• 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.

The best amplifier I have built so far

Jump to Latest
#41
I was thinking one could reduce the driver stages' gain by using local NFB from the output tube plates to the diff-driver cathodes, and perhaps plate-grid feedback around the first stage, instead of a global NFB loop. Keep the more complex reactances of the OPT out of the feedback loop. That should increase stability.
Click to expand...

Well, I investigated that and it's a lot harder than it at first appears. I think the diff-driver would have to have fairly high gain to drive that local feedback loop, which means of course there'd be more open loop gain, and now you're right back where you started.

So... Never mind. 😱
 
#42
rongon said:
There is the 'simpler is better' school.
--
Click to expand...

Yes, that's where it's at. "Simpler" can be defined in different ways. I think one ways is to use less gain stages and more followers with current gain. This would allow one to use less tubes and more silicon and making a simpler, cheaper, cooler (in temperature), and more compact amp.

Most everything that can be done has been done. One thing that I think hasn't been done is to combine a large amplification tube like a 12ax7 or similar voltage gain stage that is capacitively coupled to a high voltage mosfet concertina. Then have mosfet followers driven by the concertina. That would keep the mosfet concertina from becoming unbalanced on high drive signals. And though you would lost the traditional direct coupling between gain stage and concertina you would then have direct coupling "after" the concertina to compensate.

Finally, couple the mosfet followers directly to the output tubes using Tubelab's power drive scheme. It seems like it would work. You would have to have a separate high voltage supply going to the concertina because the theoretical maximum signal voltage would be determined by the voltage applied to that stage. The theoretical peak signal voltage out could only be a bit less than half of that voltage. That's how concertinas work.

It would look complicated in its final form but most of the complication would come in the form of silicon, which is cheap. And the silicon wouldn't detract from the sound because it wouldn't be used for voltage gain. And you would have perfectly symmetrical push pull drive because it would never become unbalanced because no current would flow after the concertina through the gates of the mosfet followers that it feeds.
 
Last edited:
#44
exeric said:
Yes, that's where it's at. "Simpler" can be defined in different ways. I think one ways is to use less gain stages and more followers with current gain. This would allow one to use less tubes and more silicon and making a simpler, cheaper, cooler (in temperature), and more compact amp.

Most everything that can be done has been done. One thing that I think hasn't been done is to combine a large amplification tube like a 12ax7 or similar voltage gain stage that is capacitively coupled to a high voltage mosfet concertina. Then have mosfet followers driven by the concertina. That would keep the mosfet concertina from becoming unbalanced on high drive signals. And though you would lost the traditional direct coupling between gain stage and concertina you would then have direct coupling "after" the concertina to compensate.

Finally, couple the mosfet followers directly to the output tubes using Tubelab's power drive scheme. It seems like it would work. You would have to have a separate high voltage supply going to the concertina because the theoretical maximum signal voltage would be determined by the voltage applied to that stage. The theoretical peak signal voltage out could only be a bit less than half of that voltage. That's how concertinas work.

It would look complicated in its final form but most of the complication would come in the form of silicon, which is cheap. And the silicon wouldn't detract from the sound because it wouldn't be used for voltage gain. And you would have perfectly symmetrical push pull drive because it would never become unbalanced because no current would flow after the concertina through the gates of the mosfet followers that it feeds.
Click to expand...

One big added benefit is that since you would operate in A2 bias for the final stage you could probably get mucho power depending on the output tubes. You wouldn't have to go with parallel output tubes to get it either.
 
#45
FWIW...

LTspice thinks the Allen Wright cascode LTP looks really (really really) good.

So there ya go!

However, I don't see it giving great gobs of gain. It looks like 20X gain at the plate of each cascode, which is about what you'd get with a 12AT7 LTP.
--
 
#46
I never said that this is the BEST amplifier in the WORLD.

BUT, it is really excellent and easy to build with readily available components.

The valves are commonly available and the Hammond 1650H transformers are available from DigiKey.

In my original build thread I described how I got around the B+ / Cathode / Heater transformer voltages. You could always get a custom transformer built but I used common 2020 components.

The hum problem could well be down to my using a balanced pre-amp in unbalanced mode originally. I've now converted the amp to balanced input and it is brilliant.
 
#47
I'm glad to hear you like the amp based on the pcb you bought.
I think implementation of any competent design is as important as the design itself.
I think the Williamson *topology* can work very well if you need lots of voltage swing for output tubes that require that.
 
#50
I think one reason the gain looks lower than expected is those unbypassed 470 ohm cathode load resistors.

Also, I modeled it as an LTP phase splitter, not as a differential driver. The phase splitter means only one input gets the input signal directly, so half the gain at each output. It looks like the original circuit was designed with the expectation that it would be driven from a balanced source.

Maybe a 10k:10k input transformer could be used to split phase...

Thanks for nudging me into modeling that circuit. It looks surprisingly good. I'll have to investigate more of Mr. Wright's circuits.
--
 
#58
I re-iterate my post.

This is an easy to build excellent amp. Easily repeatable with readily available components.

I never said that this is the BEST amplifier in the WORLD.

The valves are commonly available and the Hammond 1650H transformers are available from DigiKey.

In my original build thread I described how I got around the B+ / Cathode / Heater transformer voltages. You could always get a custom transformer built but I used common 2020 components.

The hum problem could well be down to my using a balanced pre-amp in unbalanced mode originally. I've now converted the amp to balanced input and it is brilliant.

For once, the Chinese have got it right and it's easy to build with a ready populated PCB.

The original schematic shows 5K OPTs which may be better but mine is fine with the Hammond 1650H which is 6K. I'm pairing it with B&W CM8s2's which are slightly lower than 8Ohms which would make a difference.
 
Last edited:
You must log in or register to reply here.
Top Bottom