Method for testing discharge resistors in 3 phase capacitors

[kenneth edmiston]

A couple of months ago I was troubleshooting a blown fuse on a power factor correction capacitor in a 4160v motor starter. While I was working through it I noticed that my DMM (fluke 289) would charge the capacitor to a certain degree while on the Ω function. The current from the meter would charge the cap.

Then I could switch to the dc voltage function and read a FALLING dc voltage value.
That lead me to believe that I was charging the cap, and the discharge resistor was operating properly.

I believe the internal resistance of the meter while on the voltage function is up around 1-10MΩ. Can’t exactly remember. For this purpose I’ll just say 1M.

My question is: Am I seeing a properly operating resistor, or discharging through my meter?

If the resistor was open, would I still read through the meter, making this method worthless and unreliable?

The rate of change was not aggressive, yet not sluggish the way I remember. Maybe somewhere around 1v/.5-1sec. My math is limited and I’m unsure of what would be inline with a 1M resistor. I would guess that the discharge resistor would be around the 1-500k range but I’m not certain. They say to wait 5 minutes or so to discharge a fully charged capacitor, but I’m sure that’s overkill for safety reasons.

Just a thought..
Thanks!

 

[Borek]

Perhaps I am missing something, but I don't see how estimate anything without knowing capacitance involved.

 

[kenneth edmiston]

Is the capacitance value necessary? It could be, I don’t know. But if you put a small current on it, it will charge to some value for however long you have that current present across the plates, right? It’s not the capacitor I’m testing here, it’s the discharge resistor installed to create a load when the cap bank is de-energized.

The (lightly) charged capacitor will have a voltage associated with the charge, and as it discharges the voltage will fall. I wasn’t intending to fully charge the cap, just see if it would discharge.

 

[kenneth edmiston]

But 10uf was the capacitance of that particular capacitor

 

[Borek]

Speed with which it will discharge (hence the speed of the voltage drop) depends on the capacity, if memory serves me well in a theoretical and ideal case it should change as V = V_0 e^{-\frac t {RC}} (where V₀ is the initial voltage, t time in seconds).

 

[kenneth edmiston]

Speed with which it will discharge (hence the speed of the voltage drop) depends on the capacity, if memory serves me well in a theoretical and ideal case it should change as V = V_0 e^{-\frac t {RC}} (where V₀ is the initial voltage, t time in seconds).

Ah I just added the time bit to show that it was not discharging slow enough to make me think it was discharging through my meter, which had a high resistance.

 

[Tom.G]

Fluke 289 Specs:

Input Impedance: 10MΩ
Open circuit Voltage, Res. range: 5V
Max current, Res. ranges: See Table

 

[Windadct]

But troubleshooting a blown fuse in a 4160 application will likely require the ability to apply voltage in that range. It can be tricky.

The balancing and/or discharge resistors in the cap should be on the nameplate, with a DMM this is all you should see. If they are wire wound ( common for higher power caps) they may have faulted, but you will not see this at lower voltages. If the resistor was open, you probably would not blow fuses...

The capacitors leakage - comes into play. These are typically "self-healing" caps - so the actual capacitiance after time in service may be very different than the nameplate value.

The only "real" way to ask the Mfr for a field test procedure. Very likely a Variac, connected through a Potential Transformer to the cap, and bring the voltage up and plot the current.

 

[kenneth edmiston]

Thanks a lot for the reply, it was informative.
I don’t expect anyone to read this article (5min read) but it explains what I’m asking, only he is testing for cap discharge through leakage (very slow) and I’m testing for cap discharge through resistor. Of course a DMM would be a bad choice for measuring leakage but I’m not sure if the DMM would impact the circuit enough to disregard the test method for a lower value resistor.
http://www.robotroom.com/Capacitor-Self-Discharge-1.html

I’m wondering if my meter is discharging the cap fast enough to not be able to differentiate between a working discharge resistor and a faulty one. Or if 10Mohm input impedance is enough to see that the discharge resistors are in fact closed and operating properly.

I’m looking for a qualitative measurement.

Thanks!

 

[Windadct]

NO reason you could not pre-charge to something like 24 V and then measure the decay over time, but as I posted above, there are a few variables.

 

[kenneth edmiston]

NO reason you could not pre-charge to something like 24 V and then measure the decay over time, but as I posted above, there are a few variables.

That’s what I was getting too. 24v give or take.I was just curious to see if I was missing anything. Thanks for the help

 

[kenneth edmiston]

What do the balancing resistors do?

 

[Windadct]

Basically the Balancing Caps are a voltage divider - when you put the caps in series, ( usually MV caps are a number of lower voltage caps in series - internally) - this helps ensure the voltage is more evenly distributed, since the capacitance of each of the caps may vary, esp over time. Sometimes these are used for the discharge, sometimes not.