Signal bleed at master volume 0

[psychepool]

I'm making a Diezel VH4 clone, and quite a bit of sound leaks out at master volume 0.

If I set the gain knob to 0, no sound comes out. However, when I raise the gain knob, the increase or decrease in the gain amount is reflected.

This leaking sound is thin and tearing compared to the normal sound of this amplifier.
If I start raising the volume knob here, normal sounds will start to mix in between the thin, tearing sounds.
If I raise the master volume to the normal setting, the normal sound level covers the sound that is leaking out, so it feels like I cannot hear the sound. So, when I set the master volume to 0, it feels like the bleeding sound coming out is always mixed.
The sound coming out is quite loud, about the same volume as watching TV late at night. Since noise is also accompanied by the amount of gain, it does not feel very pleasant.

When the master volume pot is set to 0, pins 1 and 2 of the pot measure 0 ohms. I can't completely trust my cheap multimeter, but I'm guessing that the problem may be due to a small error in the pot.

What should I check when volume leaks from a high gain amplifier with a master volume?

[martin manning]

The signal may be leaking between board mounted components, bypassing the MV.

[bepone]

Parasitic capacitance on any point

[psychepool]

The signal may be leaking between board mounted components, bypassing the MV.

Parasitic capacitance on any point

Could it be caused by overly dense components or wiring placement? What are the cases that mainly cause this phenomenon?

Could it be caused by incorrect ground placement?

[R.G.]

Yes, this phenomena can be caused by dense component placement and wire routing. Your note that the sound is thin compared to the normal tone is a great description of the sound of capacitive leakage. Tube grids are especially sensitive to stray capacitive pickup, which is why the conventional wisdom on tube circuit layout is to keep all connections to tube grids as short and direct and isolated from other wiring as possible. Long runs of wire carrying plate or cathode signals can couple to grid wiring that they pass. So can side-by-side components. The small stray capacitances form a high pass filter, making for the thin sound.

And yes, incorrect ground placement can cause signal coupling. The current returning to the power supply filter cap negative terminal carries an "image" of the signal in the tube it's coming from. If this current flows through a shared ground wire resistance with another tube's cathode, then the "image" of the first tube's current is injected as a voltage into the second tube's cathode.

There is a large body of knowledge about how to run wires and components to avoid crosstalk.

[bepone]

The signal may be leaking between board mounted components, bypassing the MV.

Parasitic capacitance on any point

Could it be caused by overly dense components or wiring placement? What are the cases that mainly cause this phenomenon?

Could it be caused by incorrect ground placement?

in high gain, if you make very wide chassis and put 7-10cm between ECC83s and put wide all other components, and not mixing wires, you will not have crosstalk,
you are limited to channel switching relay which will insert little bit

[psychepool]

Thank you for your reply. Parasitic capacitance is a concept I have never heard of before. Because I like making small-sized amplifiers with large output, I have always produced a high-density board layout, and there have been no major problems so far. This build's symptoms are particularly severe, and it's a pity that if I had known about this concept, I would have paid more attention to the layout.

All cables heading to the grid use shielded cables, but doesn't the shielded cable prevent all of this? Could it be caused not only by wiring but also by proximity between components?

First, I will try modifying the ground position. If component density or placement is an issue, it means the amplifier will have to be rebuilt, so I'm hoping that relocating the grounds will solve the problem to some extent. Anyway, I don't expect it to disappear completely, but I want to improve it to a certain degree of satisfaction.

This amp is a 35-40 watt amp using a Hammond 1760j and 6l6gc pair, and in fact, if you attenuate it to the level you would use at home through an attenuator, it is barely audible. However, since 40 watters are not used only at home, I would like to alleviate the phenomenon beyond a certain level.

Please mention some points that need to be carefully isolated. Would it be possible to isolate the shield wire connected to the grid from the wires heading to the plate or cathode?
What should I try other than isolating the wire? If improvement is not possible in the current layout, the board will have to be recreated with a new layout. I am asking this question because I think placement should be taken into consideration when creating a new layout.

Below is the circuit diagram of the amplifier I am making.

VH_lite_schematic.jpg

This is the layout.

Diezel_VH_Lite_Layout (1).jpg

Below is an image of the mid-production process. The reason I didn't post the current status is because I'm in the process of modifying various wiring, so it's messy and not easy to identify.

The size of the chassis is about 310 x 160mm.

20230924_171113.jpg

The four 10K/560pF/39K EQ coupling cap/slope resistor sets on the board have been confirmed to cause oscillation in the signal line, so they have all been removed from the board and hardwired to the pot.


If there are any points that could be improved upon revision, please give me advice.

[bepone]

Every component acts like antenna and receiver.. parasitic.
It is not enough "just" to follow the schematic, actually in high gain schematic means nothing, more important is practical implemantation.
In high gain you are dealing with 200Vpp signals around the preamp, and all is in account and it is critical, components placements,components types, grounding, orientation, heaters, distance between board and the chassis...

