Hi,
When testing a driver stage for for example KT88, KT120 or KT150 in a Push Pull setup I would like to use “something” to test the driver without the final tubes in place.
What circuit (R - C) could simulate the final tubes in order to make proper measurements possible on the driver stage? I intend to do this with a real life setup (already available) and not using a software simulation.
I’m currently testing a driver and without any load everything looks find, but I know thing will be different with final tubes in place.
I’m looking forward to your suggestions.
Regards, Gerrit
When testing a driver stage for for example KT88, KT120 or KT150 in a Push Pull setup I would like to use “something” to test the driver without the final tubes in place.
What circuit (R - C) could simulate the final tubes in order to make proper measurements possible on the driver stage? I intend to do this with a real life setup (already available) and not using a software simulation.
I’m currently testing a driver and without any load everything looks find, but I know thing will be different with final tubes in place.
I’m looking forward to your suggestions.
Regards, Gerrit
Grid leak resistance in parallel with a small cap equal to the input capacitance of the output? For most intents and purposes, the resistor is enough I think.
Simulated Load capacitance to load the driver stage:
Beam Power tubes used in Beam power mode have the least capacitance at the control grid.
Beam Power tubes used in Ultra Liner mode have more capacitance at the control grid.
Beam Power tubes used in Triode mode have the most capacitance at the control grid, and it might be significant.
The control grid capacitance for Pentode Power tubes, in Pentode, Ultra Linear, and Triode mode works the same way as it does for Beam Power tubes.
Beam Power tubes used in Beam power mode have the least capacitance at the control grid.
Beam Power tubes used in Ultra Liner mode have more capacitance at the control grid.
Beam Power tubes used in Triode mode have the most capacitance at the control grid, and it might be significant.
The control grid capacitance for Pentode Power tubes, in Pentode, Ultra Linear, and Triode mode works the same way as it does for Beam Power tubes.
NickKUK,
Good point!
AB2 would require at least an additional series network of a steering diode, zener diode, and lower value resistor to ground; in parallel with the grid return resistors. plus and parallel caps to simulate the control grid capacitance if necessary.
(The lower value resistor has to simulate the control grid resistance when grid current is drawn).
Good point!
AB2 would require at least an additional series network of a steering diode, zener diode, and lower value resistor to ground; in parallel with the grid return resistors. plus and parallel caps to simulate the control grid capacitance if necessary.
(The lower value resistor has to simulate the control grid resistance when grid current is drawn).
Last edited:
For the KT150 the resistance in grid 1 circuit at fixed (clamp) bias is 51 K.
Interelectrode capacitance Input (nominal) is 20.5pF.
Parallelling these values as a "load" would suffice?
Regards, Gerrit
Interelectrode capacitance Input (nominal) is 20.5pF.
Parallelling these values as a "load" would suffice?
Regards, Gerrit
Some findings of today:
A balanced 6SN7 (push pull) stage driven to 50 VRMS without load drops to approx. 25 volts with this load.
With a MOSFET source follower after the 6SN7 the output voltage is not affected at all, it remains 50 VRMS. And even better: the THD is not affected by the MOSFET either.
Regards, Gerrit
A balanced 6SN7 (push pull) stage driven to 50 VRMS without load drops to approx. 25 volts with this load.
With a MOSFET source follower after the 6SN7 the output voltage is not affected at all, it remains 50 VRMS. And even better: the THD is not affected by the MOSFET either.
Regards, Gerrit
Not fully clear to me what you mean.
Do you mean that (without MOSFETs) the gain dropped 6 dB ?
Or available output voltage dropped from 50 to 25 V ?
Some sort of schematic diagram could make some sense.
Do you mean that (without MOSFETs) the gain dropped 6 dB ?
Or available output voltage dropped from 50 to 25 V ?
Some sort of schematic diagram could make some sense.
Hi Artosalo,
When loading the 6SN7 with the suggested load the signal drops 50% (voltage). Of course I load the 6SN7 AFTER the coupling capacitor. Doing the same with the MOSFET source follower leaves the output voltage intact (more or less as expected). I have no proper drawn diagram richt now unfortunately.
