New SE With Triode Mode KT120/150

[IslandPink]

I only build choke input power supplies, is that clearer?

Hang on - you built a Steve Bench shunt-regulated power supply - did you not like that one ?

[Cressy Snr]

It's easy to play almost endlessly with PSUD chasing a notionally "good response". But whether it correlates to what you would hear is another matter. Sometimes extra R can smooth out ringing, but does that actually sound better than having less resistance?

It looks like your stepped current jumps from 25% to 100% Steve - I've always seen 75% suggested as the initial load and that's what I've used. Instinctively it feels like the voltage should settle much quicker with a smaller current step so that might be distorting things a little compared to some sims? There again if you can get a good response with a larger step that bodes well.

PSUD will get you in the right ballpark, and it sounds good so job done . Or not as the case may be .

Thanks Simon,
I'll change the first value for the step change.
You users of 'normal' computers have had years of experience with PSUD, so you'll have to be patient with my rookie questions.

[Cressy Snr]

So,
trying to put my disparate pieces of information together into something coherent; is this reasonable?

You need low impedance in a power supply, otherwise your regulation suffers, and choke input supplies are supposed to give that. It then follows that the chokes themselves have to have a low DCR. But to get the high inductance, which is necessary for decent ripple rejection and low DCR the choke needs to be wound from thick wire, and you need lots of turns of this thick wire to get 10H or more with low resistance.

This means that your choke(s)are going to be big and heavy, so your average smoothing choke like the one I have, is not going to cut it, both in terms of low DCR and current carrying capacity. Because it is in series with the rectifier output, there is a large AC current directly across it, due to there being no preceding cap, which would have shunted that AC to ground before it got to said choke. Given that raw AC current flying about, your choke has to be rated to carry twice that of the draw from your audio circuit? which means bigger and heavier still?

The one I have is rated for 400mA and I'm only drawing 200 at best, so it just about passes the current test I would have thought, but its 80R DCR is a bit high for your low impedance criteria. Introducing an RC filter further down the line, increases the supply impedance even further, and makes the regulation worse.

My Sowter, 20H/50mA, CB25s in the pre stage PSU have a 450R DCR, (900R when you add them together) so they are frankly useless as choke input supply inductors but I'm using glow tubes as a crude shunt regulator on the output, chiefly in order to get the supply down to the 200V needed for the MH4s, but I suppose they might be reducing the output impedance of the supply somewhat.The Sowter chokes and the one in the power stage PSU don't get hot if that means anything significant.

As I said, I'm only using choke inputs because the mains transformers would produce between 500 and 600V if I were to use cap input supplies. But the chokes I have are not suitable either, so as is usual the whole thing is compromised. Just when you think you're onto a winner, you realise you've fooked up again, due to lack of engineering knowledge. The way the PSUs are can't be helped and I'm working on a low budget. To me it sounds far better than it has any right to do, given what I now know.

Being on a budget and using proper choke input supplies are mutually exclusive aims, it would seem.

To get low impedance power supplies whilst on a low budget would mean electronic regulators then?
unfortunately I have no room either in or on the chassis for that sort of thing.
I'm at the limit of what I'm prepared to do now. This amp looks beautiful and sounds great as it is, so in that sense it has to be a keeper.

[Nick]

Being on a budget and using proper choke input supplies are mutually exclusive aims, it would seem.

Yes.

[Nick]

But ‘slugging’ the music comes into my head again with large caps.

I have been thinking about this and feel I have to post an opinion on it before it becomes dogma.

I think there is a danger of putting effect before cause here. Adding caps to the power supply will not (assuming good caps) reduce the HF response of the amp, and will not alter in a bad way the transient response of the power supply. If the caps are worst ones maybe, but just bypass them. However if its found that increasing the caps in the power supply alters the sound, its worth thinking why that would be. Adding caps makes the supply better in that its closer to a battery, and so it has more current capacity. I can think of a couple of reasons why that would alter the sound, but not because of the power supply, but because of the amp its feeding. Improving the LF response of the power supply will mainly increase the LF response of the amp. If the amp has problems with its LF, either due to coupling or bypass caps adding phase shifts at LF or the transformer becoming saturated with the LF signal then the better power supply will make this more audible. Also, I find that a amp will sound balanced if the LF and HF are both matched. So if its flat from 20Hz to 20kHz it will sound fine. But also if its flat (say) from 100Hz to 10kHz it will also sound flat. But if improving the power supply means its now flat from 40Hz to 10kHz, the HF response will now be noticeable and the frequency response of the amp will no longer seem balanced.

No one ever said this was simple. (Or if they did they were wrong)

[pre65]

Being "picky" Steve.

On the PSUD diagram your caps are 2 ohm, is that a guess, or a known figure ?

And re the "step" , do you need to add that with an indirectly heated rectum-frier ?

Had you considered a current tap after choke 1 to make sure it stays in choke input mode ?

Shoot me down if necessary 'cos I'm no expert either.

[Cressy Snr]

Thanks for that explanation Nick.
I get all sorts rubbish off the internet.

[Nick]

Thanks for that explanation Nick.
I get all sorts rubbish off the internet.

That may be more of it

[Cressy Snr]

This is the final version of the power supply for the output stage:

I reduced the resistance in the RC filter, to 150R; not ideal, but in the absence of another choke, it’s the best I can do. To keep the ripple low, the associated capacitor was increased to 300uF by putting another one of my JJ, 2x100uF caps in parallel with the existing 100uF half of the main dual-section cap.

The amp doesn’t hum and it sounds great. It’s the best I can do with the budget and the iron I had available. I certainly feel that I’m now getting the best out of it.

[simon]

That response looks pretty damned fine Steve - it looks to settle in 0.4 seconds without any ringing or overshoot. Job done, pour a beer, enjoy.

[simon]

There used to be a couple of guides on how to use PSUD knocking around on the net, I'll see if I can find them.

[izzy wizzy]

But ‘slugging’ the music comes into my head again with large caps.

I have been thinking about this and feel I have to post an opinion on it before it becomes dogma.

etc

No one ever said this was simple. (Or if they did they were wrong)

I'm with this thinking too and that was a good explanation. It could be a thread on it's own.

So,
trying to put my disparate pieces of information together into something coherent; is this reasonable?

etc

Being on a budget and using proper choke input supplies are mutually exclusive aims, it would seem.

To get low impedance power supplies whilst on a low budget would mean electronic regulators then?
unfortunately I have no room either in or on the chassis for that sort of thing.
I'm at the limit of what I'm prepared to do now. This amp looks beautiful and sounds great as it is, so in that sense it has to be a keeper.

And this too nails it and applies even more to commercial gear where indiviual component cost should be low, packaging needs to be "small", weight needs to be "small" coz shipping costs and technology in the form of regulation can "solve" all these issues well. It leads to different sonic outcomes but that's not a bad thing but a different thing and is easily marketed.

Your final sentence covers all you need though so well done Steve.

[JamesD]

A small word of warning when using expensive or rare valve rectifiers - There is a specified minimum resistance for the transformer and rectifier circuit that helps limit the current surges through the rectifier - low R psus sometimes do not respect this and it can lead to premature failure of the rectifier... if your using cheap or known robust rectifiers you can pretty much ignore this but its worth keeping in mind for expensive or rare rectifiers... The best way of dealing with this is to add the required resistance to both legs of the transformer / rectifier circuit not to the post rectifier part of the circuit

As Nick says - it ain't simple and you have to find out what works for you...

[izzy wizzy]

And psud can help by adding resistance to the transformer winding.

[Nick]

The best way of dealing with this is to add the required resistance to both legs of the transformer

Or to the primary.