EL-34's won't bias.

[Amled87]

I've recently built a JCM 800, all of my plate voltages are correct. I've got -60v at bias resistors, 470vdc at the plates of my el-34's. The issue I'm having, is after I flip the stand by switch and the tubes start pulling current they don't stop pulling current. Bias starts at 0 and then goes up until the tube starts to redplate (around 300-400ma) and it keeps going til the tube self-destructs if I let it. However, I shut it off.

All of my bias components (caps, screen resistors, bias resistors, bias pot and etc read in spec and show no sign of malfunction. Adjusting the pot also makes no difference, as mentioned the bias starts and keeps going higher and higher to the point of redplating. Any one have any clue? Here's the schematic I used aswell.

Forgot to mention, this happens with any tubes. All of my voltages are 3-5% within range on the schematic.

 

[metalockpick]

I would a) reduce the value of R82 to ~68k-82k or so.
That bias supply doesn't have enough current at the plate voltages you are running
b) change the 47k resistor value to set your low voltage minimum, I'd try 44 Volts at the absolute lowest if b+ is ~475.
c) set your bias at the maximum voltage that allows a low volume clean tone with a natural sustain...too high and you get a choked sound - crossover distortion. Too low and you throw away power for gain, which gives lower volume overdrive at the cost of high power performance and tube life. For these tubes & conditions, I'd want to see about -53 Vdc.

 

[Amled87]

I would a) reduce the value of R82 to ~68k-82k or so.
That bias supply doesn't have enough current at the plate voltages you are running
b) change the 47k resistor value to set your low voltage minimum, I'd say 44 Volts at the lowest if b+ is ~475.
c) set your bias at the maximum voltage that allows a low volume clean tone with a natural sustain...too high and you get a choked sound - crossover distortion. Too low and you throw away power for gain, which gives lower volume overdrive at the cost of high power performance and tube life.

I think you've misunderstood. The bias fluctuates constantly and there is "no" setting it anywhere. I'm not sure how other 2204's operate if that bias supply isn't adequate as all of them have the same bias supply. I've currently got a 220k in place of the 120k. However, I'm not getting any negative voltage now at my bias resistors. Thank you for your help though! I'll try and lower values and see if I can get the bias to stabilize.

I assume you meant to change RB2 as there is no R82 on the schematic?

 

[metalockpick]

Yeah. Old eyes. And Marshall has many different versions of jcm800.
Stable circuits typically use lower resistance bias pots and resistors, though you can probably get yours to stabilize once the voltage drops, once you get the current problem sorted, as when those tubes start to go into bias runaway, the bias voltage is going to be dropping like a rock (by definition)

 

[Amled87]

Yeah. Old eyes. And Marshall has many different versions of jcm800.
Stable circuits typically use lower resistance bias pots and resistors, though you can probably get yours to stabilize once the voltage drops, once you get the current problem sorted, as when those tubes start to go into bias runaway, the bias voltage is going to be dropping like a rock (by definition)

One issue I've found (without the EL-34's installed is I have 230v roughly at the 220k resistor (installed instead of 120k) but after the diode I have only millivolts of voltage. That seems to be the issue as I'm supposed to have -56v at the bias resistors according to the schematic. I just changed the diode, and it's still millivolts.

So, first reduce the size of RB2. I've got a 82k handy (I'll try that and see if I'm getting -voltage at the bias resistors.

 

[Amled87]

Yeah. Old eyes. And Marshall has many different versions of jcm800.
Stable circuits typically use lower resistance bias pots and resistors, though you can probably get yours to stabilize once the voltage drops, once you get the current problem sorted, as when those tubes start to go into bias runaway, the bias voltage is going to be dropping like a rock (by definition)

Swapped the (on schematic it's 120k RB2) resistor out with a 82k and changed the diode out in the bias circuit. Still only getting negative millivolts at pin 5 of the power tube sockets. I'm using the red probe directly into pin 5 (from the top) and black to ground. Nowhere near -56v at the bias supply. Any other tips or ideas?

 

[metalockpick]

The circuit was working previously, the values before were not allowing enough current to go through.

I think your circuit is the bias circuit from the 50 watt version anyway, which has a lot lower voltages and half the output tubes. At least one of the JCM 100Wt versions had a separate transformer tap for the bias circuit.

This can work, you will likely have to reduce your resistor values such that the power draw on the system when the amp is on standby will be significant enough to get you relevant voltages when the amp is operating.

Diode should be a1N4007 or similar.

If the voltages are too high, one of the capacitors may have bought the farm, shorting out and giving you those very low values

 

[Amled87]

Yep... a 220k precision voltage to current device.

I think your circuit is the bias circuit from the 50 watt version anyway, which has a lot lower voltages than half the output tubes. At least one of the JCM versions had a separate transformer tap for the bias circuit. This can work, you will likely have to reduce your resistor values such that the power draw on the system when the amp is on standby will be significant enough to get you relevant voltages when the amp is operating

The transformer I'm using has (two 325v taps and two 6.3v taps) I'm using one 325v tap for the HT and 6.3 for the heaters. The second 325v tap I'm using for the bias supply using only one wire of that second tap is giving me 230v or so at the bias supply before the diode which should chnage it to negative voltage?


AS-3T325 - 300VA 325V Tube Transformer

Here is my power transformer, it's rated at .4A for both high voltage taps, and 4A for each 6.3v tap.

