JJ 6550 max volts only 600?

[KRKozBassman]

If an SVT plate voltage is 660 volts why is it OK to put in a JJ 6550 if the Max voltage on a JJ is only 600 volts?

http://www.jj-electronic.com/images/...s/pdf/6550.pdf

[Justin Thomas]

For the same reasons it's okay to put a 6V6GTA in a Deluxe Reverb @ 420V. Voltage isn't what kills tubes, HEAT does. There's a lot more wiggle room in plate voltages being too high if you bias within reason - which basically means, if you're not dead-set on running your tubes as hot as possible, usually following the "interweb-lore rule" of 70% max plate dissipation. Personally, I think 70% should be an upper extreme, not an "ideal" setting.

I think running the screens safely and according to spec is a lot more important than plates. Usually when tubes die, it's a grid failure of some sort - often the screen grid softens, sags, and shorts to something else inside the tube.

Justin

[Enzo]

Plus 660v is only 10% higher than 600v. Tubes certainly are not worried about 10% variations.

[KRKozBassman]

Thank you! According to the schematic i'm setting the bias at the stated 24mV (across 10ohm resistor) which works out to be only 46%. 35w/660v*0.46 = ~24. Is this correct? Is this theoretically going to make the tubes last longer?

[g1]

It is a 1 ohm resistor, not 10 ohm.
The factory setting is .072V at the test point, which is for 3 power tubes, so it works out to .024Amps per tube.
That factory setting was for 42W 6550's so it was about 38%.
If you use 35W modern 6550's, the .072 setting at the test point will work out to 46% like you figured.

[KRKozBassman]

It is a 1 ohm resistor, not 10 ohm.
The factory setting is .072V at the test point, which is for 3 power tubes, so it works out to .024Amps per tube.
That factory setting was for 42W 6550's so it was about 38%.
If you use 35W modern 6550's, the .072 setting at the test point will work out to 46% like you figured.

Sorry but it's 10 ohms per tube - mine is an SVT-2 pro and an SVT-CL have the same set up with the LED/OP amp circuit. It doesn't have user test points on the back, you have to take the bottom off and check across 6 resistors.

[The Dude]

I guess I'm not understanding what the goal is here. Why not just use the LED bias procedure recommended for the amp? Also 24mV across 10 ohms would be only 2.4mA. I would think that would put you into crossover distortion territory (assuming we're talking about cathode resistors?).

[KRKozBassman]

I guess I'm not understanding what the goal is here. Why not just use the LED bias procedure recommended for the amp? Also 24mV across 10 ohms would be only 2.4mA. I would think that would put you into crossover distortion territory (assuming we're talking about cathode resistors?).

Sorry about that - it's 0.24 volts across a 10 ohm resistor - duh!

My goal is really to optimize to the best performance possible I am a engineer that works on automated machines ( I started working on tube and solid-state amps back in my early 20s - I'm 48 now) and over the years the manufacturer has not always been right so I guess I'm just being picky...

[soundmasterg]

The original SVT was biased pretty cold, and while the plate might have been at 660V, the screen was much lower....maybe around 350V, I forget without looking at a schematic. Anyway, having the screen a lot lower like that is a good way to get a lot of power and have the tubes last a long time too. I used some Svetlana (Winged C) 6550C's in the vintage SVT that I had, set at the factory recommended bias, and they were still going strong in that amp 10 years after I put the tubes in. The amp saw weekly gigging use for the whole time, and all tubes were still working perfectly when I sold the amp a couple years ago.

Greg

[g1]

Ok, SVT-2PRO. You mentioned SVT so I assumed you meant the vintage one. And the 24mA you mentioned worked out nicely to 1/3 of the vintage ".072V" spec. so I figured that was where you got the number from.
As you mentioned, the schematic shows .24V across 10 ohms. I would assume this is the correct level to make the light green, you can verify.
As far as tube life, with it biased at the stock spec., you should get the tube life they found to be the longest possible without having a negative impact on the performance.
Once again, it depends if they designed the amp with the 42W version or 35W version of 6550.
If stock was the 42W tube, using a 35W tube at the stock idle current will be 'relatively' hotter (% of max.W). Again, if it was designed with the 42W tubes, I would not expect the 35W version of 6550's to last as long as the 42W, and perhaps will not deliver the full power the amp is rated for.

[bob p]

Sorry but it's 10 ohms per tube - mine is an SVT-2 pro and an SVT-CL have the same set up with the LED/OP amp circuit. It doesn't have user test points on the back, you have to take the bottom off and check across 6 resistors.

I don't understand why you're pulling the chassis, flipping it over and checking voltage on the current sense resistors, when that amp is designed with a bias monitoring circuit so that you don't have to bother doing that. The official bias procedure for the SVT-2PRO is to adjust bias using the idiot lights on the rear panel, which are engineered to provide the right information. Comparators in the protection circuit will take care of the rest.

419xxh2 PA driver.pdf

[KRKozBassman]

Ok, SVT-2PRO. You mentioned SVT so I assumed you meant the vintage one. And the 24mA you mentioned worked out nicely to 1/3 of the vintage ".072V" spec. so I figured that was where you got the number from.
As you mentioned, the schematic shows .24V across 10 ohms. I would assume this is the correct level to make the light green, you can verify.
As far as tube life, with it biased at the stock spec., you should get the tube life they found to be the longest possible without having a negative impact on the performance.
Once again, it depends if they designed the amp with the 42W version or 35W version of 6550.
If stock was the 42W tube, using a 35W tube at the stock idle current will be 'relatively' hotter (% of max.W). Again, if it was designed with the 42W tubes, I would not expect the 35W version of 6550's to last as long as the 42W, and perhaps will not deliver the full power the amp is rated for.

The schematic says 6550A and the data sheet for 1972 GE 6550A is 42 watts so I going with that and the Tung Sol 6550 reissue I have in the amp also states 42 watts. So that would make it only 38% (42w/660v*0.38 = ~24). I think my OCD is satisfied.

[KRKozBassman]

I don't understand why you're pulling the chassis, flipping it over and checking voltage on the current sense resistors, when that amp is designed with a bias monitoring circuit so that you don't have to bother doing that. The official bias procedure for the SVT-2PRO is to adjust bias using the idiot lights on the rear panel, which are engineered to provide the right information. Comparators in the protection circuit will take care of the rest.

[ATTACH=CONFIG]44525[/ATTACH]

The LED's have a range they work in when green only is on that is approximately 15mA (24%) to 30mA (29%) and I just like to be precise. I know it doesn't really matter, it's just me.