Power Supply Switching Pulses

[tech2014]

I don't have much experience with viewing switching pulses in SMPS power supplies. I have a tv with no VS voltage and want to test the pulses on the mosfets that control the VS. What kind of shape should I see. I have a pic below of what I am seeing but it looks more like spikes then pulses. Any help would be greatly appreciated.?

[Alastair E]

Re: Power Supply Switching Pulses

Are these the PFC or the Main PSU MOSFETS?

What other tests have you done...?

In all the sets Ive done--cant ever say Ive checked the drive on the Gates of a PSU MOSFET.
--Probably as they can be touchy devices and years ago testing such places has caused severe blow-ups due to the scope-probe adding load/changing drive to the Gate....
They are something I never touch--never the Gate directly when powered....

I would expect a drive signal of around 10-15V square-wave rather than 0.6v as you have

Prolly either the Chip or the MOSFETS themselves.

You dont say what set or what model--so its not possible to help much further....

[tech2014]

Re: Power Supply Switching Pulses

its a Samsung plasma TV. The power supply is a BN44-00161A. The power is intermittent. I replaced a couple domes capacitors, tv worked for 2 days then failed again. I shorted the PS_ON to ground to trick the p/s so I could get voltage measurements. As soon as the p/s turns on, the relay clicks once then clicks again. I was able to get voltages and all the voltages are good except for the VS is only 2 volts. I am testing the p/s alone with no load.I get 400v at drain of QS801 but nothing on QS802. I have PFC voltage but no VS. I have the p/s schematic below.

[tom66]

Re: Power Supply Switching Pulses

You need to jumper VSON high for Vs/Va to come active (only PSON is no good)
Note PSON is shorted to GND but VSON is connected to 5V... it's odd... don't ask me why...

[tech2014]

Re: Power Supply Switching Pulses

I have a standalone DC power supply. If I put 5v on the VSON pin would this be sufficient. If so I will try that and check the VS and VA again.

[tw2005]

Re: Power Supply Switching Pulses

If the FETs have not shorted have you tried a reflow yet on those? Generally these crack joints, then you get intermittent power. If left they start cycling inevitably the fets can blow.

[tech2014]

Re: Power Supply Switching Pulses

I haven't tried reflowing but will do that , thanks.

[tw2005]

Re: Power Supply Switching Pulses

This is a list of parts i was putting together for the 50" model, same PSU board except they used FETs with a higher rating.

I found all the semiconductors and fuse blown. those diodes are underside

F802 T8AH, 250V
QS801,QS802 FQA18N50
DS801,802 kds184
.( RX867, RX869), (RX862 and RX878). 20ohm (2012)
RX863 and RX868 10kohm (2012)

DX803,DX804 KDS184
RX802, RX803, RX832, RX833 100 (2012)
QX801 and QX802 FQPF11N50

RX804 and RX805 10k ohm (2012)

[tech2014]

Re: Power Supply Switching Pulses

ok I checked the switching pulses on QX801 and 802, which controls the VA and I get VA voltage. I get this(see pic below). I get noise/spikes on QS801 and 802, which controls the VS. Should I just replace QS801/QS802.



TOM, my VS_ON is always 0V. I tried connecting 5v to the VS_ON and the VS voltage stayed at 2V. The PS_ON is a constant 4.2V. When I short it to ground, the relay clicks on, then off. I was able to get all voltages except VS.

[tom66]

Re: Power Supply Switching Pulses

Set your scope to DC coupled for looking at the pulses. Are you using an isolation transformer for this test?

The pulses look reasonable, but they are quite large, over 30Vp-p, which is abnormal for a driver circuit.

It is possible you are looking at a transformer gate drive output, which would have a higher amplitude than the switching pulses to the transistors. The pulses at the transistors shouldn't exceed 25V peak and more than about 20V continuously, or they can be damaged. A good design won't even need to go above 20V - the transistors will be well and truly saturated around the 11-12V mark - the higher voltage helps improve the switching time, but not much else.

If you are getting gate drive pulses and no fuses are blowing, your transistors are probably ok -- there are some exceptions, but in general transistors usually fail shorted or leaky, which will blow fuses. It is very rare -- not impossible, but rare -- for MOSFETs to fail in other ways (BJTs tend to fail more often with leaky junctions, but my experience of FETs has been in all but one case the transistor shorts, and in that one unusual case the transistor was used in a linear regulation application for which it was not designed.) I'd then focus on the coupling cap for the half bridge driver, as that's a common failure point on Samsung plasma PSUs. Seen quite a few fail, will cause no Vs/Va but everything else appears to work.

[tech2014]

Re: Power Supply Switching Pulses

Hi Tom. I know I am going to get slack for this, but I don't use an isolation transformer. I don't do many tv repairs so can't justify the cost. If you know of a decent priced one I will buy one just don't want to spend $200 on one.

I took the pulse reading off of the gate of the mosfets. So the spikes I am getting on the gate of QS801/802 is ok and I should not replace them. At the start of the repair, there were 2 domed caps in the p/s so I replaced them and I tested all of the electrotytic caps with an ESR meter and the tv worked for 2 days then it crapped out again.

Which cap is the coupling cap?

[tom66]

Re: Power Supply Switching Pulses

How are you not using an isolation transformer? Are you floating your 'scope? Yeah, that's pretty dangerous, if you are, as anything on your scope is now LIVE and that includes the thin plastic case not designed for non-earthed applications. I picked up an isolation transformer for about £25 in the UK, from a test equipment reseller. Maybe check around building site auctions etc - they tend to use a lot of them...

That waveform is not normal for the gate of a FET, it is way too large in amplitude and as I said, could damage the FET.

Coupling cap is connected to the transformer, usually where the transistors meet or on the other side. Its basically ac-coupling the transformer drive so you don't get flux-walking effects.

[tech2014]

Re: Power Supply Switching Pulses

Ok I will buy an isolation transformer. I did some reading on the gates of the mosfets and now I get about 4v p-p(see pic below)

Maybe I had did something wrong the first time. I get nothing but small spikes on the gate of QS801/802.

[stj]

Re: Power Supply Switching Pulses

http://na.suzohapp.com/amusement_pro...ies/80-2000-00

iso-transformers are standard in arcade equipment.
you will probably find them for $5 on ebay.

[tom66]

Re: Power Supply Switching Pulses

Note: You want to isolate the DUT, not the scope. Isolating the scope is the same as chopping off the ground lead. No safer.

So the transformer must be rated to run the power supply - I use a 750VA one myself - you could probably get away with a small one but only if you ever use it with low load (unloaded power supply) or with a light-bulb in series to limit load current.