• 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.

Single stage line with anode-cathode feedback

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
A pointless circuit is one where additional complication is added for no particular advantage or motivation. If revintage builds this circuit, compares it to a circuit without the complication (or to just the opamp used alone), and shows a demonstrable performance advantage, that's a different story.
 
As revintage has made no claim whatsoever regarding the performance of the circuit, rather he simply enquired "Has anyone tried the leftmost topology before?", it seems odd to suggest that he should provide evidence in support of it showing a "demonstrable performance advantage".

The circuit could hardly be described as complex.

"Just use the op-amp" conveniently ignores the possibility that the OP might actually want to use a valve for whatever reason, or assumes that any reason he might have is specious.

It does seem a shame that a quite reasonable open-ended enquiry should be treated so disdainfully, and certainly not conducive to sensible discussion. I had assumed that would be regarded as worthwhile, perhaps I am in error.

Surely it is possible to state that one cannot see any advantage in doing something one way, as opposed to another, without using pejorative terms like "pointless"? Or am I in error again?
 
As revintage has made no claim whatsoever regarding the performance of the circuit, rather he simply enquired "Has anyone tried the leftmost topology before?", it seems odd to suggest that he should provide evidence in support of it showing a "demonstrable performance advantage".

The circuit could hardly be described as complex.

"Just use the op-amp" conveniently ignores the possibility that the OP might actually want to use a valve for whatever reason, or assumes that any reason he might have is specious.

It does seem a shame that a quite reasonable open-ended enquiry should be treated so disdainfully, and certainly not conducive to sensible discussion. I had assumed that would be regarded as worthwhile, perhaps I am in error.

Surely it is possible to state that one cannot see any advantage in doing something one way, as opposed to another, without using pejorative terms like "pointless"? Or am I in error again?
I whole heartily concur.
 
SY,
Yet another pointless comment from you. If you can make a feedbacked single stage tube circuit that is not an anodefollower(low inputZ and bandwith)(to the left in the first post) I will donate 1$ to "Save the Children" in your name. You have to get around the phase reversal in some way.
 
SY,
Yet another pointless comment from you. If you can make a feedbacked single stage tube circuit that is not an anodefollower(low inputZ and bandwith)(to the left in the first post) I will donate 1$ to "Save the Children" in your name. You have to get around the phase reversal in some way.

There's a few different ways it can be done.

You have two active devices; as soon as you have the opamp in the feedback loop, it's an integral part of the network which determines the transfer function, so the tube is only reducing the performance. So you could drop the tube and configure the opamp as a 12dB gain inverter. That's probably the highest performance solution.

You could sub a tube section for the opamp to keep the same number of active devices- the topology I suggested earlier would work very well and probably outperform tube plus opamp.

You could configure the tube as a cathode follower and use a step-up transformer at the input.

You could use the tube as an anode follower with a large input resistor (100k).

You could run the tube open loop with an output transformer load.

There's probably a few dozen other ways to do the same thing with your constraint of not increasing the number of active stages.
 
You could use the tube as an anode follower with a large input resistor (100k).

Terminological error on my part- what I meant was plate to grid feedback, with a 100k input resistor. It's pretty much equivalent to the circuit you drew (gain, distortion, output impedance), but without the opamp and supporting circuitry.

I appreciate the donation in my honor.
 
SY,
Yet another pointless comment from you. If you can make a feedbacked single stage tube circuit that is not an anodefollower(low inputZ and bandwith)(to the left in the first post) I will donate 1$ to "Save the Children" in your name. You have to get around the phase reversal in some way.
SY Sorry for my lack of reading comprehension in this but how is an anodefollower not an anodefollower and thus qualify for the donation ? Charles Boegli did some very good articles about the anodefollower in the 1950s . When used with a Baxsandall tone controls it work very well.
 
As I explained, I meant a simple inverting triode stage with plate to grid feedback resistor. It's an anode follower, strictly speaking, only if it's unity gain.

100k input resistor, 400k plate to grid resistor. I'd CCS load the plate and use an LED in the cathode, but there's a lot of other options to get the open loop gain up to near the mu of the tube.

So... we have the same output impedance, same distortion, high input impedance, but no opamp and supplies.
 
100k input resistor, 400k plate to grid resistor. I'd CCS load the plate and use an LED in the cathode, but there's a lot of other options to get the open loop gain up to near the mu of the tube.

SY, making anode to grid feedback, the input impedance of the stage is severely reduced because of voltage-parallel feedback type. And for make this, it is recommended to use pentodes because its higher open loop gain.

Cordially, Osvaldo.
 
As I explained, I meant a simple inverting triode stage with plate to grid feedback resistor. It's an anode follower, strictly speaking, only if it's unity gain.

100k input resistor, 400k plate to grid resistor. I'd CCS load the plate and use an LED in the cathode, but there's a lot of other options to get the open loop gain up to near the mu of the tube.

So... we have the same output impedance, same distortion, high input impedance, but no opamp and supplies.
If you refer to the Boegli articles he showed that the anode follower was useful because you could have other than unite gain with the circuit. The anode follower with unite gain is a fish out of sea water so to speak. The cathode follower thus gain favor in use in tube circuits. The anode follower is helpful if it is used as other than unite gain . As Charles' article showed .
 
Hmh, let me briefly go back to the beginning of the thread since it seems no one mentioned one possibly obvious view of revintage's circuits...

You do realise that the whole point of negative feedback is that a resistor has much less distortion than an active device? Putting active devices in the feedback chain seems perverse.

Actually, how about using the same type active device to provide an equal-but-oposite nonlinear function by putting it into the feedback loop, thereby implementing an error cancellation scheme? There is prior art (supertriode)...
 
No, the input impedance is 100k if that's the grid input resistor. The grid becomes a virtual ground.

Only if the gain of the triode was 'infinite' :) as it is, it's closer to 20 or so without NFB - so the input impedance is higher than 100k. Also, with a 400k feedback resistor the gain will be lower than 4 for the same reason. Still, this does not change appreciably the principle that is being shown.
 
Actually, how about using the same type active device to provide an equal-but-oposite nonlinear function by putting it into the feedback loop, thereby implementing an error cancellation scheme? There is prior art (supertriode)...

That was actually commercialized some years back (maybe the 1980s?). Can't remember the name of the company- they did things like put multiple solder joints, opamp buffers, and phono cartridges in feedback loops.

Open loop gain with a CCS load and a 400k feedback resistor (as load, not returned to grid) will be quite close to mu (35), so the input impedance won't be much higher than 100k- but as you point out, it WILL be a bit higher and gain will be a bit lower than 4.
 
Hmh, let me briefly go back to the beginning of the thread since it seems no one mentioned one possibly obvious view of revintage's circuits...

Really? My joke on the first page covered this aspect. :)

Actually, how about using the same type active device to provide an equal-but-oposite nonlinear function by putting it into the feedback loop, thereby implementing an error cancellation scheme? There is prior art (supertriode)...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.