GE/Multilin 469 Motor Relay on AFD Output 2

[dpc]

Anyone have any experience with application of GE 469 motor relay (or other digital relay) on the output of an AFD? It's replacing another digital relay that's been in service for a few years. CTs are on the drive output, but no PTs are available there (not surprising).

The FAQ from GE says this can be done if the CTs and PTs are on the drive output. But I can't see how the relay would get much useful voltage data out of the PWM output. I'm wondering if we could just bring in the line side bus PTs?

But GE doesn't give any advice on any adjustments required to the normal sine wave settings on the 469. I'm wondering if we can just set the relay thermal overload protection as we normally do, or if there are adjustments that need to be made.

A further complication is that there is a bypass contactor to allow running across the line and the 469 is in the circuit in both situations.

Thanks,
dpc

 

[jraef]

I see it done quite a bit with Robicon drives, precisely for the reason you mentioned: if the relay is not on line all the time, it would think the motor is cold when you went into bypass. The guy I was working for at the time was the rep for both Robicon and Multilin (not uncommon). Apparently the 469 has no problem being on the output of a VFD, but I can't a speak to any others, or how they get away with it.

 

[dpc]

The 469 has an option for variable frequency in the frequency setup. Apparently it has some type special filtering that is used in this mode. I can see where the current functions would work OK, but I'm not sure if the voltage functions or power functions would be too reliable.

I've got a call into Multilin. After explaining this about three times to three different people I could barely understand, my major accomplishment was for them to open a case number for me. I felt like taking the rest of the day off!

Thanks, Jeff.

dpc

 

[Laplacian]

From the Multilin 469 manual:

"All
of the elements function in the same manner with the following exceptions: the ratio of negative to positive
sequence current is calculated from 0 to 30%, not 40%, and the voltage and power elements will work properly
if the voltage waveform is approximately sinusoidal. An unfiltered voltage waveform from a pulse width modulated
drive cannot be measured accurately; however, the current waveform is approximately sinusoidal and
can be measured accurately. All current elements will function properly. Note, however, that undervoltage and
underfrequency elements will not work instantaneously using variable frequency. If "Variable" is chosen, the filtering
algorithm increases the trip and alarm times by up to 270 ms when the level is close to the threshold. If
the level exceeds the threshold by a significant amount, trip and alarm times will decrease until they match the
programmed delay. The exceptions to this increased time are the short circuit, ground fault, and differential elements
which will trip as per specification."

If you want accurate voltage and power measurements, you need some type of filtering on the PT input. Using the line side voltage would probably confuse the Multilin as it is independent of the VFD output voltage and frequency.

 

[dpc]

Thanks, Laplacian - I guess we can set the current-related functions as we normally would and probably just forget about the voltage-based functions when running on the VFD.

 

[LionelHutz]

A PT that is capable of handling the PWM waveform on the input side may output a fairly decent voltage waveform. I've experimented with feeding a power transformer with a VFD before and the output side of the transformer had a fairly decent sinewave. There was a higher frequency from the pwm imposed on the waveform but I'd believe a Multilin would have no problem reading it. So, if a PT provides the same amount of filtering then it may work.