SMPS Fault - Urgent Cry for Help

Would you re-post the schematic? The waveform looks interesting, but without the right schematic it is difficult to ascertain the problem.

Diagram tidied a little

Check resistor values & cross check with given color code. Hope capacitors are OK. Electrolytic capacitors got life; deteriorate over a period.

Most Important: 1N4007 is not same as BYV95C, Trr value of 1N4007 is much less; that is it can not rectify high frequency.

I have worked in a textile industry where friquency drives were the order of the day. Actually they are very troublesome when it comes to repairing. In my case a lot of them were suddenly going into flame while working and few will just refuse to work. For the few that refused to work, a particuler tansistor turned out to be the problem since when changed, the inverter comes back to live. I will sujest u check all the power transistors. Check the one as shone in ur diagram.

Dickson.

Abuja, Nigeria.

Yes Mr subramanian is correct, the BYV95C is a fast recovery diode with 1Amp forward current and with 700PIV. you have to use the same diode only or equivallent (from TV flyback transformer 500DC circuit rectifier diode is suitable). IN series are not suited for SMPS with high freq oscillators. After replacing them you can check filter capacitors also. with IN series they will get heated up as AC component is also present in the rectified waveform. After replacing the rectifiers you can check the capacitors if required you may replace them with good quality capacitor with lesser loss(tan delta) This should fix your problem

"1N"!

This circuit is only running at 22kHz, i think a 1n4007 will handle that without problems. I had no choice at the time of night I was working. I have just looked up several data sheets for these and i cannot find the recovery time specified but one sheet refers to the reverse capacitance being measured at 1MHz.

Anyway that cannot be the problem because I get the same results from both sets of diodes, old and new.

The current being drawn cannot be the problem as it is less on the faulty board (when it is not being driven) but the voltage is still low.

The filter caps is not something that I had thought of. I will pull these out and check them.

Thanks for the effort friends. I will come back with the solution when i find it

Chas

BYV95C are low forward voltage drop, this will give you some difference in VD to 1N4007 diodes.

Some part of your output cicuit is drawing more current than it should, break some tracks and do current tests. Do you have an inductive pick up probe for that Fluke scope?

Nothing to worry about imbalance of +14 & -8V it is actually ± 7 or 8V supply.

As the power driver is a single-ended Power Pulse Generated, Waveform, ripple contents, variation of output with load are common drawbacks.

Check or it is better to replce 2 capacitors marked () in red

See below:

regards,

Have a fine day

If it is then no problem!

Hi Haajee,

This is a +14V / -8V output, I have the manual and a circuit description. I will try the caps as suggested.

Hi Garth,

Yes i have inductive probes for my scope but it would not help, this is all on a pcb. Anyway it is not neccesary as there is a 10 Ohm resistor in each leg to use as a ammeter shunt.

regards

Chas

Dont check the Caps...Just change them.

They most likely have bad ESR and if you don't have a ESR meter you won't be able to tell if they are good or bad. (if the boards are more than a few years old the caps are most likely on the hairy edge anyway. So just change them out and save the call back latter)

The 1n4XXX series diodes are just plain ol AC rectifiers you must use the equivalent cross or exact replacement. (They will not work at 22 kHz)

Because of the way SMPS work (Feedback regulation through a opto coupler etc...) I have found its always easier to troubleshoot them if you have a "known good load". Supplies I have worked with in the past were much higher voltages so I would use a 100 ohm 15 watt cement resistor is series with a 60 Watt light bulb. (resistor is there because a bulb is a dead short until it starts to light some supplies cant take that kind of beating) a little math and you know exactly how much current is being drawn. I would suggest that you build a load jig for these boards that replicates the specs.

Good luck....(I would bet dollars to donuts its the caps)

Bill12780