You can use any value of resistor between the caps.
Check the dissipation at quiescent current and try to check the dissipation at continuous half maximum power output.
If in this last test you find the voltage drop too high, then add a second resistor in parallel.
eg
for quiescent dissipation on a pair of power amplifiers each biased to 110mA and with a front end draw of 15mA the total DC current through the resistor is 2*{110+15} = 250mA
On test you will find that a 0r47 will drop 117.5mVdc
Pdis = 0.1175²/0.47 = 0.029W
The voltage when delivering half maximum power will be much higher.
Check the dissipation at quiescent current and try to check the dissipation at continuous half maximum power output.
If in this last test you find the voltage drop too high, then add a second resistor in parallel.
eg
for quiescent dissipation on a pair of power amplifiers each biased to 110mA and with a front end draw of 15mA the total DC current through the resistor is 2*{110+15} = 250mA
On test you will find that a 0r47 will drop 117.5mVdc
Pdis = 0.1175²/0.47 = 0.029W
The voltage when delivering half maximum power will be much higher.
Last edited:
It's a CRC filter. The bridge diodes take the AC and rectify that into DC with a lot of ripple. The first cap bank smooths out some of the ripple, it then passes through the series resistors, which damps out the peaks of the ripple, and then goes into another cap bank that smooths out whatever is left.
Power Supplies, Filter Circuits
There's no mystery to it. It's breathtakingly simple and extremely durable. For the currents required by a typical power amp, a regulated supply would need a similar heatsink as the amp itself, hence why big brute-force filters are used for projects like this.
Power Supplies, Filter Circuits
There's no mystery to it. It's breathtakingly simple and extremely durable. For the currents required by a typical power amp, a regulated supply would need a similar heatsink as the amp itself, hence why big brute-force filters are used for projects like this.
I've read about the concepts and principles; I was trying to understand some details; e.g. why use parallel caps (C3/C4) instead of a single cap, and how the (effective) values for the CRC filter were arrived at. And why choose a CRC over a CRCRC?
(I'm, not saying any of these decisions were wrong, I just want to understand the decision making process)
BugBear
Caps in a value big enough to use a single or pair for this circuit are extremely expensive. Paralleling is much simpler and quite a bit cheaper. (And everything can fit on the PCB for easy mounting.)
Most of the DIY amps here are class-A, so the filter resistor must be able to pass lots of current with not a lot of voltage drop. So the ability to parallel a number of 3- or 5- watt resistors to get the value you want is (again) cheaper and easier than a single resistor in a value that isn't normally made. (0.11ohm 12W for a First Watt PSU)
CRCRC is marginally better filtration, but with more resistor losses, and an increase (of 50%) in size and cost.
Just a further thought: I'm almost certainly going to build NP's VFET amp using this PS, but I was just wondering how much improvement in SQ might one derive from building it as a dual mono i.e. separate PSs for the left and right channels. I suspect any improvement would be marginal, but maybe I'm wrong? Anybody have experience of comparing 2 amps that are identical other than one of them being a stereo amp and the other a dual mono?
Thank you. I guess I should see what will fit into the chassis before I order any more capacitors. I'm putting together four of these for a pair of F4s with 2x250VA each, in a pair of 4U chassis from the store. Right now I have 32x 22,000 uF caps for 88,000 per rail total.
I suppose if I could cram an entire farad in there I might... what's DIY for if not gratuitous overkill?
I suppose if I could cram an entire farad in there I might... what's DIY for if not gratuitous overkill?
i've built an F5 with this PSU and the IXYS rectifiers. I thought in my humble and mediocre knowledge that spending money on them and the heatsinks they need, was a good choice. I have my psu buzzing somewhere over the rectifiers. i think this is why i am getting hum on speakers. It's only audible if i have my ear 2" near the speaker. i have all cables twisted. i've added a monolythic bridge paralel to the ground thermistor as recomended in the F5 turbo PSU. i've also added a rf and emi filter but with no success. What do you suggest? i can go to the all monolythic rectifiers but i don't want to do it before i understand what's happening.
thanks
thanks