why are these cathode based output tubes drawing so much?

[Randall]

The Holy Grail Supro is getting less Holy all the time. Pulled the top chassis to install a safety cord, and found one terminal on one speaker had never been soldered. I replaced the filter cap can with F&T's on added terminal strips, it looks quite time correct, actually. Now I want to slowly bring it up on the varian with no preamp tubes plugged in, just a stand in pair of 5881's and the recto. It comes up nice, no surprises, but when I check the current draw, it's thru the roof, around 70mA ea. I measure the 200R cathode resistor at 199.2 ohms, the 35uF cathode cathode cap at bit higher (I forgot), and even lift one side to see if maybe it was shorting, no change.

So, what could be making these tubes draw so hard? I tried others, and suspected perhaps leaky grid caps, but I see no voltage on pin 5. I'm at an impasse. Here is what I find on the outputs:

pin 2 to 7 = 6.3vac
pin 3 = 387 and 388 vdc
pin 4 = 386 and 387 vdc (the 1000R cement resistor shacks out)
pin 5 = 0.01 and 0.07 vdc
pin 8 = 27 vdc

That all seems to agree with the schematic, so what am I missing?

PS I am measuring with a bias socket, and have also done the cathode resistor drop, they match.

[g1]

That may be normal. Does the schematic not show a voltage at the cathode?
Here's one with a 200R cathode resistor that shows 27V across it.

[Chuck H]

The math says 25W total tube dissipation (so, not separating screen and plate). So, class A-ish? It seems a lot of amps, Fender, Supro, Ampeg, Vox, etc. from the 50's were biased pretty much class A. Whether they operated that way under high drive conditions or not. So, like g1 said (and showed), it's probably normal.

[mozz]

Grid leak resistors? I just had a problem with leaky coupling caps, I'm gonna say measure that 70mv on pin 5 and see if it moves up. My new caps brought the leakage down to 10-15 mv on each pin 5.

[Randall]

For some odd reason, it won't post the schematic, I've tried numerous times. It does show 400v on plates and 30v on cathode, so I am even lower than that with it on a variac at 115v.

I have never really grasped the class A concept very well. How is it a tubes can dissipate that much more current
and still be OK? And how do I know if it is probably OK, or not. It sure looks like it matches the schematic.

[Leo_Gnardo]

The Holy Grail Supro is getting less Holy all the time. Pulled the top chassis to install a safety cord, and found one terminal on one speaker had never been soldered. I replaced the filter cap can with F&T's on added terminal strips, it looks quite time correct, actually. Now I want to slowly bring it up on the varian with no preamp tubes plugged in, just a stand in pair of 5881's and the recto. It comes up nice, no surprises, but when I check the current draw, it's thru the roof, around 70mA ea. I measure the 200R cathode resistor at 199.2 ohms, the 35uF cathode cathode can at bit higher (I forgot), and even lift one side to see if maybe it was shorting, no change.

So, what could be making these tubes draw so hard? I tried others, and suspected perhaps leaky grid caps, but I see no voltage on pin 5. I'm at an impasse. Here is what I find on the outputs:

pin 2 to 7 = 6.3vac
pin 3 = 387 and 388 vdc
pin 4 = 386 and 387 vdc (the 1000R cement resistor shacks out)
pin 5 = 0.01 and 0.07 vdc
pin 8 = 27 vdc

That all seems to agree with the schematic, so what am I missing? PS I am measuring with a bias socket, and have also done the cathode resistor drop, they match.

If you find the dissipation worrisome you could select output tubes that have a low emission rating and/or swap in a higher value cathode resistor say 240 - 270 ohms.

AND! it was good of you to look for a DC offset at the control grid. First thing I do when I see output tubes drawing a lot of current.

[Randall]

No, this isn't a case where straying from the original circuit is on the table. I have to get it back to him as it was originally designed with as little changes as possible.

[Jazz P Bass]

Yes it's drawing a lot of current at idle.

Now check the current draw when the amp is running into a load.
Not much difference is there?

Welcome to Class A push-pull.

[Randall]

Help me with this guys. It is connected to the speakers. So I don't understand the above question. Or class A for that matter.

[Enzo]

Class A should be straight forward. A single ended amp has to be class A or part of the waveform will be missing. Class A says the tube is conducting 100% of the signal cycle. If we drove the tube into cutoff for part of the signal cycle, then that part would not come out the speaker. We would have clipped off parts of the waveform. That is single ended, like a Champ. We can run push pull amps the same way. The two sides are out of phase like any push pull amp, and the tubes always conduct. Both tubes are pushing and pulling at the same time. Like a two man lumberjack saw.

I think what Jazz is asking then is for you to measure current through the tubes when idling, then again when cranking signal through it into a speaker. In a Class A amp, the average current is always the same.


Class B is where the two sides of a push pull amp are biased right at cutoff, so the positive side tube only conducts the positive half of the wave form, and similar for negative side. One turns on just as the other turns off as the signal crosses zero. They hand off back and forth. This is more efficient in terms of power, but you now have to worry that the handoff is smooth and right at the correct instant, otherwise crossover distortion. - distortion caused by the handoff as the signal crosses zero not being precise.

Most of our push pull amps are class AB, which is in between. One tube continues for a time as the other tube picks up the conduction. After a period, the first tube goes into cutoff and thus drops out.

[Jazz P Bass]

Wonderfully worded Enzo.

" you now have to worry that the handoff is smooth and right at the correct instant,"

Hence the reasoning behind the bias adjustment of a Class AB amplifier.
No magic number.
No mojo.
Elimination of the crossover notch: which takes the circuit from Class B slightly into Class A.
Plain & simple.

[Randall]

I can't measure with a signal through a speaker just yet, I haven't gotten to the point of running preamp tubes as I want to determine the power chassis is OK first.

What makes my head hurt here is if this is class A, how do I know? Can I say, Oh it's cathode biased and it's drawing a lot of current at idle, so it's class A and move on? Doesn't seem like it. Looking at the RCA 6L6GC data sheet, it shows in a AF Power amplifier Class A plate dissipation 30 watts max. Is this what I should be looking at when I am concerned about my current draw with this?

RCA 6L6GC Datasheet | EZ260

[bob p]

To determine if the amp is operating in Class A is pretty easy. By definition Class A means that the conduction angle is 360-degrees. Maybe this chart will help:



It should be obvious then, that your amp might not be operating in Class A. If it is drawing a lot of current then it might be operating deep into class AB. Check the conduction angle to be sure.

[Dave H]

So, what could be making these tubes draw so hard? I tried others, and suspected perhaps leaky grid caps, but I see no voltage on pin 5. I'm at an impasse. Here is what I find on the outputs:

pin 2 to 7 = 6.3vac
pin 3 = 387 and 388 vdc
pin 4 = 386 and 387 vdc (the 1000R cement resistor shacks out)
pin 5 = 0.01 and 0.07 vdc
pin 8 = 27 vdc

That all seems to agree with the schematic, so what am I missing?

PS I am measuring with a bias socket, and have also done the cathode resistor drop, they match.

Current draw is not hard. It's exactly as on the 6L6GC data sheet for those voltages (360V Plate & Screen, -27V bias)

Tube dissipation (Plate + Screen) is (27/200)/2 * (387-27) = 24.3W or about 80% of 30W which is OK for a hot running Class AB amp which it almost certainly is and subtracting the Screen dissipation from the total puts the Plate dissipation lower at about 75%, even better.

[g1]

It should be obvious then, that your amp might not be operating in Class A. If it is drawing a lot of current then it might be operating deep into class AB. Check the conduction angle to be sure.

Sounds so simple. But how do you safely do that, preferably without scoping the power tube anode? And how do you then determine whether the result is the intended stock condition?