Why i should buy a scope ( maybe not ) ?

[Chuck H]

What is your cathode resistance now? You can divide the cathode voltage by the resistor value in ohms to determine the current through the tube. Then multiply that by the cathode to plate voltage (420V) and you'll have your idle bias wattage. Divide by two and you'll know (roughly) how many watts each tube is dissipating at idle.

And... Yes, the JJ 6v6 is known to be the go to tube for a bunch of vintage amps that were already around 420Vp when 6v6's were tougher and the amps were plugged into 115VAC (USA). Now with wall voltages typically at 125VAC people are getting 450V on tubes that aren't as tough as the old ones. Enter the JJ 6v6s. It's been around a long time, but I have reservations about designing with it because it's the only tube of it's type. What happens if/when they stop making it? I guess you could use a 6l6 type tube. That was the standard fix for the vintage amps I mentioned before the JJ tube became available.

Since you replaced the power tubes and the hum went away I'll wager that your power tubes were badly mismatched or you may have even had a dead tube. This might also explain the bad tone. I don't suspect the choke in this case, even though this is another change you made, because hum from the power supply would have been 100Hz, not 50Hz.

[HaroldBrooks]

I would love it if this turns out to be a ground fault on the input jack Here we are going through these scenarios and isolation techniques, and Enzo starts with what is REALLY step one.

Good Quote and thoughts. Begin from the beginning.

I've had enough cases when there was no sound in my rig, and first thing to check is the amp plugged in, and the volume turned up on the guitar.

With hum issues should start with the guitar plugged in, generally speaking and see what happens when the volume is rolled off.

Many times it's nothing exotic causing the hum. Preamp tube swaps have helped me reduce hum in some amps, for sure. You guys always give good advice and I've learn a lot here !

[tepsamps]

So i took the other amp, with same power stage, put it to 455 V and no real problem with it except that the power tubes (6V6S) don't give the max power as usual. Generally, when i switch on the amp, i have the sound after ten seconds more or less, and it stays like this. With those power tubes, i get sound after 10 seconds, but it's not at max power, so i turn on the master and after one minute or so, the level increases, reaches the max power slowly, and that's ok, i'm at the max power and i turn down the MV because it's too loud.
Strange no ?
Are my tubes going to burst ? Or is it no big deal ?

[Chuck H]

So i took the other amp, with same power stage, put it to 455 V and no real problem with it except that the power tubes (6V6S) don't give the max power as usual. Generally, when i switch on the amp, i have the sound after ten seconds more or less, and it stays like this. With those power tubes, i get sound after 10 seconds, but it's not at max power, so i turn on the master and after one minute or so, the level increases, reaches the max power slowly, and that's ok, i'm at the max power and i turn down the MV because it's too loud.
Strange no ?
Are my tubes going to burst ? Or is it no big deal ?

Probably not a big deal. Could be though. It may get better as the tubes wear in. Or they may fail. New tubes are rarely tested or burned in appropriately before getting to market anymore. I've had several tubes, mostly JJ tubes, fail in reasonable circuits right out of the box.

A schematic would help. Not a description of the circuit as perceived, but an actual documentation of the actual parts as they're actually connected. All of them. Otherwise answering is an exercise in speculating.

[tepsamps]

Here is the schematic, i think i put everything but the tone stack, the V4 circuit, the fx loop is faulty. The tube is OK, but if i connect it i've a bad noise, very loud, unplayable. So you see i have connected both triodes in V4, and i did not investigate yet. I should connect only one triode, but is it a mistake to run double // triode on a cathode follower ?

[Chuck H]

THANK YOU! for providing schematics.

Unfortunately I'm not seeing anything that should be causing the slow charge to your signal chain. But there may be something like a wiring error or incorrect resistor value in the B+ supply that could cause the symptoms you describe. Checking preamp plate voltages may reveal that. What I DO see in the schematic is a few fundamental design flaws that might explain other issues you have been dealing with.

1)Your parallel preamp arrangement has the two paths feeding V3a with out of phase signals. This would certainly cause tonal detriment and volume oddities depending on how the gain controls are set. The most likely symptoms would be low signal level and thin tone.

2)The tone stack/master circuit eats a lot of signal. You then have that signal going to the FX send circuit which (ideally) reduces the signal and converts it to a low impedance. Usually about a volt of signal. You then have that going through a level control. Without a recovery stage and a return level control you cannot adjust for unity gain and the level control basically acts like a second master volume. But the real problem is that without a recovery stage you have only an effects level signal feeding the phase inverter. One volt of signal to the phase inverter cannot produce full power in the amplifier. V4 does not need to be parallel for any reason and it would be more typical of good design to use V4b as a recovery stage. I, personally, do not like level controls on FX loops. They are a source of problems when the levels are mismatched. But if you do design in a recovery stage and keep the input level control then the recovery stage will also need a level control so that the signal can be balanced.

