Constant Voltage Transformer Capacitor failing

[SAkieran]

Hi

On the site i am working on we have 2 CVTs. The problem is the capacitor bank, first at the one end of the bank and now at the other end, is that the bridge pieces are failing. This has caused a breakdown of the respective capacitor. The bridge pieces are discolouring, but what is strange, it only seems to be on the end of the capacitor banks. I am trying to look up what the problem can be but i am hitting a wall. Would you guys have any idea where the problem could lie.
Any suggestions or explanations would be greatly appreciated.

 

[RAMConsult]

If something carrying current is discoloring it usually means that it's carrying much more current then it's supposed to. I would get a noncontact thermometer and monitor your bridge pieces for hot spots, it could just be poor crimps. But the real question is why is the current so high, for that we have to look to the nature of the load. Is it a traditional linear load, or is a lot of nonlinear loads with high harmonic content? I suspect that harmonics are finding their way back to the CVT and are being amplified by the oscillatory nature of its operation. If so try putting a purely resistive load on and see if the temperatures go down.

 

[SAkieran]

Thanks for the reply, it ia currently ran on generator load before the mains supply is turned on. Maybe this could be causing a problem. I was also thinking the crimps could be the problem.

 

[Skogsgurra]

The most common cause for this is that the CVT's cooling is insufficient. CVTs run hot and it is important that cooling air can circulate, that it isn't mounted so that the hot core is beneath the capacitors and that cooling air is below maximum allowed temperature.

This is probably not because of high current (CVTs are current limited, anyhow) but improper cooling. Find the manual for the unit and make sure that mounting is in accordance.

A very common mistake is to mount two units too close to each other. Or that someone has put paper or something else on top of them.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[RAMConsult]

The generator adds an additional layer to this. If I understand correctly this is a small (under 10kW) consumer generator that outputs directly to the CVT whose output then goes to a main breaker in a distribution panel. "Normsl" operation is to start the generator and then close the main breaker after it stabilizes. If so then I believe that there is probably interaction between the generator and the CVT especially when unloaded or lightly loaded. This is made worse by the fact that the capacitors form part of a tuned circuit that is resonant at its rated frequency, and the fact that the generator frequency is nowhere near utility grade regulation thereby materially affecting the proper operation of the CVT. Couple the frequency variations with the poor voltage regulation of the generator and you have an interesting lab experiment. At this point I would be using an oscilloscope to Look at the various waveforms into and out of the CVT under loaded and unloaded conditions. The damping of ordinary meters may not pickup the harmonics that I think are circulating in the lines.

As a side note I tried to determine the temperature at which copper discolors and was surprised that there is very little definitive info out there except for this thread on Eng-Tips...thread330-194516. So the non-contact thermometer is your best bet to see how hot the crimps are getting, or simply recrimp them and flow some solder into them.