Making Sense Of Bias Numbers

[mickeyd415]

Hello everyone nice to be here. I have a question for anyone that might be able to help. I have a '65 Fender Twin Reverb Re-Issue and I am trying to bias this for the first time. LEt me explain what I have done and the numbers I have for intepretation. First I took the plate voltage of the amp. Tube 1- 451.4 Tube 2- 451.4 Tube 3- 448.2 Tube 4- 448.6 I got this by taking my DMM on end to ground the other to pin #3 on each power tube. Next I took the DMM and set it to Dc Millivolts and placed one lead on the board on the B+ line (Red) and placed one lead on the tube that had each side of the output transformer on it. On the brown side I got 36.01 mv on the blue side I got 33.86 millivolts. Now with all of these numbers what is the next step to determine what each tube is pulling and what do I need to adjust the bias to (I am using 6L6 GC's) Sorry this is so long for my first post but I sure would appreciate some help. I also hope I gave enough information. :rolleyes:

[Old Tele man]

mickeyd415--to which pair of tubes (1 & 2 or 3 & 4) are the BLUE and BROWN OT-primary leads connected? It might explain the slight difference in plate voltages.

If you assume the tube-pairs are "matched" and conducting equally (very doubtful!), then the 36mA for the BROWN side implies each tube is conducting only 18mA...similarly, the 34mA for the BLUE side implies each tube is conducting only 17mA. However, odds are the tubes are NOT conducting equally, so it's possible one tube "could" be conducting 20mA and the other 16mA, etc.. Regardless, the tubes are idling at a VERY cold bias level...only about 8 watts.

Class-B operation would be 50%-of-Plate Maximum Dissipation (Ppd). So, for Svetlana SV-6L6GC's (Ppd=30W) the idle should be somewhere between 15 watts (Class-B @ 50%) and 21 watts (Class-AB @ 70%)...with YOUR ears being the deciding factor as to exactly were in between...!

The equation is:

Ip.q = % * Ppd / Vp.q

where:
Ip.q = Plate quiescent (idle) current, amps
Vp.q = Plate quiescent (idle) potential, volts
Ppd = Tube plate dissipation rating, watts
% = Percentage of plate dissipation


Remember to check / recheck the plate voltage everytime you change the bias, because it'll go DOWN as idle current goes UP, and vice-versa...

[mickeyd415]

Old Tele Man,
I just re-adjusted the bias now I am getting Power Tubes 1-4: 446.5, 446.1,447.3,448.1 and Blue Side (Tube 1&2) 73.4 milliamps, Brown Side (Tube 3&4) 72.4 milliamps but the bias is up as high as it can go. So if I take these numbers I believe that the new bias numbers are Blue Side: 32.75 (16.38 watts per tube) and Brown Side: 32.41 (16.21 watts per tube). Now having said all that don't the tubes need to be about 23 watts per tube, and if so what do I do to get it up to that range. Oh by the way I have Groove Tubes 6L6 B's, I plan on re-tubing but I want to know how to do this before I buy some good tubes no the cheapies that are in them. Any help form anyone is appreciated.:) :) :)

[Old Tele man]

23 watts might be too high.

As a guideline: LOW-end wattage would be 15 watts per tube (50% of 30 watts) and HIGH-end wattage would be 21 watts per tube (70% of 30 watts)...anything in-between is OK.

Since, you're getting about 16 watts per tube, which is between 15-21 watts, you're OK. If the amp sounds OK to you, then you don't have to do anything more.

But, if the amp sounds sterile or cold, then you need to increase the per tube idle wattage closer to 21 watts. This will require changing the value of one of the bias circuit resistors...something that you might want to consider having a qualified repair tech do if you're not familiar with soldering and reading schematic diagrams.

[John Phillips]

Originally posted by mickeyd415
I am getting Power Tubes 1-4: 446.5, 446.1,447.3,448.1
Firstly, this isn't possible, so you are getting some metering error.

The reason it's not possible is because tubes 1&2, and 3&4, have the plate connections paralleled with a fairly thick piece of wire. So the plate voltage on tube 1 must be exactly the same as that on tube 2, to a very high degree of accuracy. It doesn't matter what each tube is doing individually; the result is applied to both tubes equally.

You shouldn't try to deduce anything from variances which are within the meter accuracy - at best about +/- 0.5% with most normal multimeters. That's over 2V on a measurement of around 450V.

Reading to four significant figures will only confuse you. Even three sig figs (ie 1 part in 1000) is overkill; two (ie 1%) is all you should ever need for tube-amp work. In other words, you only need to know whether the voltage is 445 or 450, not whether it's 447 or 448. Likewise with current - there's really no important difference in .1mA, and not much in 1mA either - tubes are usually not matched to better than 5% anyway.

Like Old Tele man says, there's actually a fair range of 'correct' values for most amps, and as long as it's within that you'll be fine - after that it's only preference.


(But - Class B is not achieved at 50% of max dissipation. Class B is when each tube conducts for 50% of the cycle - which actually means that at idle it's cut off, ie no conduction. The problem is that this causes severe crossover distortion, so it isn't used for audio amps. That's the reason Class AB came about - by running some current at idle, the COD can be eliminated while still giving most of the power advantage of Class B. In some amps, you can run well under 50% of max. dissipation and not get COD. I agree that generally they don't sound quite right until they're up in the 50% region though.)

[Old Tele man]

re: "...each tube conducts for 50% of the cycle..." Correct, but at IDLE, there is NO signal to take 50% of...

...so, the engineering book guidance is that Class-B idle current is set at 50% of total plate current; thus, if the maximum plate dissipation is 30W, then Class-B idle approximately at ~15W -- (note the "weasel-wording" use of approximately "~").

...the more precise method is to set the Class-B idle at the control-grid projected cutoff point...but this requires somewhat more (but not necessarily difficult) information and calculation, ie:

Vg.q = (1.35)*(Ip.q / gm) - ( Vs / mu )

where:

Vg.q = Control-grid quiescent (idle) bias, volts(negative)
Ip.q = Plate quiescent (idle) current, amps
Vs = Screen potential, volts
mu = Tube dynamic screen (triode) amplification factor, mu
gm = Tube dynamic transconductance, amps-per-volt

This equation yields a reasonably-accurate bias voltage (Vg.q) value needed to produce the desired quiescent (idle) plate current (Ip.q)...for example, a 6L6GC with screen voltage of 400Vdc would idle at about 58mA with a control-grid bias voltage of -37Vdc:

Vg.q = (1.35)*(0.058A / 0.0060A/V) - (400V/8)
Vg.q = (1.35)*(9.667V) - (50V)
Vg.q = (13.05V - 50V) = -36.95V, or approx: -37Vdc

...but, to use this equation you gotta have some idea of the tubes' gm-value!

[spentron]

Hey, the first post says 36 mv and the second swaps that to 36 mA .. unless the transformer resistance is 1 ohm, that's incorrect.

[mickeyd415]

spentron,
that is my mistake I meant milliamps not millivolts. Sorry. BTW, I just finished with the bias and all I have to say is WOW!! I now have some life out out the amp. It made a large difference, now on to some good tubes, thanks to all who helped.:dude