3)A bit of a nit pick, but your power supply filtering has like phase stages at both preamp nodes. This is generally considered bad because if the filter capacitors are not in tip top condition the preamp may oscillate due to positive feedback loops in the power supply. Some commercial designs have done this and it's a common problem with those amps that requires attention often. No reason not to avoid this in a design.

4) This one is a safety issue! You should move your "bleeder" resistor to the downstream side of the fuse. As it is, if the fuse blows there is no way for the filter capacitors to discharge. Making a potentially dangerous high voltage condition for anyone servicing the amplifier.

There are other design aspects that wouldn't be "my" personal choices. But these aren't function or safety matters. I pasted your schematics together for the purposes of clarity:

[Helmholtz]

Two remarks regarding the schematic:

Supply filtering and overall stability would be greatly improved with series dropper (5k to 10k) resistors before each of the 10µ filter caps. An additional cap without its series resistor makes no extra filter or node.

The way the LTPI is wired, it won't work (properly) as you are feeding the input and feedback signals to the same tail connection instead of directly feeding the grids. Also the right side 100nF cap AC shorts the main tail resistor.
In other words both 100n caps are wired wrong.

[Chuck H]

Good catch. The symmetry of the 1M resistors caught me off my game. I think it must be a drawing error. I moved the 1M resistors in the second image below. I believe this makes the circuit correct.

[tepsamps]

Great thanks Chucks!
I did not see the 2 out of phase signals! In the begining of this amp, i put 2 gain controls to use them in the same signal path. But i don't need this anymore, and i have 2 gain pots, so i had this idea, but not a good one. I did not try to use both gains at the same time, but i've already heard problems of level and thin tone now you tell it.
Generally i use relays to switch channels, and i found smart to have a crunch and a lead possibility without relay, only a gain control.
I have to redo my main board i think. And rethink the thing...

[tepsamps]

About the fx loop recovery stage, i remove it because i find tube amps always too loud, so it's a way to lower the power and have a master volume for both channel and more flexibility in the channel volume tuning.
About the preamp filtering, yes i know it's weird but i used the diezel VH4 design with all the preamp stages at the same voltage, to get max voltage without dropping resistors. But they don't put a cap at each stage either...I find that 10k dropping resistor at each 12ax7 for instance is too much, and you loose a lot of volts arrived at V1a. I think that in the 60's or 70's, the capacitors were more expensive than now, and to have a V1a well filtered they had to optimize the filtering and use several dropping stages because they had only 3 12ax7 PI included. But now we can use big caps at the begining of the chain, have 6 12ax7 or more and have the whole preamp well filtered after the PI and the fx loop. If i put 6 dropping stage, yes V1a will be well filtered, but at which voltage?

[Helmholtz]

I find that 10k dropping resistor at each 12ax7 for instance is too much, and you loose a lot of volts arrived at V1a.

In a Marshall the first tube runs at around 300V, which is about perfect for an ECC83. You don't need 6 dropping resistors, 2 more would be sufficient. A 10k resistor drops around 10V per connected triode. The gain difference between 300V and 350V is merely 6%.
But the dropper resistors not only improve filtering but also decouple stages and thus improve stability.

But it's your decision.

[tepsamps]

But it's your decision.

Don't be like that Helmotz! I try and i learn, for now, i had no problem with that, and it's not on every amp i make, but i prefer my 375 V at V1 to 280 V as it was before.
Thank you for the advise...

[Chuck H]

Here's a good tool for determining approximate node voltages. You DO need to know the PT specs and current at each node. Unfortunately it doesn't do branched or parallel designs.

http://www.duncanamps.com/psud2/

[Chuck H]

You should check those resistors in your power supply and make sure they are on spec with the schematic. The strange behavior you report would be consistent with having a large resistor, like a 100k instead of a 1k in there.

Also, you really don't need 100uf for the screen supply node. It's a much higher than necessary value and only serves to increase inrush current and slow charge time. With 100uf for the main filter you've already done what you can.

Limiting drive to the PI isn't reducing gain any more than turning down the master volume is. In fact they are about the same thing. Limiting drive to the PI as a means of reducing output power is sort of like putting your car in first gear and breaking off the shifter to make the car go slower. You should build the amp properly and then use the master volume. That's why they call it a master volume.

[Helmholtz]

Also, you really don't need 100uf for the screen supply node. It's a much higher than necessary value

Exactly. Excessive screen node capacitance makes the amp feel stiff and lifeless - at least to me.