Energy monitor for motors on VFDs

[Electromatic]

I have a customer that does research on wheel dynamics. They have a test rig that involves two servomotors each fed from separate VFDs. The system is 480V, 3Ø, Wye (no neutral for the drives/motors). They want to record the energy use of the rig per test run. So far, about the only solution I've come up with is a rather fancy panel-mount meter connected to the conductors ahead of the drives with CTs on each phase. I could give them one setup for the whole rig or two setups--one for each motor. Am I missing a simpler option? They don't really need the kVAr, harmonics, etc. that these meters offer.
I will say I looked at some din rail mount monitors--some that could be direct connected, eliminating the CTs. The problem I see with those is access to reset/zero-out the monitor for each run would involve them being too close to energized conductors.
I'm also looking into getting the information straight from the drives. The problem there is that they are outdated, and the customer already doesn't like or understand much of the software interface already.

Any tips would be appreciated!

 

[ptonsparky]

I have a customer that does research on wheel dynamics. They have a test rig that involves two servomotors each fed from separate VFDs. The system is 480V, 3Ø, Wye (no neutral for the drives/motors). They want to record the energy use of the rig per test run. So far, about the only solution I've come up with is a rather fancy panel-mount meter connected to the conductors ahead of the drives with CTs on each phase. I could give them one setup for the whole rig or two setups--one for each motor. Am I missing a simpler option? They don't really need the kVAr, harmonics, etc. that these meters offer.
I will say I looked at some din rail mount monitors--some that could be direct connected, eliminating the CTs. The problem I see with those is access to reset/zero-out the monitor for each run would involve them being too close to energized conductors.
I'm also looking into getting the information straight from the drives. The problem there is that they are outdated, and the customer already doesn't like or understand much of the software interface already.

Any tips would be appreciated!

How outdated?
Information about the drives may get you what you need to know.

 

[Electromatic]

They are MMC-SD Smart Drives from Kollmorgen/G&L/Danaher Motion. Perhaps not really VFDs but motion controllers.

 

[petersonra]

I agree with the other poster. Before you go to far down the path you are headed see if the drive already has this capability built in.

If not, there are plenty of low cost power meters out there.

 

[Jraef]

So those are servo drives, not VFDs. There is a difference, albeit immaterial here.

What are they wanting to monitor? The INPUT power to the servo drive? Or the servo OUTPUT kW, meaning the mechanical output of the motor? They are not the same thing. If you monitor the input electrical kW, you are reading both the motor output kW, plus the electrical losses in the motor and servo drive, some of which are fixed regardless of the load on the motor, so they skew the values if you are trying to determine the motor mechanical kW output.

If the drive is capable of providing you with the output kW (and 99% of servo drives are because that is important info), then take it out of the drive. The regular programming software is fairly complex for doing that, but they (Kollmorgen) provide a version called “PicPro Monitor” that is just monitoring the drives and is simpler to use.

 

[Electromatic]

I called Kollmorgen, and they said these drives were way too old for them to offer any support. They advised I look at their online literature. I have waded through some of the PicPro manuals but didn't see a straightforward method for collecting this data.

The overall complication is that this rig is at a university for research. Students come and go from year to year, and they all seem to only know enough--whether it's the software or the hardware--to run the tests they need. (They are mechanical engineers, so not the brightest!)
The current students just said they need to know the energy consumed during each test run. I quoted installing some affordable energy meters that pretty much do it all: demand, var, harmonic distortion, etc. I will put these ahead of the drives, so yes, they will end up with readings that include system losses. They also needed it done in a couple of weeks, so my research time was limited.

Thanks as always for the input!

 

[jim dungar]

The current students just said they need to know the energy consumed during each test run.

They may be looking more for correlation and not causation.
As long as the value is consistent for tests within each student team it may be useable versus needing to be accurate/meaning full for each production run control.

 

[Jraef]

I called Kollmorgen, and they said these drives were way too old for them to offer any support. They advised I look at their online literature. I have waded through some of the PicPro manuals but didn't see a straightforward method for collecting this data.

The overall complication is that this rig is at a university for research. Students come and go from year to year, and they all seem to only know enough--whether it's the software or the hardware--to run the tests they need. (They are mechanical engineers, so not the brightest!)
The current students just said they need to know the energy consumed during each test run. I quoted installing some affordable energy meters that pretty much do it all: demand, var, harmonic distortion, etc. I will put these ahead of the drives, so yes, they will end up with readings that include system losses. They also needed it done in a couple of weeks, so my research time was limited.

Thanks as always for the input!

