Instructables
In my sophomore year of college at the University of Minnesota, I started into my main electronics classes, and needed a good power supply for working on lab projects at home. My roommate Adam told me about somebody online who had converted an ATX computer power supply into a lab bench power supply, so I decided to do the same thing. You can also check out this link for a very similar guide by their user Abizarl. I have also documented this project on my website at http://www.mbeckler.org/powersupply/ if you are interested.

Warning! There are several large capacitors in ATX power supplies, that will store a dangerous charge for a long time. Please let your power supply discharge, completely unplugged from the wall outlet, for a few days before opening it up. You can probably be seriously hurt, so please be very careful. I am not a lawyer, but I hereby release myself from as much liability as I can, for any sort of injury you sustain, or any trouble you get into.
 
Remove these adsRemove these ads by Signing Up

Step 1: Background

First, a bit of background on a typical ATX power supply:

Computer power supplies are Switch Mode Power Supplies (SMPS), which use high-frequency switching circuit elements to provide a high-quality output voltage, with good energy efficiency. One side effect of this technology is the minimum load requirement that each power supply has. In order to function properly, the power supply needs at least a very small electrical load connected to it. In other words, ATX power supplies will only work if you have something connected to it. We will be using a power resistor to provide this minimum load.

Also, modern power supplies do not simply have an OFF/ON switch, they have what is known as a "soft" power switch. This normally makes no difference to the user, as the computer behaves the same, but when you shutdown your computer, the motherboard can turn off the power supply when it has finished shutting down. This requires us to add our own power switch to the power supply chassis.

To protect our circuit from accidental (and careless!) short circuits, we will install some fuse-holders and fuses, which will disconnect the circuit supply lines if too much current flows. The size of the fuses are up to you, but a 1 amp fuse will work just fine for most circuits. You really should put fuses on all supply lines.

Update: While the diagrams show fuses on all voltage rails and no fuse on the ground line, when I actually built my power supply, I was young and foolish and only put a fuse on the ground wire. It's much safer and a better idea to put fuses on all signal lines and not the ground line. Thanks to many emails and messages on Instructables about this oversight.
1-40 of 429Next »
I have 20 pin ATX supply and I need 12 volts and at least 50 amperes from it for my experiment. So can a connect all yellow wires of my supply(12volts, 16amp each) to make a single 12 volt, 50 amp power supply? I have a total of 5 yellow wires in my supply...
I don't think the 16A rating is for each wire...Even though you likely have many red 5v wires, the current rating on the label isn't for each wire, but for 5v overall. Most newer ATX power supplies have two or more separate 12v supplies, which complicates things to the point where I am not sure if you can just connect all the wires together like that and expect it to work. If you need a single voltage, high-current like that, I would suggest searching for a dedicated 12v power supply, instead of trying to rig up something this way. Good luck.
turbiny24 days ago
Still experimenting if it works correctly or not then i'll add switches and variable dc converters.
The problem is after i plug it it wont start working for another 10/15 seconds and when it starts the voltages are rather high 12-14;5-6;3-3,8
Is it because of the missing ressistor?
photo-28.11.14 12:03.jpg
matthewbeckler (author)  turbiny20 days ago
I have heard that if your voltages are rather high, that that is due to having zero or very little load on those voltages, and that if you add a proper load that they will drop down to the correct voltages right away. A switching power supply is sort of like trying to keep an air-filled balloon at a certain height above the ground by bumping it upwards periodically. If there is very little gravity (very little load) the balloon (voltage) doesn't go down very quickly, and even the smallest bump upwards will tend to overshoot the desired height (output voltage). Also, I think the specification is actually for 3,3 volts, not 3,0 volts, but still 3,8 volts is too high.

I'm not sure what is causing the start-up delay. You have tied the green wire (PS On) to a black wire (GND)? You might try adding a resistor load to the 5v or 12v wires.

