Power supply

This is not a plc question but I am sure with the amount of knowledge that is on this forum you can help me with the answer.

I work in an electro-plating facility and we use rectifiers to get our DC for plating these are just like any DC supply but on a bigger scale (15vdc 10000 amp) any way they work by firing SCR's to a three phase transformer out of the other side through diodes with a cap. bank ect.... anyways I have one rectifier that when not under load the voltage at the anode and cathode jumps up to about 18 volts under load it drops back down to around 10 volts like it should. the controls for the rectifer are just potentiometers (sorry if it is misspelled) and a SCR control board the control voltage does not change ever. I do not expect that anybody has this exact sceaneiro but like I said it is the same as any other variable DC supply. If anybody can give me some ideas of what could be causing the voltage to jump up when not under load it would be helpful.

thanks in advance

There are several things that could cause this.
Transformer voltage increase on one winding
Diode leakage (reverse bias) could add to voltage
Capacitor has changed value so charge increases.

If uncommon then it sounds like something is breaking down. My suspicions are in the transformer and/or caps, probably the caps mainly because they will charge from the ripple.

It sounds like you might have a bad diode or two in your rectifier group.

I will assume you are using six diodes to get a three-phase full-wave rectifier. If all six of the diodes are working you will get six overlapping conduction zones going into your filter cap. If you draw this out you will notice that there isn't a whole lot of dead area between the conduction zones. More correctly, you almost always have a diode that is conducting and able to supply current to the filter caps.

Now let's assume the worst case that will still supply voltage. All the diodes but one are blown. You have something that will be conducting about 1/3 of a cycle (I think?). You are still supplying current to the filter caps and the peak votage is still the same but the available energy is lower.

Without a load on the system your steady state DC voltage will settle at the peak voltage from the AC input to rectifier. Your filter caps (or bus caps if you will) will charge and keep the voltage at this level. However, as you start to draw current from the filter caps the voltage will settle to the RMS voltage being supplied from the rectifiers. If you don't have all your diodes conducting the RMS voltage will be lower than what you expect.
It could just as easily be one of the SCRs causing a problem as one of the diodes.

It comes down to something Ron said in his post:

My suspicions are in the transformer and/or caps, probably the caps mainly because they will charge from the ripple.

Click to expand...

If you have a blown diode or SCR you will have some SERIOUS ripple and the output voltage won't be able to stay where you want it.

Keith

I like that better, had a thought along those lines but lost it.

I have worked on this type of sytem, but it has been many years. The SCR's where on the secondary side of the transformer. Each SCR (three in total) had a very small thermal cut-out, physically attached to it. Each Thermal fed a 120V relay. After that I forget. But what happened in this case was the thermal cut-out contacts where loose (broken) and you could only notice this as the machine was running. It also fried the relay. I do realize that is not exactly what you are talking about, but if you have a copy of the schematics, I would gladly help.
John