[psychepool]

I tried removing the wire connected from the master volume to the power amplifier (Phase Inverter input), and this thin tone bleeding appears the same as when the wire was connected and the master volume was set to 0.
In addition to the sound leakage problem, low frequency noise is also present.
Is this the basis for judging that Parasitic Capacitance is not the cause but is caused by the ground?

The current ground layout is like this. If you find any points that may cause problems, please let me know.

Diezel_VH_Lite_Layout_ground.jpg

[bepone]

which wire did you remove? and after that what test did you make? it is difficult to follow in text mode.

[R.G.]

I tried removing the wire connected from the master volume to the power amplifier (Phase Inverter input), and this thin tone bleeding appears the same as when the wire was connected and the master volume was set to 0.
In addition to the sound leakage problem, low frequency noise is also present.
Is this the basis for judging that Parasitic Capacitance is not the cause but is caused by the ground?

This test proves that the signal is not coming through the wire you removed; that is, it's coming through a different path. Might be ground; might be through the power supply; might be just capacitive coupling to anything in the PI or output stage.

Do channel volumes affect the loudness of the bleeding signal? What about shorting the input jacks?

Do you use the chassis for ANY grounds except one safety ground point and one connection to the first filter cap negative terminal?

[psychepool]

This is the current status.
On the preamp side, most of the wiring except for channel 3 has been removed, and all pots for other channels have also been removed.


The wiring is a bit messy as I tried to temporarily change the earth point.

vhgut.jpg

Even when the jack is not plugged in and all knobs are set to 0, there is quite a bit of idle noise. This symptom feels strange. Other amplifiers manufactured in almost identical environments have noise levels that are close to mute in this state.
There was low frequency noise, but it disappeared when I connected the main filter cap (OT 2nd center, Power Tube Screen) to the chassis with the PT B+ center ground.

The grounding location on the chassis is roughly like this.

vhground.jpg

Would the fact that the symptom persists even after removing the master volume wire be evidence that this is not related to the parasitic capacitance caused by the overcrowding of the preamp and is a power-related problem such as the ground?
Could it be a bad power supply capacitor?


When other channels were connected, manipulating pods on unused channels did not affect the active channels.

[psychepool]

While the issue has been resolved to an appropriate level and I will let you know the results, I would like to ask additional questions.


Thinking that incorrect ground connection was the cause, I tried connecting the ground in various ways, but there was no change.
Thinking that interference from components around the PI input was the cause, we changed the layout and isolated the components around the PI input by distancing them, but there was no change.

I was curious what would happen if I connected the PI input to ground, so I connected it.
Signal bleed is almost gone. It felt like about 97% of it was gone and you couldn't hear it very well unless you listened close to the cabinet, so you could say it was virtually gone.
I thought this meant that the problem was not caused by ground problems, PI input interference, etc.

But I looked at the circuit diagram wondering why it wouldn't disappear if I set the master volume to 0.
The circuit I edited has a resistor of 56~680K connected in series between the master volume of each channel and PI input.
Since this was in the circuit diagram I was using as a reference, I added this to my edited circuit without much thought.
But because of this resistance, even if the master volume is set to 0, a resistance of 56K or 680K is connected between the signal and the ground, and this seems to be preventing the signal from being completely muted.

mv_res_03.jpg

I removed this resistor and connected MV pin2 directly to the PI input.
The bleeding signal finally disappeared.
As the bleeding signal disappears, now I can enjoy the original heavy sound without the thin tearing sound mixed in, even at low master volume settings. Background noise also disappeared.

The original VH4 had a global master volume between the end of the FX loop return and the PI input, so I am not sure whether it was judged that there would be no problem if such a structure was selected for the preamp master volume.
As this resistance disappears, there is a feeling that the tone has slightly brightened, but it feels negligible.

I would like to just omit this resistor if there are no other problems caused by omitting it.
So I'd like to ask you a few questions.

1. What is the purpose of this resistor? Is it simply a filter for highcut? Or is it for channel mixing and separation?
I am asking this question because it does not seem to play a large role for the purpose of high cut, and it does not seem to play a special role for that purpose since other channels are completely separated by the relay even without it for the purpose of channel mixing and separation.

2. If this resistor is removed, will there be a problem when using a specific channel or switching channels?
This resistor is connected right behind the relay that shorts out the master volume out and is eventually combined into one and connected to the next stage (FX loop in the original amplifier). Could there be a problem caused by this omission?

3. Most amplifiers connect the master volume out directly to the next stage (FX loop or PI input), but amplifier circuits with a series resistance connection structure like this are also often found.
Do all amplifiers that adopt this structure have signal bleeding problems? Or is it normal for there to be no signal bleeding even if a structure like this is chosen?


Please reply.

[sluckey]

If only one channel select relay will be energized at any time you don't need those resistors so just remove them.

[bepone]

every ressistor act like antenna/receiver so this big value (hundreads of kohms) was collecting signal around,