Diode is a 1N4007, I've changed the caps out for 22uF 450v just the be sure they never were underated at that point in th circuit as well.

 

[metalockpick]

Wow. Should be stellar when you get the bias sorted. Other than the values, and the 47pf capacitor (lose it, shield your preamp wires), everything here looks correct for what you are looking to do.

 

[Amled87]

Wow. Should be stellar when you get the bias sorted. Other than the values, and the 47pf capacitor (lose it, shield your preamp wires), everything here looks correct for what you are looking to do.

Preamp wires are shielded (I'll test with and without the 47pf ) however no matter what I have done so far I can not get the correct bias voltage at the resistors now.

Correct me if I'm wrong, but losing so much voltage after the diode (cathode is connected to the 82k resistor now and the anode is connected to the rest of the bias circuit) would indicate either an issue with the diode? Voltage goes from 230vac to only millivolts after the diode. The issue seems to be no negative bias voltage which cause the power amp tubes to runaway slowly approach max amplitude before they redplate and malfunction?

 

[metalockpick]

If you check after the diode and voltage is low:
1) could be the diode.

2) could be a shorted filter cap. Disconnect first cap and measure again.

3) check other caps.

That transformer seems very beefy. This amp likely won't sag voltage (good), though I can see how it is intolerant of that particular bias circuit

 

[metalockpick]

One more thing-
Those 22n phase inverter caps need to be 630VDC.
That's because at turn on, they get full B+ (470) on one side, and max bias on the other momentarily. All old Marshalls need this changed out, else you take a properly biased amp, turn it up and play a power chord and the tubes go into bias runaway from internal cap shorts in the poor quality, underspec mustard Mullard caps

 

[Tubetec]

With 30 plus years in the tube amp repair game I have to disagree ,
Mullard Mustard caps are quite reliable
Around 250 volts at the plate of the phase invertor , to minus 50 v at the grid of the power tube , the actual issue is the bias supply is taken from the cold side of the standby switch in some vintage Marshalls, so the power tubes have 0v at the grid the moment standby goes to on .

 

[Amled87]

One more thing-
Those 22n phase inverter caps need to be 630VDC.
That's because at turn on, they get full B+ (470) on one side, and max bias on the other momentarily. All old Marshalls need this changed out, else you take a properly biased amp, turn it up and play a power chord and the tubes go into bias runaway from internal cap shorts in the poor quality, underspec mustard Mullard caps

Disconnected caps from bias supply, still only millivolts at the 5k6 grid (pin 5) grid resistors. All of my voltage is 100% correct at every section of the amp (preamp, power amp and etc) except for in the bias supply. I can't even get -1v at the 220k resistors that feed the 5k6 grid resistors. I've changed the diode 3 times, also reduced the value of (RB2 from 220k down to 82k) no matter what I don't even get -1v after the diode.

All of my caps are Mallory 150's rated at 630vdc for that exact reason

 

[metalockpick]

With 30 plus years in the tube amp repair game I have to disagree ,
Mullard Mustard caps are quite reliable
Around 250 volts at the plate of the phase invertor , to minus 50 v at the grid of the power tube , the actual issue is the bias supply in taken from the cold side of the standby switch in vintage Marshalls

Nope. Not talking about switched vs unstitched side of standby. I'm talking about long term degradation of those capacitors from over voltage and poor sealing.

You won't see it unless you actually crank the amp up to top volume and let it blast. Turn on is fine, wait, hit power chord and failure ensues. Many, many, many techs miss this.

 

[Amled87]

Nope. Not talking about switched vs unstitched side of standby. I'm talking about long term degradation of those capacitors from over voltage and poor sealing.

You won't see it unless you actually crank the amp up to top volume and let it blast. Turn on is fine, wait, hit power chord and failure ensues. Many, many, many techs miss this.

Since I'm running my bias supply (power transformer isn't connected with a center tap) I believe I need to change my bias supply as the diode isn't being charged possibly? I could rewire it so I'm using a center-tap? Opinions?

 

[metalockpick]

Disconnected caps from bias supply, still only millivolts at the 5k6 grid (pin 5) grid resistors. All of my voltage is 100% correct at every section of the amp (preamp, power amp and etc) except for in the bias supply. I can't even get -1v at the 220k resistors that feed the 5k6 grid resistors. I've changed the diode 3 times, also reduced the value of (RB2 from 220k down to 82k) no matter what I don't even get -1v after the diode.

All of my caps are Mallory 150's rated at 630vdc for that exact reason

Best suggestion here would be to disconnect everything but the bias supply and see what you get. You were getting voltage previously, something went fishy on you. Try the RB2, the resistor and a filter cap (only) and you should get ~400+ vdc

 

[metalockpick]

Since I'm running my bias supply (power transformer isn't connected with a center tap) I believe I need to change my bias supply as the diode isn't being charged possibly? I could rewire it so I'm using a center-tap? Opinions?

Center tap should be grounded at the negative of the main filter cap

 

[Amled87]

Center tap should be grounded at the negative of the main filter cap

I didn't use a center tap as wiring the transformer with one requires both 325v taps. I used a bridge rectifier for my 470+VDC with one 325v tap and used the second tap to feed the bias supply. I could wire the transformer with a center tap however.

 

[Matador]

What do you get on pin5 with no power tubes installed?