I'm having a problem with one. I have the brown connected to the orange as it was in the plug, and the resistor on the red and grnd as it should be. But when I try to fire it up it will only spin the fan for a second and then quit. It won't stay running. Any suggestions? Much appreciated.

What do you mean by "the brown connected to the orange"?

Perhaps try moving the power resistor to a different voltage rail, such as 12v or 3.3v? Maybe your particular power supply needs the minimum load on a different voltage line? Or it needs a minimum load on two or more of the voltage lines? I've never seen any specifications for this sort of thing, so you might need to experiment with your power supply. Let us know if you figure out what's going on.

There was a brown wire piggybacked onto an orange in the 20 pin plug so I after removing the plug I tied the brown wire to an orange. Also, I've tried the resistor on the 3.3v, the 12v, and the 5 volt. Also tried two resistors. One on 3.3v and one on the 5v at the same time. I've set this one aside for now pending a solution. If I can figure it out I'll post my results.

Thanx!

matej.blagsic2 months ago
Did you use 3 fuses as you have shown in the schematic because I can only see 1 fuse on pictures
matthewbeckler (author)  matej.blagsic2 months ago
Yes, sorry for the confusion. When I originally created this project and instructable (and took the photos) I only had one fuse, and it was on the ground wire. After that, I learned more about best practices and a lot of people suggested that it's a bad idea to fuse the ground wire, as you always want that one connected. Instead, it's better to put individual fuses on each power rail. I updated the schematic to show a fuse on each power rail, but I haven't rebuilt my power supply.

Thank you for your reply. I really needed that to finish building this power supply, as it will be only power supply as the "lap power supplies" are much more expensive. :

harshesh2 years ago
Will a 10W 100Ohm Resistor work as the power resistor ?
When I use the 10W100Ohm Resistor and connect the green line with the black line, Some noise comes for a second and then stops .. Sometimes the noise comes for 5 seconds and then the Fan starts but then immediately stops ! .. Is there something wrong with the Power Supply I am using or is it the resistor ?
matthewbeckler (author)  harshesh2 years ago
Hi! Somewhat counter-intuitively, the larger the resistance you have, the less current it was draw from the power supply. A 10 ohm resistor will draw 500 milliamps (mA), a 100 ohm resistor will draw 50 mA, and a 1000 ohm resistor will draw 5 mA. Presumably there's some minimum current draw that your power supply requires on the 5v line (and maybe the 3.3v line too?) and if you don't draw enough current (resistor is too large) then it won't stay on.

It really sounds like the behavior you describe is due to insufficient load current, or perhaps some other problem. I would try a power resistor with a smaller resistance value to try and draw more current. You could try connecting two of your 10W 100 ohm resistors in parallel, producing an equivalent resistor of 20W 50 ohms that will draw 100 mA, and dissipate P = V * I = 5 * 0.1 = 0.5 watts shared across both power resistors. Good luck!

I always wondered why all tutorials use or suggest to use specifically a 10 Watt 10 Ohm resistor for dummy load. In high school I was taught: P=V*I => P=V*V/R
In this case: P= 5*5/10 = 2.5 Watts So I always wondered, why use a 10 Watt resistor instead of 5 Watt. Even considering possible spikes the resistor wouldn't be damaged. A constant load that produces 5 Watts of heat or more would be necessary to damage it.

matthewbeckler (author)  myouknowwho11 months ago
Yep, only 2.5W burned at 5v. I would guess that 10W 10Ohm resistors are more standard and easier to find, if you can find a power resistor at all. They are not used for many applications and can be difficult to find.

[head-bulb.jpg]

What about use a incandescent bulb instead? Maybe hader to find soon becouse of led lamps tough.

