Regulated Supplies in Tube Amps

[fdegrove]

Hi Alex,

Thanks for the sim.

There's more that I need to tell you; the Ri of the ECC807, the tube used in the original design, is 41K approximately and the µ is 150 per section.

When I first looked at the circuit together with an employee who's an enginineer in electronics we were puzzled by the design.
Especially by the high anode load on the top triode just like yourself.

We both thought it couldn't work.

In practice the reg does a fine job, we had a DVM and a scope hooked up for days in a row and voltage never varied more than 0.5V when we changed line voltage from 200 to 250VAC.

In real life and in context of the load, the White follower, this reg performs really well at keeping B+ constant.
Of course, I'm well aware that with the high value caps sitting behind it, it really doesn't have to do any hard work...

Still, as they say, the proof is in the pudding.
The o-scope showed this stage to be impervious to the outside world environment and the follower was just doing what is was supposed to do.

As Tim pointed out earlier a penthode can be used for the error amp.
Problem with most data we have access to regarding penthodes is that you need to do the hard work, figuring out amplification factors etc.

Another concern against penthodes was noise, the low noise SQ ones ususally don't come cheap.

But I could well be wrong about this, maybe it's easier than I imagine it to be.

It's rather difficult, IMO to judge regulators outside well defined operating conditions.

Anyway, thanks again for all the work.

Cheers,

[EC8010]

Runeight, I will have to read a lot of posts thoroughly later, but that last pair of diagrams you posted looks highly plausible...

[runeight]

We both thought it couldn't work.

It looks to me like the large plate resistor gives the cascode amplifier a lot of gain and this gives the regulator its good characterisics and also its resonance.

that last pair of diagrams you posted looks highly plausible...

Glad to hear it, EC. Gut reality test is critical. Who knows, maybe you'll be a simulator jockey before this is over.

The question that remains unanswered for me is whether or not the regulators affect the audio quality. Recognizing that their ability to reject line noise is essential and that simulations have some limitations, I haven't really found anything else that says regulators are important for class A circuits. Please help.

Also, gentleman, is there a way to make a negative regulator out of a tube series pass positive regulator (other than moving the ground)?

Perhaps the next topic should be using CCSs in tube amps. I know that there are very reall affects on the audio from doing this. Is it worth some investigation?

[fdegrove]

Hi,

The question that remains unanswered for me is whether or not the regulators affect the audio quality.

From experience, I'd answer with a resounding YES.
Would I turn this into a blanket statement? Definetely not!

There's a lot of ways to achieve your goal, one way is multi-stage PS design. Whether one uses chokes, regs or resistors the main goal should be good isolation from the mains giving the circuit proper to do its job.

For class A circuits the shunt reg may outperform the series reg but it wastes a lot of current along the way...not MY concern but still.

Also, gentleman, is there a way to make a negative regulator out of a tube series pass positive regulator (other than moving the ground)?

Sure, and it can be used to bias the tubes...not too sure sure what you mean by moving the ground though.

Perhaps the next topic should be using CCSs in tube amps. I know that there are very reall affects on the audio from doing this. Is it worth some investigation?

Certainly is...it's the how that interests me most.

One great source for CCSs is Gary Pimms' site.

I've always wondered about tubed CCS as I firmly believe that semi-conductor regs or ccs do have a sonic fingerprint I'm not too happy with...just my taste.

Cheers,

[runeight]

From experience, I'd answer with a resounding YES.

I certainly believe this, but I don't have enough practical experience to atest to it myself.

From this little analysis, it seems to me that regs do only three things for class A devices:

1. They maintain a steady B+ both with respect to line changes and with respect to component drift (which we know tubes do).

2. They are very effective at filtering out line noise. This can also be done with LRC stages, but it may take more of them to do it than using a regulator.

3. A good regulator may make it possible to eliminate large electrolytics on the circuit side of the PS. Or at least significantly reduce their size.

Other than the noise issue, which is quite real, I don't really see that the other items are that important. Even if I use more sophisticated models for the electrolytics, they don't seem to be so terribly bad. But, as you all know, I've been wrong before . . .

not too sure sure what you mean by moving the ground though.

I mean this:



But, this seems like a really bad idea because the pass device has to float the entire PS. So, I'm trying to understand another way to do it.

I guess the next topic will be CCS. I've already done the basic work on this (including tube CCSs). Should be a piece of cake.

But, you guys have to help. Thanks very much for your comments on this thread. You all have been very kind to me.

[exurbia]

Hi,

Runeight, this is a question that may or may not be strange

are you a locomotive man, engineer or conductor ?

just wondering

[runeight]

Unfortunately, none of the above. I'm just a railfan and, at one point in my career, a super-detailer of modern US diesel locomotive models and a model railroader.

I'm not a member of too many forums, but you are the first person to ask about the runeight!

[exurbia]

just gotta love them old warbonnet E's

[runeight]

yes, i agree. i liked the PAs too.

now, my part of texas is heavy UP with a lot of trains per day.

we should probably move this to the unrelated section or somebody might start making note of our off-topic conversation.