Rotor Earthfault Analysis in 600mw Synchronous Generator

That is a question for the manufacturer!

CALL THEM!

<...its puzzling...>

That situation will persist so long as there is reluctance to use the telephone. Take the risk!

With the unit being microprocessor controlled, any fault is going to cause the system to alarm and possibly shut down. The unit is a contiguous machine that will protect itself if there are any alarms.

What did the manufacturer say?

For one thing the filter circuits including the snubber circuits on/in both AC & DC are capacitive-coupled to ground.

Any fault in the AC side is directly reflected in the reconstructed DC waveform especially when without properly operating filter systems.

A failed AC snubber circuit can allow "drag-over" and passing/transferring of AC waveforms into the DC excitation which will cause severe disruption of the excitation, erratic generator loading, and/or damage to the rotor windings.

Operating the generator without AC snubber protection will cause very noisy operation, distorted DC waveform(s) and loss of generation capacity and/or rotor or other component damage.

I would not attempt to run a generator of any size with any excitation system snubber circuit disconnected.

A 600 milli-watt generator? The generator on my Schwinn bicycle....

Don't mind me. I'm just being a linguistic as..... again.

One of the things I've noticed over the years I've worked with massively powerful systems is the diagnostics monitoring devices work very well as long as the system works within the designed parameters. Once something fails you hope that the diagnostic system is what actually failed. This can be quickly and easily repaired. More often than not the massive system failed and also took out some of the diagnostic system with it.

My point here is that until you find and fix the root cause of your problem suspect any and all telemetry readings from a failing machine.

I agree entirely, in the 70's I worked for a large US computer company, that (correctly) designed all its equipment to test various parts as it "Boots".

The problem was, that if any part failed a test, the Boot-up procedure simply stopped and posted an error code, which had to be looked at in a manual.....not a problem you might say....but if the manual was not to hand........

I eventually made a suggestion, that took several years to implement for all the equipment we had, that it would be better to test the relevant voltages for the processor in some way to be within specs, before running the processor up fully, then test the other parts, posting a message in English on the console, sometimes in LED display, explaining firstly what area was being tested, the test that was running at each point and the results if an error was found.....but it would test everything without stopping....giving a list of erros at the end.

It was FAR easier to fault find/fix such machines after that.....though no actual studies were made of "Repair speed" as far as I remember.....

My personal take was that some of the less intelligent technical staff, could now fix machines!!

We usually only replaced PCBs anyway......not that difficult!! Only good old hardy DTL Logic, with external diodes, at that time. Also easily repaired if no PCB readily available.....

Welcome to the digital world of circuit protection. Errors in the protection circuits have no bearing on reality. This just makes troubleshooting that much more fun. Go with the manufacturers recommendations. While they could be wrong, you will still get your best answer there.