Optimal ripple voltage on b+ supply

[nhuwar]

I was wondering what is a good percent ripple to shoot for when designing my b+ supply filter.

Thanks
Nick

[sorenj07]

I don't know about percent but under around 500mV should be fine, assuming it's push-pull.

[SY]

That's dependent on which stage, what the PSR of the stage is, what the overall S/N requirement is...

For a raw supply, 5% ripple is a good figure. That's probably good enough for the plates of a p-p output stage. For more sensitive stages, you'll have to determine what your requirements are, then design appropriate filtering or regulation. A popular power supply design tool (and justifiably so) is PSUD2 from Duncan Amps.

[sorenj07]

Originally posted by SY
That's dependent on which stage, what the PSR of the stage is, what the overall S/N requirement is...

For a raw supply, 5% ripple is a good figure. That's probably good enough for the plates of a p-p output stage. For more sensitive stages, you'll have to determine what your requirements are, then design appropriate filtering or regulation. A popular power supply design tool (and justifiably so) is PSUD2 from Duncan Amps.

My only gripe with the program is that it doesn't understand how current loads from tubes gradually increase from 0 as they warm up... Stepping loads can only do so much to simulate this gradient in reality. It took me a lot of grief and warnings about negative current draw before someone explained this to me.

[SY]

Yup, no question, software is no substitute for common sense. It's a nice tool, though, and the ripple predictions are damn accurate.

[nhuwar]

Thanks for the input I actually downloaded that app a while ago just havent used it yet. The power supply in question is pp final stage plate ps.

Oh Sy I should be getting the Morgan Jones book in 2 day's ordered it yesterday but amazon is very slow to ship things I've found.

Thanks

Nick

[Wavebourn]

Originally posted by sorenj07
I don't know about percent but under around 500mV should be fine, assuming it's push-pull.

If push-pull when well balanced don't produce audible hum when idle it does not mean it does not intermodulate the signal and ripples.

[Sheldon]

Originally posted by sorenj07


My only gripe with the program is that it doesn't understand how current loads from tubes gradually increase from 0 as they warm up... Stepping loads can only do so much to simulate this gradient in reality. It took me a lot of grief and warnings about negative current draw before someone explained this to me.

I think you are talking about two separate issues. It is useful to know voltages for all possible operating conditions. When modeling circuits, I almost always use a current source for modeling. Then I look at the supply output with my target currents and with very small currents. I use the latter to determine the highest voltages any component will see, so that I won't have over voltage failures.

The step function is for modeling possible ringing behavior - a performance issue, not a safety issue. My understanding is rudimentary in this area, but it's good to check at least a couple of step changes, especially if you are using inductors in the supply.

My experience mirrors SY's as regards accuracy. This requires measuring your transformer primary and secondary resistance, as well as unloaded input and output.

Sheldon

[nhuwar]

I have actually just been playing with the program an with my current filter cap the output is remarkably clean but that doesn't mean i won't change it once I hear it.

what I have is 550 micro farad at 1000vdc but I dont know how to find the capacitors resistance.

Nick

[sorenj07]

Originally posted by nhuwar

what I have is 550 micro farad at 1000vdc but I dont know how to find the capacitors resistance.

Nick

wow, what kind of capacitor are you using? or is that a big capacitor network?