Car light bulbs (back up/brake/dome, etc) are incandescent, durable, and run at ~12v. They also put off significantly less heat (less power waste) than that fat resistor.

matthewbeckler (author)  JohnB142 months ago

As far as I know, the whole point of the constant minimum load is to always draw a small amount of current from the PSU to keep it from turning itself off. If the PSU requires at least, say, 500 mA current draw at 5v to keep it running, then it doesn't really matter whether you dump that power into a resistor as heat, or an incandescent bulb as a little bit of light and the rest as heat.

The real question is to figure out how much minimum current draw is needed on one or more of the power supply rails. I've never seen any definitive "minimum load" data, and for a hobbyist level project like this, the power resistor(s) aspect is much more of a "try things until it works reliably" process than anything else :-) Maybe there are some power supply datasheets we could check? I'm not really sure, but that would be useful.

Extraneous heat matters to the components inside of an enclosure. I realize we can use cooling fans to expel it, etc., which is using more energy to expel wasted energy (and another noisy fan harshing my zen-like bench experience). Using an external incandescent "Power On" indicator as a load seems appealing to both heat management as well as my tree-hugging sensibilities. ;)

I agree that "minimum current" data would be useful.

arush26 months ago

hi friend

Currently I'm working with some dc motors and i need negative voltage with high current, but when I connect the GND of the "power supply 1" to the 12V of the "power supply 2", to use the GND of "power supply 2" as -12V, i have a short circuit, and therefore the "power supply 2" is switched off. Is there some way to get -12V with high current?

thanks

matthewbeckler (author)  arush26 months ago
Most likely your power supplies are connecting GND to the AC earth ground (ground plug on your AC outlet), which produces a short from 12V to earth ground on PS2.

Why do you need negative voltage? Most applications don't need negative voltage. If you are driving DC motors, you should be using an H-bridge circuit in order to drive the motors in either direction, without needing to use negative voltage. More details here: http://en.wikipedia.org/wiki/H-bridge

There are a lot of different H-bridge chips that are really easy to work with. Which chip to select depends primarily on how much current your motor needs, and what voltage (6v? 12v? 24v?).

thanks for your response :), the problem that i have is due to the driver that im using to control the motor, i have a analog output signal of -10v to 10v from the PC (data acquisition card), and the driver only converts this signal in "high current voltage" (the driver is similar to use the h-bridge but using two transistors), then i can control the movement speed and the rotation direction using one output, but for the driver i need -12V and 12V (with this voltage is enough because i don't require high speed) with high amperage, but when i use the 2 atx in parallel to get 12 and -12v volts one of them is turned off :S.... then i dont know if it is possible to connect two atx in parallel, or is necessary to do other thing

matthewbeckler (author)  arush26 months ago

I see, sounds like old PLC-style control systems :-) How much current do you need? If it's just a few amps, you could probably build your own from an AC-AC transformer (down to about 15 volts AC) and then a full-bridge rectifier to generate the positive and negative DC voltages, and finally a 7812 and 7912 linear regulators (with heatsinks and in/out capacitors). If you need more current, you're probably going to have to drop some cash into a real power supply, which may be a little more difficult to find such a bipolar power supply.

Alternatively, I have used some fancy power supplies in EE labs where the negative terminal of the power supply wasn't permanently connected to AC earth ground (it included a little wire that you could use to connect them if you wanted). If you had two such power supplies, or a power supply with two outputs, you could connect them together in parallel like you mentioned.

Zarix7 months ago

I know some people have been having trouble getting their PSU to stay on. And this was my solution. I could not find any information on the current rating of the 12v CPU plug in the newer supplies. But when I put my load there it works for me. Maybe that is where the load needs to be for some newer supplies to stay on.

stuffandwires7 months ago

hello, this is a nice project that i'll try to make in these next days. Regarding the fuse matter, i see only a fuse holder in the front. Is that the only first fuse that you had connected to ground? The rest of them are in a 3x fuse holder or individually?
thank in advance

matthewbeckler (author)  stuffandwires7 months ago

The way I built it, there was only one fuse, on the ground line, which isn't a great idea. What would be best is to put a fuse on each power rail (3.3v, 5v, 12v, etc). If you could find a 3x fuse holder that would really help to reduce the space needed for fuse holders. Good luck with your build!

