• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Automatic bias board.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm sure it works fine - the real challenge with a bias servo however is how it behaves under dynamic conditions. Without seeing a schematic it is hard to predict how this will be. Additionally, if the amplifier has a global feedback loop, any servo (not necessarily the one shown) could potentially cause instability or at least LF ringing if its own time constant interferes (ie gets too near) with those in the amplifier itself.
 
Last edited:
That's a dead ringer for Guido Tent's board which is based on the van der Veen circuit. It has an extremely long time constant and will not cause LF instability in a properly-designed feedback amplifier. It performs well with very little drift in bias even with very loud music signals with long, sustained bass notes.

ETA: I have one of Tent's boards in the amp in my bedroom.
 
Apparently (from various descriptions) Pavel's board is using the 0 to 2X idle sensed cathode current control method (like Morgan Jones describes in his book), while the Guido-Tent board is using a narrow sensed current range around zero AC input signal (not sure HOW, since it doesn't have a grid input AC signal to monitor?). The PC board layouts are NOT the same anyway, if you look carefully.

Description of operation (method 2, I suppose)
Optimization of operating conditions for push-pull output stages of ...

Some discussion here:
Bias servo for class AB

However I see an issue in the -description- (method 2) on that audioamp.eu link, and also with the general 0 to 2X scheme description.

The class AB tube is operating at low to zero current range in the sampling region, so its gm is varying significantly. This means that Symmetric Vin operation around the zero Vac input region has -Asymmetric- current variation around the idle current that is to be measured. This would mean sampling should be from 0 to something LESS than 2X current to make up for the gm variation (to avoid "pulling" the integrated sense value upward).

While I see the audioamp.eu diagram has got it sampling from from 0 to 25 to 60mA, the Opposite correction! This may just be description semantics however. Just means the idle condition (Vac input = 0) is more like at 35 mA (not 25). One hopefully tests the idle current with an actual meter, using 0 Vac input signal (while adjusting the bias set pot on the circuit board).

My 2 cents...

.
 
Last edited:
Correction to the last line above.

Setting the idle current with 0 Vac input will actually see the dynamic operation occur at something Less than that set current, because of the gm variation pushing the average off center. (causing the sensed "2X" current to be larger than expected for a truly linear tube. So the circuit will reduce the bias current to compensate, by dropping the grid V slightly more negative) And that now dynamic idle current will vary some with signal amplitude up until the "2X" current limit is passed. Should still be OK if the integrator is slow enough.
 
Last edited:
The capture of the cathode current is the same as Guido Tent. Measurement section for each tube measures the actual current around a very small window around the bias hence is not affected by any music signal (Guido Tent). The wiring diagram was on the Internet. When setting, the cathode current of one tube is measured. The current of the other three tubes is set automatically.
The setting is to measure the cathode current of one tube. The current of the other three pipes is set automatically. The signal on the grid tube has no effect on the setting. An audio signal may or may not be. However, it is good to check the signal at the output of the amplifier, such as an oscilloscope. However, the module is correctly set even without the oscilloscope.
Pavel audioamp.eu
apologize google translator :)
 
Some link to the schematic?

From what I see, it is measuring cathode current across a 10 Ohm resistor, limited by 0.6V Si diodes. So the measurement interval is 0 to 60 mA. There is no actual grid signal input to determine the AC drive zero crossing. So it must be operating off the cathode current alone.

Tubes have a variation of gm with current, at 0 current they will have 0 gm. Instead of the rectangular integration expected from a linear amplifying device (with constant gm), you get a triangular integration from the gm non-linearity. The integrator output will vary the bias V until the triangle has equal area above and below the (new dynamic) DC set point. So it controls the bias V to get the average DC cathode current (with AC signal applied within the limited range) at about 25 mA.

So setting the unit up for 30 mA DC idle with no AC signal applied, will see a drop to 25 mA when AC signal is added. A long integration time will keep things more stable by averaging over the signal variation effects.

right ???

To do this more correctly, as Wavebourn mentioned in the DIYaudio link above, one would need a sample and hold function and the AC coupled grid drive signal input to determine the AC zero crossings to trigger the sample and hold.

.
 
Last edited:
Good to hear. I didn't spend a lot of time comparing the details (obviously) but the layout did look very extremely similar at first glance.

My God! You put a lot of efforts! I can only say it's very appreciated. It crossed my mind to try such device in an amp eventually. But buying a "copycat board" wouldn't be my first choice. In this case, I think Guido's design kept attention from a "fan" ;)
 
AB-Q

Some link to the schematic? <snip>
yes, you have described it exactly.
The image is from the output of the GRD module.

66d14533776e57a9a658de7db532321f.640x360.jpg
 
I don't see why it wouldn't work for 4 tubes, one channel. I assume you could also use this to bias 4 mono SE amps.

The dimensions are 100x57mm from audioamp.eu.

I've seen others talk about a 0 - 60 ma current window. Does that mean I can't use this module to set an idle current of 90ma? Or would I have to change the 10R sense resistors to 6 2/3R?
Thanks
 
I believe you can adjust the reference pot. on the board (center) to change the idle current. It will affect all four channels. The other way would be to put some shunt R across the 10 Ohm sense R (parallel summing). Could just add a shunt R from cathode to ground at the tube, I think. The latter might be better for high current TV sweep tubes, just to avoid smoking the 10 Ohm.

---------------------

Pavel:
Looking at your posted waveform, I had imagined that the audio signal would produce some 2nd harmonic of itself in the correction waveform, from the Asymmetric tube gm around the current set point. But I don't see any obvious effect in the waveform. Is this a non effect? (and so no bias current drift between no AC or quiet and AC audio present?) In which case the scheme must be quite stable at holding the idle current. ( I was conjuring up compensation schemes for the gm variation..., like altering one of the limit diodes slightly by mis-match of Vf or adding a series Schottky diode)

.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.