Regulated Supplies in Tube Amps

[runeight]

BJT Series Pass Regulator

One more simulation for this part of the thread. Here is a simple BJT series regulator.



Setting C=0, we get the following behavior for delta-V/delta-I:



Same behavior as the tube regulator. Set C=470u to get:



This regulator shows about 3.5ohm for audio spectrum. About the same as the tube regulator except that it is a little better at low frequencies. And it uses many less parts.

If we use a large cap at the output, C=330u we get about what we got before except that the better low frequency response is obvious.



This regulator looks like an RC stage for frequencies above 100Hz and then acts like a near constant resistance below that value. I won’t bother posting the log graph. I think this illustrates it well enough.

So that finishes off this part, namely resistive analysis.

These results are exactly what any circuit designer would expect when regulating for a class A circuit.

Next installment will discuss the reactive piece of the problem. That is if anyone wants me to go further.

[SY]

Why not include adjustable 3-pins in the analysis (with some varying output and reference capacitance)? It's trivial to use a high voltage BJT to keep the I/O voltage difference within ratings. I think the results might look interesting.

No yawns, I find the model info interesting and don't really have the capability to do that. My experience is more on the practical, hands-on end, but experimentalists need theorists. Thanks for some interesting work so far!

[EC8010]

I clicked on one of those little icons, and a menu came up, so I clicked on "show picture." The icon went away, but nothing else happened. It's certainly stopped me from commenting...

[dhaen]

EC8010,

The images are deep-linked to another site. They should be there automatically. Presumably your web-browser couldn't contact that site, at the time you tried.
Larger images can be posted this way.

Cheers,

[EC8010]

This time, it all burst into life, diagrams and all.

I'd comment that there's more to a regulator than its AC output impedance. It's worth remembering that the mains is not a nice stable voltage, and it does jump around a lot, particularly when people put kettles on during the adverts. Additionally, a handful of active devices with associated Rs and Cs is often cheaper and smaller than LC filtering that can achieve the same effect. Regulators also nail motorboating. Personally, I think there's a lot to be said for circuits where the only disturbing influence is the audio.

[runeight]

There is definitely much more to a regulator than the AC resistance.

I'm going to try to look at the phase shift part of the problem next. Although, I don't think there will be anything much to say about it.

Your points about line regulation and motorboating are two of the best reasons that I can think of. I'm trying to think of how to analyze the pre-filter portion of the PS. Also, regulators stablize against component drift over time.

What yet puzzles me are statements about the "sonic quality" of regulators. I did this little study, partly for my own sake, because sonic quality implies that normal RC supplies have frequency dependent impedances that affect the audio quality that are "corrected" by regulators.

Given that the load effect of regulators in class A ciruits is so small, I'm not sure that I understand this at the moment.

[Original Burnedfingers]

runeight

I must have prompted this thread when I commented on my view of wanting/needing a regulated supply. No offense meant there and I for one certainly have enjoyed your connents and circuit diagrams in my 5692 topic.

Is a regulated supply actually needed in a class A line stage? This is a good question and one that might not be able to be answered via graphs and charts but with an educated ear and gut feel. Your charts seem to indicate that a regulated supply is probably a waste of time and parts. The ear might seem to tell a different story. Audio to me is more than what can be proven with charts and graphs. Different circuits may perform the same and even test the same. Why do they sound different?

A "Quicksilver KT88 and a 8417 amplifier are both capable of the reproducing the same signal, test the same, and yet sound different. Why? Is it possible that there is more to this hobby than what can be measured with a chart or graph? Different capacitors seem to result in changes that can be felt/heard. Why is it not possible to percieve that regulated supplies can also have an effect?

A heart felt welcome to you by the way. I am very certain that you have and will continue to contribute in a very positive way.


Joe

[runeight]

Joe,

Thanks for the welcome.

At this moment I sit somewhere between the subjective and objective view of audio. I believe that audio quality is a measurable thing, although the measurements that are most comonly made may not tell the whole story. On the other hand, any honest physicist will admit as to how many times we've been surprised by discoveries of things that we never even conceived of before. As Hamlet once said, "There are more things in heaven and earth, Horatio, than are dreamt of in our philosophy."

Part of my purpose in doing this little study was to understand this for myself. I am still willing to agree that regulators have an impact on the audio of even class A devices. But, so far, I don't believe it comes from their resistive properties as load regulators. So, if there is an effect, it must be something else that is at work here.

I hope to do a little more investigation and post the results. I trust you all will indulge me a little more.

[EC8010]

Regarding regulator "sound", you might like to think current rather than voltage. With your proposed use, the audio current will be very small, but there are plenty of applications where the audio current can be quite large, and the behaviour of a regulator as it strives to hold voltage constant in the face of changing current could well be significant.

Oh, and welcome to the forum. We need more physics.

[dhaen]

Yes, the regulator's effect will be application dependent to large extent.

I echo EC8010's welcome, and the comment about physics. There are an increasing number of "compitent" members, which can only be a good thing.

Cheers,