Connect with us

New build, new questions _ Lab PSU

Discussion in 'Power Electronics' started by Canobi, Jan 26, 2015.

  1. Canobi


    Jul 25, 2014
    Hi guys n gals

    So, I'm starting an online elecronics engineering course and a lab PSU would be rather handy right about now. I started looking around with the view to buying one, but gave up as I don't have the kind of cash that any of the half decent ones go for, but since I've already made almost all of my lab equipment so far, it was inevitable I'd end up having a crack at this too.

    The heart of it is a micro ATX PSU I bought by accident a while back, it's not even been out of it's box until now so it's still good as new. The case it's in is an old RS PCB exposure box that died just recently, it got stripped down to be upgraded, but I ended making a completely new one so was going spare for one project or another.

    I paired it with an equally old project box which I thought would make a great console for mounting the binding posts and such:

    View attachment 18327

    Couple of things I need clearing up on.

    There's are two thin red and orange wires I can't identify, they shared pins with a couple of thicker wires of the same colour (circled in red):

    View attachment 18327

    Does this mean they need to be tied with the thicker wires in a similar manner to the binding posts?

    If the -12v and +12v lines are combined, does it give 24v, or magic smoke?

    If it does give 24v, is there any reason a step down module could be inserted inline so the output was adjustable?

    It seemed a cheap and simple alternative solution to making my own adjustable circuit as time is of essence and in short supply at the moment.

    As to the step down module, I got one with adjustable V and A, and my hope was to swop out the pots on the module (circled in red) for something more finger friendly I could mount. I can just see those tiny screw adjusters on the little blue pots getting pretty tedious in quick time:

    View attachment 18329

    I'm assuming I'll need to find the base value of the pots in order to replace them with the equivalents but not sure how to go about it.

    Also, not quite sure what fuse rating I should use with the outputs, and if one rating fits all, or if each output need their own particular rated fuses (ie 3.3v = 5A, 5v = 3A etc etc).

    Any help and thoughts would be most appreciated, thanks (c:
  2. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    Jan 21, 2010
    Several answers in no particular order:

    Yes you will get 24v between +12v and -12v. But don't connect them together!

    The voltages generated by PC power supplies are not precise. You may find that they vary by up to 10 percent, or even more in some cases.

    Some PC power supplies require a load on one or more rails to keep the voltages correct or indeed to prevent the supply from shutting down.

    Some rails on PC power supplies are capable of large currents, whereas others are not. As an example the +12v rail is likely capable of up to tens of amps but the -12v rail may be capable of under an amp. This means you need to be careful of both huge currents from accidental shorts as well as not overloading other rails.

    In principle you can regulate the 24v from the rails as you describe but the available current would likely be low.

    Bench power supplies almost invariably have current limiting, often have floating outputs, and have well regulated outputs. PC supplies don't have these things.

    If you are doing things which require 5 or 12 volts then a PC power supply is an option as long as you are careful. As a variable supply they are less useful.
  3. Canobi


    Jul 25, 2014
    Thanks for taking the time to answer :)

    Would you mind elaborating on how to achieve 24v using the -12v +12v without connecting them? My head can't quite get around that one lol.

    I didn't realise the power from ATX PSUs fluctuated like that, I'm now thinking of using some voltage regulators on each rail to stabilise things a bit, I have a bunch of adjustable regulator modules which I could set to the correct voltages, and stick a 3A fuse inline on each one.

    I believe you are correct, and this ATX PSU will also most likely need a load to run. I found quite a few other ATX PSU projects and almost everyone used a 10w 10R on the 5v rail, though an acquaintance of mine has had to put one on each rail for his to work apparently. Hopefully not in my case, but will have to se when I get there.

    I have a 25w 150R resistor on hand that I got from the R&D guys at work (they didnt have many values in the higher ratings), but not sure if it's a little overkill. If my logic is correct, too high a load will make the PSU work harder or something as it thinks more power is being drawn so naturally want to avoid that if it is a possible issue.

    As long as I can get around 1A from the 24v rail that will suit me fine, there aren't that many application's I'll need it for, but I wanted to cover all the bases to give me some relative flexibility.
  4. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    Jan 21, 2010
    Sorry, I'm still on the move.

    Measured from the ground rail, the 12v rail measures 12v. From the same ground rail, the -12v rail measures -12v. So, from the -12v rail, the 12v rail will measure 24v.

    Just try it and see :)

    However, if you use the -12v rail as ground, the 5v rail will measure 17v.
  5. Kiwi


    Jan 28, 2013
  6. Canobi


    Jul 25, 2014
    Nice build kiwi, very tidy job.

    And thanks to all who chimed in to help me out, much appreciated :)

    While the gubbins of mine are all new, the casing is a bit more "frankenstiened" from items I had on hand.

    I also swapped out the trim pots on the variable current/voltage module for external ones. In my case a couple of Vishay p16 103 trim pots I rescued from the scrap bins at work, luckily for me, the values happened to match.

    I also added a resistance meter as I thought it would be useful in finding parallel values on the fly, and I can think of other situations where it might come in handy too.

    I'm just waiting for three binding posts and a fuse holder to arrive so I can finish assembling it, but here's how it looks thus far.

  7. Canobi


    Jul 25, 2014
    Hi all

    Finally got the PSU all wired and connected today so fired it up.

    Got the green standby light, but when I turned it on, nothing happened. No red "on" light, and the vent fan I installed didn't spin up either so it's obviously not "waking up" for some reason.

    Could it be the load resistor values aren't high or low enough?

    I'm using three 5w 33R in parallel to get 15w 11R for each rail (standard seems to 10w 10R on the +5 rail, so very close).

    Could it be the sense wires are miss located?

    I tied the two thinner sense wires to the load resistors together with their corresponding rail colours (one red, one orange) as I read that they look for a voltage to read what is actually being supplied to the rail, or words to that effect IIRC.

    Scratching my head over this one and any help would be most welcome as everything seems to be within spec and wired correctly.

    Just in case this snipit of info is of help in any respect, the ATX PSU I'm using is 300W.

    Thanks :)
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day