Valve Amp Power Supplies

[Nick]

Well, the delay is the rate a 50uf cap will discharge when loaded by 80ma. Remember you have put a current source as the load, so what you are seeing is the cap discharging. So not really of much direct relevance to the amps frequency response given it will never do that as a load to the cap. Well, I guess it will when you go from zero signal to full on signal, but its not going to happen in the space of a cycle of signal frequency.

[Cressy Snr]

Cheers Nick.

[Cressy Snr]

Getting too excited there. Regulators are to do with supply impedance. That much I know.

[Cressy Snr]

Out of interest try shorting the 220r I can't see the point of any resistance you don't need in the supply to the output stage.

Removed the resistors and the bass performance has improved. Not in terms of slam or attack, but different bass instruments are better defined and more tuneful. It's not massive, but it is certainly worthwhile. A 6080 series regulator like the ones I had on my now almost forgotten 6B4G monoblocks, might improve things further I'd wager. That will have to wait for now but it'll get done before the summer.

[simon]

Removing the resistor also removed a second filter which would have significantly reduced ripple, but you haven't commented on this so perhaps wasn't audible?

I've often mused how small the ripple should be ideally. It can be chased in isolation, less is better right? But there are consequences - more R/C/L/filter stages - which aren't necessarily a good thing. "Just enough" is right I suppose.

[Cressy Snr]

Well that’s the thing isn’t it. I’ve never figured out how much ripple should be removed passively and how much actively with a voltage reg, so I’ve all but abandoned trying.

Nick, Andrew, Mark, JamesD and Paul have by far the most experience in this field. I have very little. That’s the main reason I started this thread

There is still silence but no longer a palpable silence with this version. With the setup I had at the beginning I had never heard such quiet, if that makes sense. Difficult to describe.

[simon]

No I think I understand what you mean. And it seems to demonstrate the trade off - subjectively improved sound (less R in the supply) but at the cost of a very slight increase in background noise (less R in the supply).

I've spent a lot of time over the years going round in circles in PSUD trying to get low ripple and a fast step response. Some loads just didn't seem to lend themselves to this, or getting a quick response at all even, at least using the iron I already had.

Pretty much all the amps I've built have two filter stages as they're easy to get quiet. The one I do remember building with a single LC (plus small first cap to tune the HT) was the Bugle. And that kicked surprising butt for a 45SE. You might even have heard it on the old horns a long time ago Steve.

[IslandPink]

Shunt !

[Cressy Snr]

Shunt !

How dare you sir! Same to you!

[ed]

not sure how relevent this is, but back in the day I came to the conclusion that choke input was the way to go. This is 2008 and looking back I'm not sure if I understood it all back then..I certainly don't now...but with minimum understanding and armed with psud I discovered this:

from this

ps-572-mk1.jpg (230.23 KiB) Viewed 3098 times

moving to this was a huge improvement

ps-572-mk2.jpg (229.67 KiB) Viewed 3098 times

[Nick]

Why would you decide the improvement (I assume its a heard improvement) was the result of the choke input instead of the change in B+ voltage?

[ed]

yes there was a huge improvement in what I heard...
but I was chasing more voltage...
I think I was guilty of witnessing the removal of the ripple and immediately thinking that must be better, and that may have clouded what my ears were telling me....

so, in my defense, I wanted more volts and I wanted rid of that ripple....so objectively I had acheived both and subjectivity went out of the window.....until everybody else that heard it said it was the biz....

which of the two things actually brought the improvement I'll never know because I couldn't isolate the changes to 1 iteration. I'd like to think it was both changes in equal measure, but that's not very scientific and is at odds with good engineering practice.

[Nick]

It would be interesting to see the effect of a step change in load, in theory the choke input should have greater regulation.