilt protecting resistance carried over from tube rectifier era designs.

thank you for the info. I will check it out momentarily. Also, the input at these pics is a 500Hz sine, 1/2 volt I think...

[Gtr0]

just checking with my dmm...

plates @ idle = 458VDC
plates @ vol 3 = 399VDC
plates @ vol 5 = 395VDC

G2 @ idle = 453VDC
G2 @ vol 3 = 364VDC
G2 @ vol 5 = 356VDC

vol on 10 is little difference from 5 ( a few volts max).
Difference between left and right side output valves (v4 and v5) seems to be within a few volts as well.

I did not confirm directly, but if I remember correctly, the sine starts to show squaring off starting at around 3 on the volume pot. I can verify this if you need?

[Helmholtz]

I would first sort out the socket issue.

The pictures show no signs of oscillation but surely grid current limiting as grid drive is huge.
With a bias voltage of -42V, grid conduction will start when the grid signal exceeds 84Vpeak-peak.
PI asymmetry and grid currents cause asymmetric duty cycle.

I don't see much sense in scoping amp signals with the fuzz pedal as it only complicates the interpretation of the waveshapes.

[pdf64]

That’s fine, those measurements help to eliminate a super stiff HT as being a contributory factor here.

[Gtr0]

I would first sort out the socket issue.

The pictures show no signs of oscillation but surely grid current limiting as grid drive is huge.
With a bias voltage of -42V, grid conduction will start when the grid signal exceeds 84Vpeak-peak.
PI asymmetry and grid currents cause asymmetric duty cycle.

I don't see much sense in scoping amp signals with the fuzz pedal as it only complicates the interpretation of the waveshapes.

not sure if you missed it? I said I was using 500mV 500Hz input signal. The whole fuzz pedal thing was while playing... it is when it starts to red plate with a fuzz in front.