TOP224 based power supply

Hi guys,

I have some boards with a SMPS using a TOP224. The boards came with this IC damaged. When I replaced it went damaged again. I attach a similar design of the power supply I have, you can see on page 7, but my one have 12V and 8V as main outputs and probably 12V as auxiliar output to control pin of IC.

After that I found also optocoupler U2 (CNY17-3) and C5 (47uF) damaged. After replacing them the same problem occurs. All diodes with good measurements, transformer seems OK with good inductance, we were checking also zener in case it drained too much current but we don´t find anything else strange.

We opened line between D3 and C4 and put there only 8V to avoid too much power on the control pin of the IC (it says 9V as maximum) but again IC damaged. In that moment we measured a peak of 20V on the main 12V output (C3).

Now we are thinking if this design needs a load connected on the 8V output which I think there isn´t. Can be possible that without load it switches too much getting overvoltage at output and burning the IC?

SMPS are notoriously difficult to repair. Most of the time you gotta hope for failed secondary smoothing caps.

If it's on the Primary side, you need a Variac to make sure you don't blow up the thing when checking it on Mains. Only experience teaches you that and that can't all be explained in this thread.

I deal with SMPS daily on CRT here.... luckily they are mostly all the same design, which I understand. I know the various circuit values and parts out of my head. So, it's mostly a breeze for me.

tvtech said:

SMPS are notoriously difficult to repair. Most of the time you gotta hope for failed secondary smoothing caps.

If it's on the Primary side, you need a Variac to make sure you don't blow up the thing when checking it on Mains. Only experience teaches you that and that can't all be explained in this thread.

I deal with SMPS daily on CRT here.... luckily they are mostly all the same design, which I understand. I know the various circuit values and parts out of my head. So, it's mostly a breeze for me.

Click to expand...

Don't you assemble 'kits' to repair specific models, that's what I used to do - there are usually a small number of specific models that you see a LOT, and once you've repaired enough of them you know pretty well all the faults. So it's easier and faster to make a'kit', and then change all the parts before you bother looking any further. I seem to recall I had about three 'kits' for TV's?, two Grundig (one bipolar, one FET), and one Tatung.

And yes.....many years ago I thought I was clever. Designed my own SMPS.... using Viper 50. Bought 100 and blew 6 up. So I stopped. Still 94 in a tube here somewhere.

We couldn't get the SMPS transformer 100%

Looking back now, I think my PC board layout sucked.

I would attempt another SMPS design. Maybe. Nope.

It's an art that few master.

As a side note...Nigel you rock. Good man and I enjoy our discussions.

Especially when we share common ground.

tv

Smile did not come out right.

Supposed to be

Hi,
after connecting a load at output the IC is not burned anymore but look the scope signals. For some reason the voltage at control pin is not constant so is discharging until it gets 4.7V and at that point the power at output stops. We have checked for capacitors, opto and resistor at that feedback circuit, we don´t know why the voltage is not stable at 5.7V at that point.
The other signals are voltage outputs 12V, 8V (feedback) and voltage at zener (green line)
Look page 3 of datasheet.

If you are getting 20V on the 12V fedback supply after a single cycle, I'd say the transformer is designed for a lower mains voltage or it's incapable of operating without a high current load on the output.

The IC has no chance to properly start up and achieve regulation under the present conditions.

I'd expect that to an extent, theres no compensation in that system so theres going to be some oscillations.

Finally we supplied the board to be tested at original load and it worked fine!!! Thank you everybody.