Parallel non adjustable PC PSU's ???

[NeilP]

I can always get a lot of PC PSU's but they are never voltage adjustable...If I smoke a few in experimentation no problem...but still would rather avoid it if I can..

Trouble is, although they may have stable outputs...they are always different voltages..by 0.3 volts or so variation...This batch of 4 I picked up this morning vary from 11.3 to 11.9 volt open circuit on th 12 volt line....OK...so cheap DVM so maybe they are all a little higher...but at least I would be expecting the error to be consistant.

I have been running three in series to give 34 volts to run an iCharger 3010b...but was wondering about running another three in parallel to up the current.

OK...so i could choose the PSU's to string together to voltage balance each string to give as close as possible voltage...but was wondering about balancing the parallel strings as is being done with Meanwell PSU's

I believe diodes are involved and I would imagine they go in series with each string...but that is where my knowledge stops. So probably high current rectifier types, stud diodes or similar...but???

Anyone shed any light on how to balance up these strings...or not bother just parallel up the series strings and hope of the best?

Neil

[fechter]

If you put them in parallel, there will be a tendency for the one with the highest voltage to do all the work. I don't think diodes will help or are necessary. The supply has diodes as part of the output stage, so there won't be much issue with back feeding the lower supplies.

If you take 2 supplies in parallel, for example, the highest voltage one will be doing all the work at light loads. As the load increases, eventually the one doing all the work will reach the overload point and either shut down (hiccup), or the current will fold back. If the current folds back, the second one will kick in and take up the slack. If it goes into hiccup mode, I don't think you'll get much more than what a single supply would do by itself.

It would be possible to hack the circuit to add some adjustment, but evey model is a bit different.

Another way to make them share more evenly would be to add some resistance between each supply and the common connection point. This could just be a fairly long piece of wire. As the highest one takes load, the voltage at the end of the wire will sag a bit due to resistance, so it can drop to the point where the lower one can contribute power. They won't be even, but you can get more than what a single one can do.

[NeilP]

OK, thanks

I have 6 prep'd up here, so will probably just choose them to give the closest matched no load voltage, and see what happens. If anything goes wrong, can always get plenty more from the local dump

They are all different units, so working out the circuit one each type is not an option...I may get a few the same, but more likely not. So trying to find the voltage control circuit on one type, then cooking it...just too much wrk..and beyond my knowledge

[NeilP]

In digging through the pIle of PSU's I had, I came across one that did have an adjustable voltage output. No outwardly visible signs, but two pots on two separate daughter boards of the main one
It is a NMB-MJPC-270A1 as found in Sony Vaio PC towers.

So I only need one to adjust the voltage of one string to match the other

It is strange that of all the PC PSU's all the 12 v lines were within three quarters of a volt...except one...up at 13.5 volts..could be faulty I suppose...wont use that one just in case


Gary, what would I need to look for ...would it work......to add one ov your current limit boards to a PC PSU I imagine these SMPS all work in a similar way,?

[dak664]

No load voltage is not a reliable measure as leakage, capacitor size, minimum switching duty cycle, etc, have a large effect. I''d dedicate a cheap HF current meter to each to monitor and adjust under load. Velcro works well for mounting them on the PSU case.

[fechter]

Gary, what would I need to look for ...would it work......to add one ov your current limit boards to a PC PSU I imagine these SMPS all work in a similar way,?

Yes, they all have some similarities. I'd look for a TL494 chip or something similar. From that, and a datasheet, you can figure out where the voltage regulation feedback is. An adjusting pot is the other thing to looks for, as the limiter circuit works by lowering the voltage when the current tries to exceed the limit.

[NeilP]

OK, thanks..The chips are easy to find on here...no pots on any except one.

Looked at the chips...datasheets...and of the 3 I have seen. different numbers...but pinouts all seem the same. will try and work that ....my spoke nipples arrived from China today....2 weeks after the spokes...they forgot to send the nipples !!


I did have a box of small meters kicking around somewhere from years back, easy enough to shunt them up to give me a few ammeters

[NeilP]

Ok, so here is a pinout
http://circuits.datasheetdir.com/66/TL494-pinout.jpg


Which pin should I be tracing back from Pin 3 'Feedback' or 13 'Output Control' or one of the others

[NeilP]

Ok, so here is a pinout




Which pin should I be tracing back from Pin 3 'Feedback' or 13 'Output Control' or one of the others...what am I looking for ? a resistor...that maybe needs changing for an pot?

[fechter]

I'd guess pin 1 or pin 16. Ususally pin 1.

If you measure them while it's on, you should see something like 2.5v on one of them typically.
If you trace pin 1 (or 16) back, you should find one resistor to ground, and one that goes to one of the outputs. The one that goes to ground can be replaced with a potentiometer to make it adjustable.

Here's a somewhat typical schematic (they vary widely in style):

230W switching power supply schematic.jpg

(247.98 KiB) Not downloaded yet

[NeilP]

OK, thanks. I'll examine some of the dead units and see what I can see

I zapped one of a string of three of the units two days ago...I had about an hour before I went out somewhere..and I thought...lets connect up another three that i have stripped the wiring back on to just the ones I need. I had them all lined up, output sides isolated from the earth and a good no load stable 36 volt output. Thought I would try to load them up so ...rushing and not doing a proper job....I was holding the pair of bare output wires to a light bulb to load them up. As I pulled the wire closer, I do not know what touched what amongst the other wires but there was a tiny spark and they all failed


At that point I put them all away and went off to work.
So now I only have one string of three working again I have actually found two that are adjustable, so hopefully will find some more like that...back to the dump again tomorrow

edit: dead unit now replaced.

Now will have to find a pair of ammeters and a load to try and parallel them and balance the output of each string. If it all goes pop...well I pick up some more from the dump...again

[NeilP]

Well I have started another thread along a different angle with these PSU's...just to create a sting in series...6 or so long. But current limiting and initial connection seems to be the problem..as soon as I connect the string to the pack, one or more or the supplies drops out.

So now need to look at current limiting on these...with that TL494 chip control that and is it going to be an easy hack?


viewtopic.php?f=14&t=28632