But how do the cathode and plate resistances (of each 130K) fit in this? Are they not in parallel with the 1M load for ac-signal?
Rayma
Please ignore my first and later diagrams except for the last one, in Post #14,
I have omitted the 1M resistor on the grid of V2b.
If you see anything wrong with the new diagram, I wish to know what you think and how to fix it.
Thanks.
Please ignore my first and later diagrams except for the last one, in Post #14,
I have omitted the 1M resistor on the grid of V2b.
If you see anything wrong with the new diagram, I wish to know what you think and how to fix it.
Thanks.
Last edited:
No, the 1M grid resistor is not connected to the second stage cathode or plate. A grid terminal is (to first order) an open circuit.
Do you mean the first section 220k and 3.16k resistors? The composite output resistance of the first stage, which includes
these resistors as well as the tube parameters, is in series with the 1M of that circuit.
The basic topology, ignoring component values (as revised in post #14), is now similar to various known tube amp designs.
At this stage one would use the tube curves from a data manual to determine possible DC bias points, which are
more critical when the second stage is directly coupled to the first stage. You would first need to know the maximum
output tube grid signal needed for full output. This depends on the OT, power tubes, rated output power, etc.
Perhaps someone here would like to undertake that effort.
OK, thanks.
But see my post #9 in which I point out that the 1M grid resistor in the schematic in post #1 should not be connected to ground but to the bottom of an additional cathode resistor. I would think that in that situation the 2 x 130K will be part of the ac-resistance the preceding stage will see.
Yes, that would be a different circuit, which was not shown in this thread. To derive the circuit equations for such a driver stage,
you would draw the circuit on a piece of paper, and then derive the circuit equations using algebra and Kirchhoff's laws.
Old school tube guys would say that the 1M grid resistor is bootstrapped by the cathode signal, and is effectively large enough to ignore, especially with the 20% parts of that era. But if you want the exact equations, you'd have to draw out the circuit and derive them. Of course you can try SPICE and simulate the circuit, but that does not give you the theoretically correct component values.
you would draw the circuit on a piece of paper, and then derive the circuit equations using algebra and Kirchhoff's laws.
Old school tube guys would say that the 1M grid resistor is bootstrapped by the cathode signal, and is effectively large enough to ignore, especially with the 20% parts of that era. But if you want the exact equations, you'd have to draw out the circuit and derive them. Of course you can try SPICE and simulate the circuit, but that does not give you the theoretically correct component values.
To Robert H. Gribnau
You are correct. The original circuit of Post #1 would work better if the 1M resistor went from the grid of V2b to a node between two cathode resistors, the upper one larger than the lower one. But I see now that it was unnecessary to begin with, and I don't even recall where it came from. I think perhaps that the drawing was a work in progress, had errors, but I was distracted for a time and later posted it by mistake. I have since posted an updated diagram, Post #14, in which that resistor has been omitted. The new diagram should be getting closer to viability.
Thanks.
You are correct. The original circuit of Post #1 would work better if the 1M resistor went from the grid of V2b to a node between two cathode resistors, the upper one larger than the lower one. But I see now that it was unnecessary to begin with, and I don't even recall where it came from. I think perhaps that the drawing was a work in progress, had errors, but I was distracted for a time and later posted it by mistake. I have since posted an updated diagram, Post #14, in which that resistor has been omitted. The new diagram should be getting closer to viability.
Thanks.
Yes, this is now a Dynaco style driver circuit, with the voltage amplifier directly coupled to the Concertina phase inverter.
This driver circuit (with a 7199 pentode/triode tube) was in all their commercial tube amplifiers (the Stereo 35 used a different tube),
and even in the Dynaco MKVI mono 130W tube amplifier, which used four 8417 output tubes in each mono chassis.
This driver circuit (with a 7199 pentode/triode tube) was in all their commercial tube amplifiers (the Stereo 35 used a different tube),
and even in the Dynaco MKVI mono 130W tube amplifier, which used four 8417 output tubes in each mono chassis.
Rikaro,
I don't know what you mean by "DC coupling". The signal from V2a to V2b is biased positive through the 22pF capacitor and 22k resistor, according to the original RCA circuit (including part values). Is that what you mean by "DC coupling"?
I don't know what you mean by "DC coupling". The signal from V2a to V2b is biased positive through the 22pF capacitor and 22k resistor, according to the original RCA circuit (including part values). Is that what you mean by "DC coupling"?
Actually that is a high frequency stability correction circuit, which reduces ringing, and has nothing to do with DC biasing
or DC coupling of the tubes.
or DC coupling of the tubes.
DC coupling (as opposed to AC coupling) here means that the PI grid is directly connected to the plate of the preceding gain stage.
Consequently the PI grid DC voltage is defined by the plate voltage of the preceding stage.
As the PI cathode voltage follows its grid voltage within 1...2V, also the cathode voltage is fixed.
The technical phrase "direct coupling" means that the first stage is connected directly (hence the name) to the second stage,
without blocking capacitors, transformers, etc. between the stages. This can be done with either tubes or transistors,
but requires additional design effort. This approach can eliminate one low frequency pole from the transfer function,
enhancing the low frequency stability of the circuit.
without blocking capacitors, transformers, etc. between the stages. This can be done with either tubes or transistors,
but requires additional design effort. This approach can eliminate one low frequency pole from the transfer function,
enhancing the low frequency stability of the circuit.