Runaway bias with HRD

[norris2002]

some weeks ago I started a thread detailing problems with a hot rod deville. turned out the 82k plate resistor on the PI was open. I replaced this, and the amp worked, briefly.

Now there is low volume and all notes are distorted. I have taken the amp to work to get one of the engineers here to have a look (telecom system design, experience with building valve amps in the 70s).

We have located the problem to the power tubes and surrounding grid resistors / bias / etc. The main problem seems to be with the bias. The lowest I can adjust the bias is around 75mA, and if i try to adjust lower, the mA starts to rise, quite rapidly. I thought this was due to heat affecting a resistor somewhere. a quick experiment with the multimeter after powering down the amp showed that one of the 220k resistors (R59 or 60) that go from grid to C- supply was changing value quite a lot as the device cooled down. I replaced both 220k resistors, and the problem is still there.

Why would the bias supply be so unstable. Is this something that could be tube related (I can get a spare set of 6l6's from somewhere I am sure) or is it more likely to be some other component. The two coupling caps to the power tubes are blocking DC as they should, and when I adjust the bias trim pot, the voltage is changing as it should, but current draw is not. Are there any 'usual suspects' to check in the case of bias issues?

[Structo]

Have you tried different power tubes?

[Alexo]

How 'bout the screen resistors?

[Bob-I]

Is the bias voltage stable? You may have a bad C- cap or resistor in the ckt. Make sure you measure the C- voltage at the tube socket itself to make sure it's not drifting there.

[PCollen]

some weeks ago I started a thread detailing problems with a hot rod deville. turned out the 82k plate resistor on the PI was open. I replaced this, and the amp worked, briefly.

Now there is low volume and all notes are distorted. I have taken the amp to work to get one of the engineers here to have a look (telecom system design, experience with building valve amps in the 70s).

We have located the problem to the power tubes and surrounding grid resistors / bias / etc. The main problem seems to be with the bias. The lowest I can adjust the bias is around 75mA, and if i try to adjust lower, the mA starts to rise, quite rapidly. I thought this was due to heat affecting a resistor somewhere. a quick experiment with the multimeter after powering down the amp showed that one of the 220k resistors (R59 or 60) that go from grid to C- supply was changing value quite a lot as the device cooled down. I replaced both 220k resistors, and the problem is still there.

Why would the bias supply be so unstable. Is this something that could be tube related (I can get a spare set of 6l6's from somewhere I am sure) or is it more likely to be some other component. The two coupling caps to the power tubes are blocking DC as they should, and when I adjust the bias trim pot, the voltage is changing as it should, but current draw is not. Are there any 'usual suspects' to check in the case of bias issues?

Put the amp in STBY, and then measure the bias VOLTAGE min and max values as you adjust the bias pot from min to max at a) the junction of the two 220K bias resistors and then also at b) pin 5 of each power tube making sure that the voltage is the same at both pin 5's. then try and set the bias at the reference voltage of approx -54.5 Vdc. Don't worry about the tubes, as in STBY no plate voltage is applied to them. If the C bias voltage tracks OK, and can be set at -54.5 Vdc, then set the bias voltage at MAX NEGATIVE value ensuring that both power tube pin 5's are the same and that the value is MORE NEGATIVE than -54.5 ( should be significantly MORE NEGATIVE) , take it out of STBY, and measure your DC power supply voltages against the expected values noted on this schematic as well as verifying the value of your C bias voltage again:

http://support.fender.com/schematics/gu ... ematic.pdf

If your tube plate voltages (pin 3) and screen grid voltages (pin 4) are OK (around 480 Vdc) and your bias voltage is set to -54.5 Vdc at pin 5 of both tubes, then I would suspect a bad tube OR you may have one or both tubes that simply draw more current than expected. Your target cathode current is 60mA (per TP30 ) calculated by the 60mV referenced drop across R66. This is the SUM of both tubes cathode current at quiescence when being biased with no signal applied to the amp (and I like to turn ALL controls to zero as well. But you should be able to adjust this downward using the bias control to set the C voltage to a more negative value than -54.5 Vdc

[norris2002]

PCollen, I have got some numbers from your suggestions, and all the voltages seem a bit high, but not massively. I am in the UK, and this amp is set to 240v (not 230 V like they were originally shipped to the UK).

When in standby, the bias at the 220k junction will adjust from -51.3 to -62.7 and reading at pin 5 of the valve from -50.1 to -61.3. Both valves read the same.

From now on C- readings are taken from the valve pin 5. I set the C- to -61.3 and turned the amp on. Other voltages are as follows:

B+ 514
Z 511
Y 467
X 409
C- -60.3

V4p3 511
V4p4 510

V5p3 511
V5p4 510

bias (read from 1R cath resistor) 26.9mA


I then set the C- to read -54.5 (on pin 5 or V5 whilst on (not standby)).

B+ 508
Z 506
Y 462
X 406
C- -54.5

V4p3 507
V4p4 506

V5p3 507
V5p4 506

bias (read from 1R cath resistor) 52.9mA


Does the extra 20V everywhere stop an amp working? I think not. I will pick up a set of 6L6's this weekend and test with them next Monday. Hopefully that is the only problem, but I am hesitant to put my hopes on this.

[PCollen]

PCollen, I have got some numbers from your suggestions, and all the voltages seem a bit high, but not massively. I am in the UK, and this amp is set to 240v (not 230 V like they were originally shipped to the UK).

