daisychaining 2 (or more) ATX computer PSU

[rdpzycho]

anyone messed around with a newer ATX powersupply with 2 12v rails?
would it be possible to daisy chain the 2 rails to get 24v?

some of them does have 2 separate converters inside the PSU but the grounds are common. if you cut away the common ground then you can wire it in series.

[rdpzycho]

messing with power electronics is really not for the faint of the heart. already got electrocuted by a few kilovolts during a homebrew ESD testing.

I already did modify a single PSU from 12V to 48V. I rewound both the transformer and the inductor. daisychaining would be easier though.

there are also other techniques to increase the power output from a PSU, but you may be burning lots of transistors in the process.

never mount daisychained PSU's in a metal chassis. most of them use the chassis as ground (either Analog or Earth Ground). insulate them if you should. some cheap PSU's also don't feature current limiting (I don't remember if it's specified by ATX Spec.), overload them then the fuse will blow. mid-end to high-end PSU's would happily accept shorts.

[KLeeper]

• As mentioned before, grounds are the critical part of these mods. Carefully isolate the case from all points of contact with the board and any heat sinks. When assembled, check with a ohmmeter for shorts.
• ATX supplies may require a load resistor on the +5vdc in order to turn on.
• Power on line (green wire) must be tied to ground. http://www.technick.net/public/code/...nmth_atx_power
• Lowest rated supply is maximum rating for series configuration. Add them in parallel configurations.
• For more info: http://www.wikihow.com/Convert-a-Com...b-Power-Supply

[emtffkev]

Hate to revive an old thread, but.....

Is it possible to daisy chain 2 sets of 2 power supplies in series (24v / roughly 3A), and then daisy chain those two sets in parallel (24v / roughly 6A). These aren't ATX supplies they are ones I've scrapped from some OLD dot matrix printers and they only have a 3A output but 12v each. I have 4 of them...

[NC Cams]

Yes, reread post 39 about grounds as this is the only fly in the ointment.

You can minimize problems by feeding them off of the same power strip and keep all them on/of the same high side polarity.

If the P/S's are not IDENTICAL models, that could pose a problem/issue.

[MrWild]

I'm modifying a server redundant power supply. Each individual power supply is the most robust I've ever torn apart, and each plug into a main box, not have wires coming out of them. The cases are stainless steel, but they have the same problem of using the case as a common negative for several parts of the circuit. The back circuit board they plug into ALSO has the grounds running to the case, so even though one supply was redone, I have to redo where it plugged in also.

I'm in the process of pulling the secondary power transistors from the receptical circuit board now. After, I'll do the series wiring on that board for the two supplies Will end up with WAY more power than I need, but I need 24volts and the power supplies were free,,,,, 300w each, and all I'm doing is powering 24v. air selinoids for a Kurt draw bar and air breake on my mill. Maybe it's time to look into low voltage Halogens for lighting.

[MadKad]

If I was to use 2 or even 3 PSU's would that be the right amount of power to run the driver and 3 stepper motors?

I just wounderd with the dirergrams showing just one stepper motor on

[MadKad]

sorry, can any one help with this question aboth /\ /\

Thanks

[dpuch]

You have to add up the power requirements for the 3 drivers.
If your running 3 x 3A @ 24V you will need:
2 12V (= 24V) power supplies in series that are rated at least 9A (= 3 x 3A for 12V) In this example you would ideally have a 3A fuse on each driver.

For more voltage add power supplies in series, see post #13
You amperage is limited to what the lowest 12V amperage output power supply

[AlienSteve]

The voltage at any succeeding 5v tap depends on what two points you measure the voltage.

The 5v from the "bottom" P/S from where you're referencing ground is going to read 5v.

If you measure from +5vdc on the next one up to the ground on the "Bottom" P/S, it should measure 10vdc. If you measure it to that respective P/S's ground, it should be 5vdc.

If you have at third supply stacked, it will read 5v at its own ground but 15vdc with reaspect to the "bottom" ground referenced P/S.

Still don't understand?

Hook 3 batteries (9v will do nicely) in series. Then measure the voltages at the various terminals and you'll see how the voltages relate. Arbitrarily keep the ground at one point and simple move the + probe.

You'll now understand how a -9vdc is generated from a +9v battery.

Incorrect. The common line of PS2 is at 12V, so the 5V line on that will read at 5+12 = 18V above the ground/common line of PS1. The common line of PS3 is at 24V, so its 5V line will read at 5+12+12 = 29V. Etc.

[AlienSteve]

Hate to revive an old thread, but.....

Is it possible to daisy chain 2 sets of 2 power supplies in series (24v / roughly 3A), and then daisy chain those two sets in parallel (24v / roughly 6A). These aren't ATX supplies they are ones I've scrapped from some OLD dot matrix printers and they only have a 3A output but 12v each. I have 4 of them...

Hm. Yes, I think you can, if you use isolation diodes between the supplies. They may not be necessary, but it is a good idea. Use shottky diodes for low voltage drop.

Otherwise if one supply loses power, the output of the supply in parallel with it may cause a fault by supplying voltage into the unpowered supply's output. Of course if that happens and your circuit is drawing more than one supply can provide, you may end up with fireworks. Like having one battery die first in a series string and reverse charging that battery.

There is also the potential problem of slightly different output voltages causing one supply of a parallel set to coast while the other does all the work.

I have a vaguely similar situation. I have a bunch of 28Vac 3A transformers from old Apple Imagewriter II printers, I'm going to parallel them by having separate bridge rectifiers on each secondary, going to the same bank of smoothing capacitors. I will also be putting a smallish series resistor on each, between the separate bridge rectifiers and the common capacitor bank.

If I weren't so broke, I'd just buy a $20 transformer... 10A at 24Vac (about 36V rectified and filtered).

[kartika_1]

Hm. Yes, I think you can, if you use isolation diodes between the supplies. They may not be necessary, but it is a good idea. Use shottky diodes for low voltage drop.

Otherwise if one supply loses power, the output of the supply in parallel with it may cause a fault by supplying voltage into the unpowered supply's output. Of course if that happens and your circuit is drawing more than one supply can provide, you may end up with fireworks. Like having one battery die first in a series string and reverse charging that battery.

There is also the potential problem of slightly different output voltages causing one supply of a parallel set to coast while the other does all the work.

I have a vaguely similar situation. I have a bunch of 28Vac 3A transformers from old Apple Imagewriter II printers, I'm going to parallel them by having separate bridge rectifiers on each secondary, going to the same bank of smoothing capacitors. I will also be putting a smallish series resistor on each, between the separate bridge rectifiers and the common capacitor bank.

If I weren't so broke, I'd just buy a $20 transformer... 10A at 24Vac (about 36V rectified and filtered).

You can use 2 ATX PSU in series , but the second ATX ground must be isolated . Cut the the pcb on 2nd ATX so it's not connect to ground , then connect positive lead (12V) of 1st ATX to the negative lead 2nd ATX (12V) . Now you get 24V DC from negative lead 1st ATX PSU to positive lead 2nd ATX PSU , its become 24V DC and more than 10A output . Or use 2 ATX PSU in serie but the ATX PSU must fully insolated , do not connect each other .