computer power supply?

[jaencer]

This is probably obvious, but what if i took TWO similar 12v computer supplies and tied the 0V of one to the 12V of the other.. would I then have 12-0-(-12)? or would i have two blown fuses?

[Westerp]

Originally posted by jaencer
would I then have 12-0-(-12)? or would i have two blown fuses?

You would indeed end up with 12-0-12 Using two of
these you get a 5A 12-12 split power supply for just $ 33.30.
I use cheap chargers for most simple/fast projects and with a
choke/cap/c filter there pretty clean.

p.s. When you do use computer psu's do lift one from ground and don't let
the metal casings touch eachother as they are connected to the 0 and you
would still end up with sparks

[Rainwulf]

And you will have to remove the earth connection from the one "below" earth.

[jaencer]

Originally posted by Rainwulf
And you will have to remove the earth connection from the one "below" earth.

you've confused me...
the ground "below" earth (actual earth ground) is the ONLY ground we're talking about removing, right?
on the mains AC plug to the outlet, i was going to just hacksaw off the round ground pin part of the AC plug, from one of them. (i'll have fuses like everywhere, too)
Then the +12V output of one is soldered to the 0V output of the other, and this junction becomes output ground. (the other two output wires are +/- 12V)

we on the same page?

[Westerp]

As long as one of them is not connected (lifted) from ground, the pin you mention
it will indeed work. When you leave it connected you will in fact short the 12v rail
from this psu through the common ground at the outlet. Also you will have to
connect the 'pwr-ok' to 0/ground and the 'pwr-on' pin to a switch.

[pacificblue]

deleted

[jaencer]

Exactly :unconfused: thanks, will report back with results.
this couldn't possibly be too bad sounding, i mean, the thing will put out 40 Amps :holy-s#*t: of regulated power (per supply rail!) (and the ripple is only a few mV on my scope). besides being really big and ugly, that's not too shabby, and would probably be perfect in a LM1875 chip-amp.

Edit: BTW, there should be no difference which of the two PSUs gets their ground chopped off, right? It was going to be the one with its 12V being output-ground.

[norazmi]

Hi,

This topic should go to Power Supply section actually, there is more expert there to explain how to make psu work and perform better with amplifier.

Actually i`ve experience with mod psu`s to use with power amp. Its great result with huge of current it can provide from it. But will need some adjustment with it. I`ve done it before with modified current limit and voltage limit and rewinding the trafo.

This will need some knowledge to mods psu and perform better with your needs and until now i`m still using it to power up nmos200 power amp design by quasi, dx amp by destroy_x and many others, the result is awsome and the sound like more bass and treble, its looks like that psu`s supply with huge current and great performance. Voltage drop is too small is about +- 2v at full power.

I would like to suggest to who have some advantage to make own board offline smps can go through Luka design its perfect to get power up 1kw and maybe more.

btw, for anyone who mods psu can have more current by winding it back with correct core and cable and correct winding technics.

p/s - be carefull with high voltage and do not try with simple thing and rush, do some search and read and learn from it.

regards,
Azmi.

[454Casull]

The PSU I'm using right now is a Corsair TX750 which is a great unit but is horribly noisy. Even after regulating the +12V down to ~+9.5V using an LM317, there was horrendous crud on the output of an opamp-based buffer that I built. I had to use a wall-wart supply.

[jaencer]

I wouldn't have used a regulator for that, you didn't first try a good 'ol CLC filter? if you can get an old crossover from some speakers, often they have a nice big inductor. Try CLC of about 2,200uF, about 1.5mH, and 2,200uF, it'll kill even the nastiest ripple, and you can add a snubber (like .47uF + 1R or whatever) to try to undo some of the high frequency response loss.