This all seems good to me.
Regards, Gerrit
When loading the 6SN7 with the suggested load the signal drops 50% (voltage). Of course I load the 6SN7 AFTER the coupling capacitor. Doing the same with the MOSFET source follower leaves the output voltage intact (more or less as expected). I have no proper drawn diagram richt now unfortunately.
This all seems good to me.
Regards, Gerrit
Hi Artisolo,
Thanks for your drawing. This is my diagram indeed, except that I have used a 1K resistor from each cathode to a 10M90S CCS in order to get as much lower THD. I get minimum THD with 6.3 mA (for both triodes together).
My measurements do not confirm your simulated output figures. I cannot explain why, but I know from previous experiments my findings are quite right. Whether its a 50% drop or slightly more or slightly less makes little difference.
I would like so know what others may have measured.
Regards, Gerrit
Thanks for your drawing. This is my diagram indeed, except that I have used a 1K resistor from each cathode to a 10M90S CCS in order to get as much lower THD. I get minimum THD with 6.3 mA (for both triodes together).
My measurements do not confirm your simulated output figures. I cannot explain why, but I know from previous experiments my findings are quite right. Whether its a 50% drop or slightly more or slightly less makes little difference.
I would like so know what others may have measured.
Regards, Gerrit
gerrittube,
Class A1 is usual.
But Class A2 (has grid current) is also possible.
Both of those are Class A.
It all depends on the driver swing, output tube bias techniques, quiescent operating voltage and current, active corrections, and active bias, etc.
Yes, Class A1 makes the driver's job easier versus Class A2.
And, self bias for the output tubes makes the driver's job easier.
That is because with self bias the control grid resistor to ground, Rg, can be a large resistance; versus fixed bias (or fixed adjustable bias).
A typical Beam Power tube type, may have a specification for Rg maximum of 220k or 500k for self bias; but Rg maximum specification of only 50k or 100k for fixed bias/fixed adjustable bias (for the same tube type).
Class A1 is usual.
But Class A2 (has grid current) is also possible.
Both of those are Class A.
It all depends on the driver swing, output tube bias techniques, quiescent operating voltage and current, active corrections, and active bias, etc.
Yes, Class A1 makes the driver's job easier versus Class A2.
And, self bias for the output tubes makes the driver's job easier.
That is because with self bias the control grid resistor to ground, Rg, can be a large resistance; versus fixed bias (or fixed adjustable bias).
A typical Beam Power tube type, may have a specification for Rg maximum of 220k or 500k for self bias; but Rg maximum specification of only 50k or 100k for fixed bias/fixed adjustable bias (for the same tube type).
Last edited:
The Valve Wizard -Long Tail Pair
Output impedance: The output impedance when both outputs are equally loaded is:
Zout (equal loading) = Ra || ra
If only one output is loaded, output impedance is approximately equal to:
Zout (one output loaded) = Ra / 2
Do you only loaded only one output?
Output impedance: The output impedance when both outputs are equally loaded is:
Zout (equal loading) = Ra || ra
If only one output is loaded, output impedance is approximately equal to:
Zout (one output loaded) = Ra / 2
Do you only loaded only one output?
kodabmx,
Suppose a push pull driver Does Use a Constant Current Source directly connected to the direct connected cathodes.
That causes a few things, I will mention two of them:
1. The 2nd harmonic of each of them is partially cancelled due to the linearizing feedback from one cathode to the other cathode.
2. That push pull driver stage is always Class A, all the way until it is clipping (until one tube completely cuts off).
Not too many Hi Fi amplifiers sound good when the driver clips, so in a practical sense, the stage is always Class A.
Just the way I see it . . .
Suppose a push pull driver Does Use a Constant Current Source directly connected to the direct connected cathodes.
That causes a few things, I will mention two of them:
1. The 2nd harmonic of each of them is partially cancelled due to the linearizing feedback from one cathode to the other cathode.
2. That push pull driver stage is always Class A, all the way until it is clipping (until one tube completely cuts off).
Not too many Hi Fi amplifiers sound good when the driver clips, so in a practical sense, the stage is always Class A.
Just the way I see it . . .