Power supply problem

[vidc]

hi,

i made dual power supply with two separated tranfsormers. here is schematic:


problem is when it's not loaded,i measure around +26 Volts V+ to GND, and around -26.xx Volts ,V- to GND.but problem is when i connect amp to it,the negative rail drops to zero,positive remains ok (cca 26 Volts).Transformer for the negative rail have enough power i presume...i got it out of my old 30W hi-fi...,here i need about the same power for one power supply rail
what could be the problem here?why should power supply voltage drop after connecting it to load??
thank you

sorry for bad picture quality,it was edited in mspaint

[bob91343]

Maybe a short, or poor connection. Measure at the capacitors. Measure to see if the load is shorted.

[richie00boy]

The bridge rectifier on the lower winding is the wrong way round.

[vidc]

Originally Posted by richie00boy

The bridge rectifier on the lower winding is the wrong way round.

well,i found this schematic on internet because it was similar to mine and modified a little.i think that the bridge orientation is ok,i presume it should be ok because i actually get what i was expecting when the load is not connected to it...
@bob: i measured it right on caps,same thing...tried it on two different amp PCB-s,same problem...i doubt that load is shorted
now,i'm thinking it could be the capacitor...but i don't have enough experience to be sure.
is it possible that cap broke in a way that it is working "properly" when there isn't load connected and doesn't do it's job when actually some current is needed?

[bob91343]

The diagram is correct.

I suspect a high resistance somewhere, like a bad solder joint or similar. Load the negative section with a resistor to make sure the problem isn't in the load itself. In fact, measure the current to see what's coming out.

I had a similar problem a few years ago where everything looked perfect but no output under load. It was a bad solder joint, even though it looked good. Must have gotten a bubble or something in the solder and the connection wasn't being made. Yet there was enough leakage that with no load all looked okay.

[vidc]

Originally Posted by bob91343

The diagram is correct.

I suspect a high resistance somewhere, like a bad solder joint or similar. Load the negative section with a resistor to make sure the problem isn't in the load itself. In fact, measure the current to see what's coming out.

I had a similar problem a few years ago where everything looked perfect but no output under load. It was a bad solder joint, even though it looked good. Must have gotten a bubble or something in the solder and the connection wasn't being made. Yet there was enough leakage that with no load all looked okay.

ok,i will try that today.
thank you!!

[AndrewT]

power it up through a mains bulb tester.

Check each half completely independently.
Then create the centre tap at the output.
Check again. You should get 52Vdc across the whole doubled up PSU and +-26Vdc when using that centre tap as the ground (black probe).

[vidc]

problem solved.
the load was shorted in the end.the problem was that the loads were TDA 2050 and LM 3886 chipamps...and i put them on heat sink which was screwed on chasis.chasis is grounded,and tab of both chips is on V- potential,so it was shorted through the heat sink.i found that after puting 2W reistance and everything was ok,then checked datasheets of both chips and found out that i was complete idiot


thanks anyway,i appreciate your help!!

[AndrewT]

and the tester would have lit up to tell you something was trying to draw too much current.
It would also allow the PSU to power up to full voltage and when you added the amp/s it would have powered up at very low voltage and give you time to measure what was wrong.

[vidc]

Originally Posted by AndrewT

and the tester would have lit up to tell you something was trying to draw too much current.
It would also allow the PSU to power up to full voltage and when you added the amp/s it would have powered up at very low voltage and give you time to measure what was wrong.

i guess it would,but its little complicated.i'm living in a dormitory so i'm little short with equipment...but in the future i'm planning to get bulb tester (and other equipment) to make my life easier

thanks