Drastic Volume Change on Jazz/Rock Switch - Dumble ODS #102 Clone

[erwin_ve]

Thanks Electron.
Am I correct to assume the voltages look in line with a 102? Anything that stands out?

When measuring your voltage/current for biasing with the cathode resistor you also measure screen grid current. Your tubes in current state are biased around 46% instead of 49/50%. Try bias a little hotter, this will bring your B+ around 460v and your v1 around 195 and v2 around 202v. My personal taste is that will get you more in the ballpark of toneheaven. YMMV ofc.

Erwin

[frusciante89]

I did some more measuring and testing. Voltages do seem correct, but I still feel like something is wrong. For example, although the clean channel sounds good, as soon as I turn the PAB boost on, the sounds gets extremely brittle and fizzy... any further debugguing idea? I'm running out of options..

[rootz]

Personally, I do not like PAB on clean at all. Spiky without much low end. But never fizzy, but I guess that depends on the gain setting.

But please post a clip. I think your rock/jazz acts normal. I do get that behaviour in almost every ODS derived amp.

[martin manning]

I did some more measuring and testing. Voltages do seem correct, but I still feel like something is wrong. For example, although the clean channel sounds good, as soon as I turn the PAB boost on, the sounds gets extremely brittle and fizzy... any further debugguing idea? I'm running out of options.. :(

Lifting the tone stack boosts the level significantly above 100 Hz, which may be causing some HF oscillation. You could look for that with a scope. Might be lead dress or component placement on the PCB.

[martin manning]

I think your rock/jazz acts normal. I do get that behaviour in almost every ODS derived amp.

I agree, and so does the simulation :^)

[rootz]

I did some more measuring and testing. Voltages do seem correct, but I still feel like something is wrong. For example, although the clean channel sounds good, as soon as I turn the PAB boost on, the sounds gets extremely brittle and fizzy... any further debugguing idea? I'm running out of options..

Lifting the tone stack boosts the level significantly above 100 Hz, which may be causing some HF oscillation. You could look for that with a scope. Might be lead dress or component placement on the PCB.

Agree, if it sounds fizzy it’s time to get the scope and start measuring.

I find myself somewhat surprised if I see threads like these without screenshots from scope output. Measuring is one of your best clues in fault finding, apart from instantly recognising a problem because of experience.

[frusciante89]

I did some more measuring and testing. Voltages do seem correct, but I still feel like something is wrong. For example, although the clean channel sounds good, as soon as I turn the PAB boost on, the sounds gets extremely brittle and fizzy... any further debugguing idea? I'm running out of options..

Lifting the tone stack boosts the level significantly above 100 Hz, which may be causing some HF oscillation. You could look for that with a scope. Might be lead dress or component placement on the PCB.

Agree, if it sounds fizzy it’s time to get the scope and start measuring.

I find myself somewhat surprised if I see threads like these without screenshots from scope output. Measuring is one of your best clues in fault finding, apart from instantly recognising a problem because of experience.

In this example, what could be some useful oscilloscopes tests? Input/output waveform changes?

[rootz]

What you could do is measure the fizz you hear. What is it made of or how does it look like. Once you know that, you can start to pinpoint the exact location where it originates.

[llemtt]

I think it's acting as it should. Bode plot below shows R vs. J with TMB all at noon.

That sky plot looks wrong to me..

my .02
Teo

[martin manning]

I think it's acting as it should. Bode plot below shows R vs. J with TMB all at noon.

That sky plot looks wrong to me..

my .02
Teo

Hi Teo, what makes you think there is something wrong with it? The model is pretty straight-forward: Tone stack only, 39k source impedance, with appropriate log ("power") and linear taper pots, output taken from the top of the volume pot.

[llemtt]

Hi Martin, here is my simulation plot including first valve tube stage taken at the top of volume pot (I have also done real measurements and it's very close)

Screenshot 2026-03-30 at 17.25.50.png

actual output impedance of first tube is more 50-60k and there's the bass shelf of the cathode capacitor but your plot shows too much bass anyway, so difference should be in pots values and tapers, here are the pots I simulated

RF TS102.jpg

Teo

[martin manning]

Thanks for posting that. I consider the frequency response of the previous stage to be a separate EQ (here a 6 dB bass cut), and model only its output impedance. 39k is typical 12AX7 with 100k plate load, so maybe you are considering a high-plate stage, which would be 50k? In any case, I'm looking at the response of the tone stack by itself. What do you see for the difference in dB between Rock and Jazz? Is it similar to what I showed?

[llemtt]

Sorry but I can't find my old spice circuits with the Jazz switch (my amps use the Jazz switch to change the mid capacitor), but yes I remember there is a change in volume, more or less depending on the tone controls settings.

We must consider that almost all D amps are meant to be used "at the sweet spot" (where the power amp is compressing and clipping) so actually the Jazz mode doesn't change the volume so much but instead is a cleaner more Jazzy mode.