mothball8 months ago

I have LCD Controller that need 12V4A. Could I connect through 12V PSU directly?

matthewbeckler (author)  mothball8 months ago
That should work, if your PSU can provide > 4 amps for the 12v line (check the label on the side of the PSU for the current rating).
ionsight1 year ago
do you guys think a 160w unit would be practical for use testing my arduino and various small electronics. It could be a nice compact test supply
matthewbeckler (author)  ionsight1 year ago
Most power supplies' wattage rating is an overall power rating, for all the different output voltages (mostly 5v and 12v though). You should check the label on the side of the PSU to see how many amps of current it can provide on the 5v rail. This will directly indicate how many arduinos you could run. I think most USB devices will draw less than 500mA, unless they are a high-power device like some some cell phones and tablets, which can draw up to 2 amps sometimes, depending on how you connect them. FWIW, the USB and ATX specification for 5v rail is the same for both, 5v +/- 0.25v.
sir small doubt regarding the power supply!
im done with my led cube project and im using our clg LAB RPS to power my cube.now im planing to use pc smps.can i dircetly connect the smps 5v rail same as rps? smps has very high current ratings,does it make any damaga to my cube?
thankyou
matthewbeckler (author)  sdevi reddy1 year ago
I am not sure what you mean by "RPS". I would expect that the SMPS 5v rail will work correctly with your LED cube project. Normally, a device such as this will only draw as much current as needed, regardless of how much current the power supply can provide.
tnq sir!
RPS- regulated power supply.
nwlaurie1 year ago
Just built it over the Easter weekend - it works exactly as expcted. Many thanks for an excellent circuit diagram (I know it's not complicated but it's still really handy to have a decent diagram!).
HINT: if you have a choice of old power supplies, pick a QUIET one!!! And test repeatedly as you build.
I opted o remove some of the unwanted output cable bundles at the root, if you do this make sure you have a really powerful soldering iron or you'll end up doing more cooking than clearing (I used a gas-powered one for this bit).
Thanks again,
Nick
nwlaurie1 year ago
Just built it over the Easter weekend - it works exactly ad expcted. Many thanks for an excellent circuit diagram (I know it's not complicated but it's still really handy to have a decentdiagram!).
Thanks again,
Nick
Crucial971 year ago
Hi sorry for getting back to you late, i was testing my power supply last night and a resistor popped. so i am going to remake another since we have 3 non-used power supplys
Crucial971 year ago
Hi,
I have been reading your instructable and I have built my own until i found the grey power-good wire. what did you do with your (probably grey) power-good wire?
Thanks,
Zach
matthewbeckler (author)  Crucial971 year ago
Hi Crucial97, thanks for the comment. I connected it to an LED to act as a "power is good, we're ready to go!" indicator. Check out the diagram on step 2 (http://www.instructables.com/id/ATX--%3E-Lab-Bench-Power-Supply-Conversion/step2/Planning/) and you can see how I connected the LED+resistor to the grey wire. Good luck with your project!
hschmutz1 year ago
Hy there
I know this is an older post but i will try to make this power supply for me.
What fuse do you take ?
matthewbeckler (author)  hschmutz1 year ago
Hi, thanks for the comment. Choose your fuse based on your application (what you're going to power with your power supply). The fuse is primarily there to protect your circuit from mistakes in your wiring, not to protect the power supply from your mistakes (it doesn't need protection). If you are running small circuits like an Arduino that will draw less than 500mA, then a 1 amp fuse should be fine. If you are charging RC batteries from your power supply, then you'll need a 10 or 15 amp fuse. Basically get a fuse that can handle just a bit more than the maximum amount of current than you expect to need for your application. Good luck!
Skarz881 year ago
How do you use the -12v and -5v?
1-40 of 429Next »