When in standby, the bias at the 220k junction will adjust from -51.3 to -62.7 and reading at pin 5 of the valve from -50.1 to -61.3. Both valves read the same.

From now on C- readings are taken from the valve pin 5. I set the C- to -61.3 and turned the amp on. Other voltages are as follows:

B+ 514
Z 511
Y 467
X 409
C- -60.3

V4p3 511
V4p4 510

V5p3 511
V5p4 510

bias (read from 1R cath resistor) 26.9mA


I then set the C- to read -54.5 (on pin 5 or V5 whilst on (not standby)).

B+ 508
Z 506
Y 462
X 406
C- -54.5

V4p3 507
V4p4 506

V5p3 507
V5p4 506

bias (read from 1R cath resistor) 52.9mA


Does the extra 20V everywhere stop an amp working? I think not. I will pick up a set of 6L6's this weekend and test with them next Monday. Hopefully that is the only problem, but I am hesitant to put my hopes on this.

Your tubes are biased very cold @ 13 W..you need to adjust your -C bias voltage to be LESS negative, try about -50 Vdc, to increase your voltage drop across R66 to about 80mV (indicating and average 40mA per tube quiescent cathode current flowing. Expect your DC voltages to drop further probably to around 495Vdc on the power tube plates (P3) as you adjust the bias voltage less negative. This would give you about 20W bias on the power tubes, which is a typical bias poing (66%) for the 30W 6L6GC's. The stock DeVille tubes are biased somewhat cold @ if Vp = 485Vdc and Ip = 30mA....that's a little less than 15W and doesn't take into account a loss of a few mA per tube to screen grid current which would further lower the bias point.

NOTE: Make sure the 6L6's you are using ARE, in fact, 30W-rated tubes and not 25W-rated, as some may be. If 25W rated, then the 60mV reading across R66 would be about right if Vp were 500Vdc. Read this:

http://thetubestore.com/6l6templeton.html

[PCollen]

See my embedded comments >>>>>>

PCollen, I have got some numbers from your suggestions, and all the voltages seem a bit high, but not massively. I am in the UK, and this amp is set to 240v (not 230 V like they were originally shipped to the UK).

>>>>>if the amp expects 240Vdc, and the wall voltage is LESS, then your DC voltages in the amp will be slightly LOWER than expected. If wall voltage is MORE, then your DC voltages will be slightly HIGHER than expected

When in standby, the bias at the 220k junction will adjust from -51.3 to -62.7 and reading at pin 5 of the valve from -50.1 to -61.3. Both valves read the same.

>>>> Your higher than expected voltages is also effecting the RANGE of your bias voltage. But the low end of -50.1 may be where you need to be under the circumstances.

From now on C- readings are taken from the valve pin 5. I set the C- to -61.3 and turned the amp on. Other voltages are as follows:

B+ 514
Z 511
Y 467
X 409
C- -60.3

V4p3 511
V4p4 510

V5p3 511
V5p4 510

bias (read from 1R cath resistor) 26.9mA

>>>> OK...as expected. You are throttling back the quiescent cathode current flowing in the pwer tubes


I then set the C- to read -54.5 (on pin 5 or V5 whilst on (not standby)).

B+ 508
Z 506
Y 462
X 406
C- -54.5

V4p3 507
V4p4 506

V5p3 507
V5p4 506

bias (read from 1R cath resistor) 52.9mA

>>>>>OK..as expected. Less negative bias voltage allows more current to flow in tubes, and the DC voltages will drop under the increased current loading.


Does the extra 20V everywhere stop an amp working? I think not. I will pick up a set of 6L6's this weekend and test with them next Monday. Hopefully that is the only problem, but I am hesitant to put my hopes on this.

>>>>> Shouldn't make any difference. Amp voltage tolerance is typically +/- 10-20% unless you are on the threshold of the voltage ratings of other components (cap's, etc.)

[groovtubin]

some weeks ago I started a thread detailing problems with a hot rod deville. turned out the 82k plate resistor on the PI was open. I replaced this, and the amp worked, briefly.

Now there is low volume and all notes are distorted. I have taken the amp to work to get one of the engineers here to have a look (telecom system design, experience with building valve amps in the 70s).

We have located the problem to the power tubes and surrounding grid resistors / bias / etc. The main problem seems to be with the bias. The lowest I can adjust the bias is around 75mA, and if i try to adjust lower, the mA starts to rise, quite rapidly. I thought this was due to heat affecting a resistor somewhere. a quick experiment with the multimeter after powering down the amp showed that one of the 220k resistors (R59 or 60) that go from grid to C- supply was changing value quite a lot as the device cooled down. I replaced both 220k resistors, and the problem is still there.

Why would the bias supply be so unstable. Is this something that could be tube related (I can get a spare set of 6l6's from somewhere I am sure) or is it more likely to be some other component. The two coupling caps to the power tubes are blocking DC as they should, and when I adjust the bias trim pot, the voltage is changing as it should, but current draw is not. Are there any 'usual suspects' to check in the case of bias issues?

Been modding these for 11 years, the BIAS pot is prob the cause, ( intermittent and stupid!) i`ve seen it alot, if not the bias cap is leaking, or you have a bad OT!! I HAVE seen that!

jp@Omegaamps