WARNING!!! Switchmode Power Supplies

Called out to a site the other day as all outputs were on from the PLC. The outputs were in fact not on in the software and the output indicator lights were not on. The outputs were driveing a large number of LED lights and a stack of relays. The lights were all on but not fully - about 2/3 rds on. When the lamp test button was pressed the lights came on to full brightness and all the output indicators on the output cards came on.

The relays were all on and had smashed up motors on circuit breakers etc. Fortunately, no other damage.

An electrician, good one actually, had been searching for the problems for days without success.

I went through a logical process to find the problem. It was obviously common. Checked for shorts, tested for the supply from the batteries and was within limits, tested the output of the switchmode (24/24VDC) and it was right on 24 volts. Grounded the input negative to the switchmode and brought up a ground fault on the battery charger (expected), grounded the output negative from the switchmode, I hate doing this as I always use floating power supplies from these things with never a problem, no result.

After 3 hours I was getting desperate and bypassed the switchmode. VIOLA!!! All came good. I left the switchmode out of the circuit and the system returned to normal.

I did not have a CRO with me but would have loved to have seen a reading on the output of the switchmode.

I have seen switchmodes fail before but they have always either died or gone high. I have never seen this type of fault before and I have been using switchmodes for as long as they have been around.

The funny thing is that the output cards were of 2 types. There is one output card that is NPN (sink) and one that is PNP (source). They both displayed the same fault with a "leakage"? through the transistor to turn on the outputs. Really has me puzzled.

One thing we learned about designing a switched mode power supply, is that there are Min- & Max- Load parameters. When the power supply doesn't opperate within these parameters it's becomes very difficault to determine the result/outcome.

So I think it would be intresting to know what the max demand & max supply(Switch-Mode-Power-Supply) was.

BobB said:

Grounded the input negative to the switchmode and brought up a ground fault on the battery charger (expected), grounded the output negative from the switchmode

Click to expand...

I'm not clear what you mean by the "negative input". Is that what I (in the USA) might call an AC mains neutral line?

Dan

J.A.du Toit said:

One thing we learned about designing a switched mode power supply, is that there are Min- & Max- Load parameters.

Click to expand...

Single output supplies are usually fine down to no load. (there is no Min-) Multiple output supplies frequently have a minimum load current on one output (designated as the main output) for proper operation.

I'm not clear what you mean by the "negative input".

Click to expand...

It is negative input - the unit is 24VDC/24VDC.

It is single output and was running at about 4 amps - it is a 10 amp power supply.

By the way, it has been running OK for about 7-8 years until the failure.