diyAudio (
-   Power Supplies (
-   -   CRC Power Supply design, help me understand it (

pjpoes 2nd March 2007 03:44 PM

CRC Power Supply design, help me understand it
I have designed on PSUD2 a series of CR filters in the power supply to come up with something for my tube amp power supply that I can easily fit within the case. That makes chokes out, so I thought I would look at this. I can display more graphs, but I'm not really sure what I'm looking for here. This is modeling the supply over 5000ms after a 30 second warm up and soft start initially. It choses a big curve, looking at the Voltage at the Load Resistor, that varies about .001 volt. Is this good? I'm used to seeing graphs that show flat lines, or ripple, not this, so I'm just not sure what to look for. If this is showing a voltage variation of only .001v then that seems good. Currently the amp just has a capacitor bank of I believe 8-1500uf caps rated at 425 volts. I want to insert resistors to make a CRC filter system to reduce noise. I know it will reduce the voltage, but it looks to me like all the tubes are running at the limits to too high a voltage anyway.

AndrewT 2nd March 2007 05:24 PM

your 5k load is drawing 80mA.
With a light load you will have a tiny ripple and PSUD's figures confirm this.

You can zoom in on the ripple, to see what is happening.

The current ripple in c1 & c2 show reversed effects.
It is normal for the first cap to suffer extreme ripple but you have fiddled the ESR to hide the normal problem.

A rCRCRCRC supply will inherently have low mains ripple particularly with the RC=220r*1.5mF=330mS.
You must ensure that the last C can maintain good current supply with little modulation of the supply voltage as load current changes.

Can you afford all those motor run plastic film caps? 4*1.5mF is a BIG volume and even BIGGER cost.

Increase the delay from 30S to 120S to see the near final voltages.

pjpoes 2nd March 2007 10:25 PM

Thank you I appreciate that.

I have to find out the ESR of the caps to get those numbers right, but I already have 15 Rubycon 1500uf 450 volt caps to do this with. Otherwise I would probably not have gone that route.

Thanks for the input by the way, its appreciated. I'm going to look further into the design and see what I get. I was looking at an LC network and found it to have more ripple, which suprised me, I thought chokes generally worked better than resistors.

Do you know what a Hash Choke is by the way. I see them come up for sale from time to time. They are generally 250UH, rather than a usable size like 5h. Can adding them in series, like you would a normal choke, have any benefit with a CRCRCRC filter network as I have constructed here. I guess it would sort of be a CLCRCRC, but the L would be under sized for the job.

pjpoes 4th March 2007 05:24 PM

What is the ESR supposed to be, they start at 2 ohms, which seems way too high. When I looked it seemed like most of them had a much lower esr of around .1 ohms or so. Do I have this wrong, if so, where do you get the right ESR spec, I was going by the spec sheet, and graphs sometimes available with them. Does lower ESR actually lower the ripple at light loads?

AndrewT 4th March 2007 05:46 PM

looking at that first cap in the rCRCRCRC filter.

The higher ESR than it's partners has forced PSUD2 to predict a low value of ripple current though the first cap.
I think this is misleading, and may not be a good model for your completed PSU.

If the ESr were lower or even set the same as the following caps then you will find that the ripple current will be much higher.

Do the numbers and see what happens.

If the ripple current does indeed increase, then you must choose a cap that can safely handle that amount of ripple and keep it's temperature down so that it's other parameters do not deteriorate.
A set of parallel caps can solve the excess ripple, but the model does not show you have a problem (yet).

What quiescent current does the amplifier draw?
How much current will the amp draw when it delivers significant output current?
I think you should be running a number of models to see more realistic supply conditions.

All times are GMT. The time now is 10:10 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio