2203 clone suddenly started to pop

[djseban]

Hi there guys,

back in 2013 I started a 2203 build with intention to later mod it to #34, the build was successful, mod was set in place. Later I added PPIMV. I played some gigs with the amp and it was totally okay. In 2016 I modded it with additional gain stage and the overall switchable config looks like this: #34/#36/Hot-rodded 800 - the changes are done only on preamp. Recently as I returned home I fired the JCM up, and what I heard first was some quiet windy-like sounds coming out from the cab - caps charging up after long time I reckoned. But suddenly amp popped very loudly and I quickly powered it off. Later as I procedeed to fix the amplifier, it continued to do these pops and it fried one of the tubes (measuring voltages I spotted one pair of EL34s was drawing very big bias current). I swapped the pair and I played on the amp for about 2 days - it started to pop again. Then I pulled out the turret board, cleaned it and replaced all parts with the new ones - amp pops again. After this try I rewired power supply section (with caps included) and put new sockets for the power tubes - I played it two times for like 30 minutes, and today I fired it up again - well, you guessed it, it fuc**ng pops again . I also spotted that when pop happens I can see a spark in the lower part of first power tube. Pops happen also without pre-amp tubes. Caps are JJ Electronics - 3 50+50uF and 2 100uF+100uF.

So I already rewired entire amp, the only thing left to try is capacitors. I am not sure though if I should replace all caps or try to search somehow for the faulty one. The thing that gives me doubts is that sometimes amp plays totally okay, other time I turn it on and after switching standby to on, it starts popping or to these windy sounds. My guess is one cap is shortening the circuit when voltage is turned on. When I rewired the caps I tested them all by measuring resistance - usually when cap is okay, then when you measure it resistance firstly raises and then drops - all capacitors were following that pattern.

What do you think guys? Did anyone maybe had similiar issues and cares to elaborate? I am really running out of ideas

[glpg80]

What’s your B+? First thought is that you’re over voltage of what the power tubes are rated for assuming no other problems.

Have you tried to bring it online with a variac and seeing if it’s internal arcing within the tubes or external within the chassis?

You mentioned redplating and it happening mostly when you come out of standby, what’s your bias supply look like circuit wise? Normal 2203 values?

[djseban]

What’s your B+? First thought is that you’re over voltage of what the power tubes are rated for assuming no other problems.

Have you tried to bring it online with a variac and seeing if it’s internal arcing within the tubes or external within the chassis?

You mentioned redplating and it happening mostly when you come out of standby, what’s your bias supply look like circuit wise? Normal 2203 values?

B+ voltage is practically textbook JCM 800 - 472V. Unfortunately I do not have a variac to play with. About bias supply it's also normal 2203 circuit. Only differing thing is PPIMV. Voltage-wise I set about -38V by the method with 1 ohm cathode-grid resistors, to get about 36-40 mV on one resistor. About red-plating it wasn't happening, once the amp popped it gives a spark in the tube, but once it's done it it continues to work normally

[glpg80]

Does the power transformer get hot to the touch?

I wonder, if your bias potentiometer is going open causing the tube to then conduct fully and internally arc? Have you tried replacing the bias potentiometer?

A capacitor that’s gone south and turned purely resistive could also take out the power transformer secondary if it draws too much current. It would show signs of overheating before this occurs.

[djseban]

Does the power transformer get hot to the touch?

I wonder, if your bias potentiometer is going open causing the tube to then conduct fully and internally arc? Have you tried replacing the bias potentiometer?

A capacitor that’s gone south and turned purely resistive could also take out the power transformer secondary if it draws too much current. It would show signs of overheating before this occurs.

I doubt that PT will get hot enough to check it in the time amp starts popping - I just want to save tubes as much as possible. Will it be reliable test if I pull the tubes out and turn on standby and check then? B+ will obviously be higher, won't it hurt the capacitors?

Bias potentiometer-wise I just replaced it, set the potentiometer to get the old bias setting, fired up the amp. Unfortunately amp again started to do these windy noises that usually preceed pops, but I turned off the amp before it happened. I spotted an interesting pattern - I fire up the amp, it heats for about 5-10 minutes and then I turn standby on - usually for 2 minutes it plays normally, then it starts to make noises for like a minute and then it makes a loud pop. Does it make any sense?

[danman]

Try disconnecting the standby switch and connecting the wires to complete the circuit. Power up and play the amp a bit and see if it still happens. Not unusual for a standby switch to fail and arc internally with the high DC voltages they see in tube amps.

[djseban]

Try disconnecting the standby switch and connecting the wires to complete the circuit. Power up and play the amp a bit and see if it still happens. Not unusual for a standby switch to fail and arc internally with the high DC voltages they see in tube amps.

I replaced the standby switch with a known good one, outcome is the same unfortunately. Appreciate the suggestion though.

@edit: Also swapped PPIMV pot from other build (after testing it of course). Result's the same - amp makes windy noises. Other than that I tested power and preamp tubes on other amplifier (also 2203) - they're fine. Really running out of ideas, I decided to recap and new TADs are on their way. There is actually one theory that would justify recap - original Marshalls have every one electrolyte 50+50uF 500V, therefore when we apply the voltage of 470V to it, it will bear it better than single 100uF 500V I guess. Really hope that recap will help

[glpg80]

Have you pulled all of the preamp tubes and turned it on to see if you still hear the noises? Will help isolate the sound from the arcing as they could be two different things.

[danman]

The windy noise you mention makes me think that one of the plate resistors on the preamp tubes may be failing.

[djseban]

Hey guys,

turned out to be the issue with PPIMV - the thing is I relocated PPIMV pot to front. As I did amps, I always wired it with 0,75mm finely twisted wire, but I always put the pot on back. Turns out that front placement of PPIMV pot needs signal cables (the ones going to EL34 grids through 5K6 resistors) to be more isolated from environment, otherwise the power amp will start to excite itself (I don't know proper english word for this phenomenon), hence the crackles and pops. I replaced PPIMV cables with a shielded cable, connected the shield to bias voltage and the issue is gone.

@gipg80,danman: I pulled out all preamp tubes, the issue was still happening

[danman]

"Oscillation" is the correct term. Happy to hear that you found the problem and got it fixed!

[glpg80]

Glad you managed to find and fix the problem, we all learned something new. One question though, normally shields are connected to ground reference. What’s the purpose of connecting it to negative voltage? That’s a new one on me and I want to learn.

[djseban]

Glad you managed to find and fix the problem, we all learned something new. One question though, normally shields are connected to ground reference. What’s the purpose of connecting it to negative voltage? That’s a new one on me and I want to learn.

Hey there, thanks ! About connecting shield to the negative voltage I am not sure, I tried searching in what aspect is it superior to ground shielding, but all I found was "it is better" on some forums. Majority of shielded PPIMV designs I found were done this way. If I could guess, connecting this particular shield to the ground won't produce any other outcomes than eliminating oscillations, maybe shielding will be weaker, correct me if I am wrong. Some designs like shield connected to anode voltage on preamp (hot-shield mod) take advantage of Miller capacitance and affect the tone, but this voltage is far greater (like 250V) than ~-42V of bias voltage, so I guess shielding PPIMV won't do anything tone-wise.

[danman]

I remember reading on this forum years ago that grounding the shield to the negative bias supply is just as effective as grounding directly to the chassis. I don't believe that there are any added benefits of grounding it there, it's just that it makes for a convenient and close grounding point that keeps the install looking neat and tidy.