Volume/Voltage dropping when cranking

[viniciusvalenca]

Hi guys!

I've just finished this amp (schem attached) couple months ago and everything was going fine until last two weeks when I started to notice that when I dial both gain and volume controls to the max I got some volume drops when I pluck bassy notes. It's like a very low frequency tremolo effect but just when I pluck hard mostly the low strings. It's like the power section is going to a cut off region.. the volume drops quite noticeably. I got some hum noise (at quite acceptable levels, I'm not complaining about it) on it.. so when it happens even the hum noise goes away and then comes back in less than a second. It follows the volume drop.

When checking with the DMM the power supply, I have these idle voltages:(A) 396v, (B) 370v, (C) 330v, (D) 300v.

When playing, the (B), (C) and (D) voltages drop a lot. (B) and (C) drops about 50v. (D) drops about 25-30v. Is that normal?

I've got 23v across the V3's pin 8 to ground. I swapped the 470R cathode resistor to a 560R in order to get more power section saturation. According to my math, V3 is draining around 41mA (idle).

For that scenario, I assume that the plate is dissipating about 14w (not overall power, but the plate dissipation.. which is acceptable, according to the JJ 6v6s datasheet), right?

Well, no red plate. Everything works fine.. until I crank the amp.

Does anybody here have idea what's going on with this guy?

Any help will be much appreciated.

Thanks in advance.

[Littlewyan]

If the HT drops about 50V that can be pretty normal. In my 67 Marshall Clone the HT drops about 100V when cranked!

Does it happen consistently? As in, can you make it happen every time? Is it only really low notes that cause it?

[viniciusvalenca]

If the HT drops about 50V that can be pretty normal. In my 67 Marshall Clone the HT drops about 100V when cranked!

Does it happen consistently? As in, can you make it happen every time? Is it only really low notes that cause it?

Hi, Littlewyan!

Yes, everytime when I pluck low notes. Mid and high notes are fine...

[danman]

The fact that it happens when you have the volume and gain controls dimed may indicate that you have some sort of oscillation going on in the circuit. This will rob power and cause the volume drop that you are hearing. It may not be evident when the controls are set to a lower level. Do you have grid stoppers attached to pin 5 of the power tube sockets? This can help stop the oscillation as can adjusting your lead dress a bit. Moving a wire or two around can have a huge effect sometimes. If these fixes do not help, you may need to scope each stage of the amp to determine where the oscillation is occurring.

[Littlewyan]

I think you're right Danman and I think I recall that JCM800s can suffer from some oscillation, however I always thought it was high end oscillation and not low end.

Any photos of the inside?

[viniciusvalenca]

The fact that it happens when you have the volume and gain controls dimed may indicate that you have some sort of oscillation going on in the circuit. This will rob power and cause the volume drop that you are hearing. It may not be evident when the controls are set to a lower level. Do you have grid stoppers attached to pin 5 of the power tube sockets? This can help stop the oscillation as can adjusting your lead dress a bit. Moving a wire or two around can have a huge effect sometimes. If these fixes do not help, you may need to scope each stage of the amp to determine where the oscillation is occurring.

Yes! I built this amp according to the schematic. Everything is its right place. Do you think that 5K6 resistor is too small value for the purpose?

[viniciusvalenca]

I think you're right Danman and I think I recall that JCM800s can suffer from some oscillation, however I always thought it was high end oscillation and not low end.

Any photos of the inside?

I'm not at home now... I'll try to get some pics from the inside when I come back.

[Haze13]

When you crank up the volume, the voltage after the MV is very large. More than your bias -> the grid of the 6V6 goes to zero and tries to go positive. 6V6 can do that and you can see it from the Datasheet. What this means that you have a grid current that your CF (second half of 12ax7) have to supply. When this happens (or tries to) the input impedance of the 6V6 falls down and it's dominates the grid leak resistor that you have there. The cap with that resistor makes a LP filter and when impedance of a tube falls down, it's changes to the worst (High frequency roll off) and you have a BLOCKING DISTORTION + the value of a caps goes higher and it takes time to get to the normal state when the stop playing.

I said that the grid tries to go positive, but in order to do so, it needs CURRENT! Which 12ax7 cannot produce.
So it works very hard to buffer your tone stack from the last stage, and now you ask for more current for a grid of a 6v6. This happens for a very short period of time, but actually it damages the 12ax7. Put there Tung-Sol Ri and you will be surprised how fast it will fail

Ether use 12wd7 for a CF (the 12au7 half of the tube) + lower the voltage so the grid of the 6v6 will never be positive, or use 12au7 as another buffer - for a grid but the blocking distortion will kick in when grid goes positive. You can make a good driver (another gain stage before the 6v6) but it will provide you with more gain, and you have more then enough of it, even with the NFB.

Just something to think about There are more ways to deal with it...

[viniciusvalenca]

When you crank up the volume, the voltage after the MV is very large. More than your bias -> the grid of the 6V6 goes to zero and tries to go positive. 6V6 can do that and you can see it from the Datasheet. What this means that you have a grid current that your CF (second half of 12ax7) have to supply. When this happens (or tries to) the input impedance of the 6V6 falls down and it's dominates the grid leak resistor that you have there. The cap with that resistor makes a LP filter and when impedance of a tube falls down, it's changes to the worst (High frequency roll off) and you have a BLOCKING DISTORTION + the value of a caps goes higher and it takes time to get to the normal state when the stop playing.

I said that the grid tries to go positive, but in order to do so, it needs CURRENT! Which 12ax7 cannot produce.
So it works very hard to buffer your tone stack from the last stage, and now you ask for more current for a grid of a 6v6. This happens for a very short period of time, but actually it damages the 12ax7. Put there Tung-Sol Ri and you will be surprised how fast it will fail

Ether use 12wd7 for a CF (the 12au7 half of the tube) + lower the voltage so the grid of the 6v6 will never be positive, or use 12au7 as another buffer - for a grid but the blocking distortion will kick in when grid goes positive. You can make a good driver (another gain stage before the 6v6) but it will provide you with more gain, and you have more then enough of it, even with the NFB.

Just something to think about There are more ways to deal with it...

Hi haze13!

Thanks for the input. It sounds I need to lower the voltage after MV a little, right? To be honest I'm not toward to run drastic mods. I'm looking for a simple mod.

Yesterday I double checked it across the 6v6s pin 5 and ground.. I got some peaks like 15v (everything cranked up). Is that too high? I guess so... According to the datasheet the max acceptable is 13v. In order to lower it can I just lower the value of the 6v6s grid leak resistor (220k)? Or do I need to do another mod?

Thanks!

[Haze13]

The voltage across the cathode resistor is the voltage that sets the maximum swing. If it's say 15 volts, then that's the voltage that will produce a clean output of a output tube, more and it will saturate.

[viniciusvalenca]

Thanks everyone for the inputs.

I solved the problem in my way

I swapped the gain pot (1M) for a 250k one and the oscillation was gone. Plus I noticed this mod tame the distortion a little. It was a too harsh before, specially when I got the amp cranked up... Now it's more clear and creamy. I lost a bit of preamp volume, which is quite normal for this mod, but that's fine.

[Littlewyan]

Glad you solved the issue. What Haze13 says does make sense.