Yeah, then by using one that can record harmonics, even though you may not need it, you will know that it is as accurate as you are going to get when trying to read the input current of a VFD. That’s the way to go.

 

[Electromatic]

A follow up question for this potential install: There are line reactors ahead of the drives. Should I install the energy meter CTs ahead of the line reactors or between the reactors and the drives? Would one or the other location be more accurate and/or give more useful measurements?

 

[ptonsparky]

A follow up question for this potential install: There are line reactors ahead of the drives. Should I install the energy meter CTs ahead of the line reactors or between the reactors and the drives? Would one or the other location be more accurate and/or give more useful measurements?

How about putting one on each side and see what the difference is? Are the students helping?

 

[Jraef]

Won’t matter much, other than another small addition of losses through the reactor to what you are reading.

 

[Electromatic]

I explained to the researchers how my setup would be taking readings--of the system, not directly of the motors--and am waiting to hear back if this solution will suffice for them.
Out of curiosity, with the location of the meter's CTs in relation to the line reactors: The reactors are quite close to the drives, so if I put the CTs between them, they will also be close to the drives. You don't think that the circuit location (before or after the reactors) or the physical location near the drive will cause interference? Or could it possibly be a more accurate measurement of the drive/motor usage?

Thanks again, everyone.

 

[Jraef]

Interference? No. Accuracy, maybe a little, but you are already getting the majority of the "inaccuracy" by putting them ahead of the drive anyway.
Line vs load side of the reactor is likely less than 1/2% different. If you want to know for sure, look up the maximum "watts loss" on the reactor you are picking (the mfr should provide that info), then compare that to the watts (kW) of the drive at full load. I recently did one, the watts loss on a reactor for a 20HP drive was 71W, which when compared to the drive and motor at full load, represented 0.15%.

 

[garbo]

Two possible possible cheapest ways to monitor power: go into list of my dress of parameters and locate. power consumed. Would have to subtract the power from the total power consumed during drives life span. Might be able to I install a current sensor on one if the two wires on the DC buss voltage . The 480 volts three phase is rectified to around 660 volts DC. Believe the average VFD consumes as approximately 3% of I put power that would have to be deducted. If you go with current sensors on drive output you would have to verify if they would work with low frequency when dive is running at reduced speeds.When I first started working on VFD'S back in the early 1980's we had a older digital Amprobe that never gave good readings when any drive was not running at full speed. I called Amprobe company and they told me that model was only intended to be used in 60 Hertz. Our control company ( JCI ) always placed two little split current transformers on drive input to display motor current.

 

[Joethemechanic]

You know we are loaded up with those 3 phase permanent magnet servo motors running all kinds of things. Not that it has ever been a problem, but I'm just curious what the power factor is like, and what are the major factors affecting it?

 

[Jraef]

You know we are loaded up with those 3 phase permanent magnet servo motors running all kinds of things. Not that it has ever been a problem, but I'm just curious what the power factor is like, and what are the major factors affecting it?

PM servo motors will not run without a servo drive, so the drive, just like a VFD, will present a .95 PF to the line.

 

[Joethemechanic]

PM servo motors will not run without a servo drive, so the drive, just like a VFD, will present a .95 PF to the line.

Yes I understand what the power factor presented to the line by the drive is. I was wondering about measurements taken between the motor and the drive. It can't be at unity

 

[Jraef]

Yes I understand what the power factor presented to the line by the drive is. I was wondering about measurements taken between the motor and the drive. It can't be at unity

Right, but why do you care?

 

[Electromatic]

After not hearing from the students for a while, they emailed saying the tire company that was requesting this power/energy measurement suggested simply using a couple of multimeters with bluetooth to read the volts and amps. And they sent a link to a $30 multimeter from Amazon!

I explained to them that, yes, that is possible but introduces all sorts of other complications and room for error. The meter wasn't even a clamp meter, so they were potentially looking at passing 20+A through who knows what kind of flimsy test leads!

 

[jim dungar]

After not hearing from the students for a while, they emailed saying the tire company that was requesting this power/energy measurement suggested simply using a couple of multimeters with bluetooth to read the volts and amps. And they sent a link to a $30 multimeter from Amazon!

I explained to them that, yes, that is possible but introduces all sorts of other complications and room for error. The meter wasn't even a clamp meter, so they were potentially looking at passing 20+A through who knows what kind of flimsy test leads!

Are multimeter rated for continuous use?
They definitely are not concerned about accuracy if they intend to get a power value by multiplying separate amp and voltage readings in a spreadsheet.