short circuit finding in switchmode power boards for plasma tvs

[ant17]

hi i am working on a plasma
the power board has a major short in it some where i checked directly across the main rail no short. but it blew the fuse in the house i was wondering what the main way of hunting  down a  short circuit with both digital and things like transformers and such as switch mode supplies. I have seen where some people use a milliohm meter and some use a power supply with a limited current injected to the circuit and then measure the voltage drops i don't want to damage board anymore how is it most oftenly done in the industry if someone could enlighten me would be greatful thanks

[vk6zgo]

It is a while since I fixed a lot of SMPS.
In those days they mostly used bipolar devices.
That said, the way we looked for  short circuits was using a DMM on the Ohms (not "continuity") range, with obviously, no power applied to the circuit.

A good idea is to disconnect the later sections of the circuit, & starting from the ac Mains input, check for shorts firstly between Active ("hot) & Neutral, & then if that looks OK, between the + & - outputs from the bridge rectifier.
If there is a "short" there, disconnect the switching part of the PSU, & check again.

If it now looks OK, you have narrowed the problem down to the switching part.

This is where it gets messy-----maybe the active device/s involved are faulty, or some of the passives around that area.

If, however, you don't see a "short" anywhere on Ohms range, but the thing blows Mains fuses or smokes when powered up, you may have something like a shorted turn in the switching transformer.(Sony call it a "choke" but if it looks like duck, quacks like a duck....)
This will usually kill the main active device, though!

No doubt there are things I have missed, like there being a huge load on the secondary of the switching transformer, which might give similar results to a short at the input.
No one can give you a generic "road map" through SMPS faultfinding that will address every possible situation.

[shakalnokturn]

Most of the time (not always) the short is a biggish semiconductor.
My lazy method since I have the right toys is a current limited lab supply and thermal camera. This can be easily replaced by a combination of schematics, experience and deduction tough.

[ant17]

thanks guys for your help but i do have another i read that some people use current limited lab supplies and measure the voltage drop over the components to find a short but i was wondering should it be an ac current limited supply or dc only because the inductors and transformers would show as a short  to dc current am i right  thanks

[shakalnokturn]

May I advise reading about SMPS design and looking at a few schematics?
Of course a coil will only show it's resistance to DC, and in the case of a SMPS that is usually low enough to be called a short circuit, I can't think of any design where a coil is thrown alone across a power source though...
Troubleshooting with a DC voltage and current regulated lab supply is usually the best solution.
When working on low voltage power rails make sure lab supply is voltage limited below the specified rail voltage.
Bring current up slowly to a value that makes sense, it may be easier for troubleshooting to exceed the specified current for a given power rail, less than double is usually fine. Really overdoing it will usually give you extra work but may also point you to which part of the circuit the short is in.