Variable-frequency drive (VFD) output without load (motor)

[gnurf]

I have acquired an old 1.5kW/2HP VFD [1] with aim to power a drill (which I don't have yet) with a three-phase motor from the single-phase outlet in my garage. It was supposed to be in working order, but alas it was not as the LCD display, for one, is only displaying gibberish (yes, I have cleaned the contacts etc). The LCD display aside, I'm more worried about the fact that all three output lines (U, V, and W) have no phase difference with respect to each other and are also at approx 30-40% of the input voltage. The 'V' line is also significantly more noisy than the U an W lines. Does this make sense under any normal condition/state of a VFD? All testing is without a load because I don't have a motor atm.

I'm trying to slowly hurry through the documentation [1] looking for clues, but with the return-by-date approaching fast I thought I throw this out here in hope of a definite "No" to the above question.

Thanks.

[1] https://www.schneider-electric.com/en/product-range-presentation/694-altivar-58/

 

[essenmein]

First I would suggest looking at the three phase outputs with a scope rather than a meter, that will tell you a lot more.

Now assuming the drive has started properly and is sitting there with zero modulation index, basically an "idle" mode of sorts, you would expect all the three phases to be switching at 50% duty, in unison, so the line to line voltage is zero. So the voltage would read, measured with a DC meter ~ half the DC link voltage, which if its a simple bridge rectifier, then this would be Vin(rms)*sqrt(2)/2, so should be more along the lines of 70% the rms input voltage.

****Extreme word of caution: If it is a bridge rectified input DO NOT connect your scope ground to the inverter power ground unless they are true floating differential probes or using an isolating transformer. This internal ground will be going negative wrt Earth ground/neutral by the line voltage every half cycle and you will blow something up or hurt yourself*****

Unless you really know your way around inverters, and given the screen is not working, id take it back... Does it have CAN or some other communication network you can talk to it on? if not how do you plan on setting it up?

 

[jim hardy]

Hook up three huge incandescent light bulbs and see if it'll light them.

when i look for the English instruction manual it tells me to log into twitter. for me that's too high a price.

https://www.homedepot.com/p/Philips...descent-Heat-Lamp-Light-Bulb-416743/202768698
series them up if you need 240 volt.

 

[gnurf]

essenmein,

Thanks for your kind reply. I did look at the output lines with a scope, hence my comment about zero phase difference between U, V, and W in the original post. I especially found your "idle mode" comment interesting, and zero line-to-line voltage is indeed what I measure. Furthermore, I measured

Vac(rms) = 242V (between L1 and L2, mains input)
Vdc(U, V, W) = -139V (between each output and Earth ground, or chassis. Note the negative voltage.)

I'm not sure I understand the DC relationship, if any, between the output lines and Earth ground. Is this what you meant? If so, it's about 57% of the input, which I interpret as either a suspect rectifier bridge (and therefore DC bus voltage) or, more likely, 1) a duty cycle which is not 50%, or 2) something else.

Please feel free to correct me or fill in any blanks, but know that I haven't done my homework here and do not expect to be spoon-fed this topic. In fact, I'm pulling the plug on this project and will try to find something a bit newer. I have enough junk (or 'gems' as I call them on the way into the house) in the pipeline as it is waiting to be repaired or whatever--a situation I'm sure many in forums like these can identify with.

jim,

Thank you for that common-sense suggestion. We don't have those bulbs around the house anymore, unfortunately, so I couldn't try that.

 

[anorlunda]

We don't have those bulbs around the house anymore, unfortunately, so I couldn't try that.

There's an opportunity. Long after incandescent light bulbs have been outlawed, someone could reinvent them and market them as "load testing devices" If you save a few now, they might become very valuable in the future. Just kidding, or am I?

 

[jim hardy]

If you save a few now, they might become very valuable in the future. Just kidding, or am I?

i bought a couple cases of 100 watt's several years ago. I like them for room lamps in winter when the heat is welcome.

For ceiling fixtures though i prefer low heat CFL or LED because they don't cook the wires in the fixture.

But I'm maybe abnormal that way - still have a few cans of Freon 12.

 

[essenmein]

essenmein,

Thanks for your kind reply. I did look at the output lines with a scope, hence my comment about zero phase difference between U, V, and W in the original post. I especially found your "idle mode" comment interesting, and zero line-to-line voltage is indeed what I measure. Furthermore, I measured

Vac(rms) = 242V (between L1 and L2, mains input)
Vdc(U, V, W) = -139V (between each output and Earth ground, or chassis. Note the negative voltage.)

I'm not sure I understand the DC relationship, if any, between the output lines and Earth ground. Is this what you meant? If so, it's about 57% of the input, which I interpret as either a suspect rectifier bridge (and therefore DC bus voltage) or, more likely, 1) a duty cycle which is not 50%, or 2) something else.

Please feel free to correct me or fill in any blanks, but know that I haven't done my homework here and do not expect to be spoon-fed this topic. In fact, I'm pulling the plug on this project and will try to find something a bit newer. I have enough junk (or 'gems' as I call them on the way into the house) in the pipeline as it is waiting to be repaired or whatever--a situation I'm sure many in forums like these can identify with.

jim,

Thank you for that common-sense suggestion. We don't have those bulbs around the house anymore, unfortunately, so I couldn't try that.

Looks like the inverter is not running.

I assume one of your L1/L2 is a neutral, ie single phase input, rather than a line to line.

In which case the voltage you are seeing is the inverter internal power ground to chassis potential with the output stages are not switching or active. If you checked with a scope, it would be abundantly obvious if they were switching (xxkHz square wave with ~340V p-p).

So all the values you measure make sense for a single phase inverter with a powered off output stage, your DC meter would sort of average the half wave voltage seen between the internal ground/negative and chassis/earth due to the rectifier.

 

[essenmein]

Oh the dc relationship I was talking about would be between internal power ground (ie the negative of the DC link capacitors) and the phases, now that I think about it at 50% duty (zero modulation), the three phase outputs would be mid point of the DC link, which, on average at least, would result in 0V between UVW and Earth ground.

 

[gnurf]

Looks like the inverter is not running.

I assume one of your L1/L2 is a neutral, ie single phase input, rather than a line to line.

In which case the voltage you are seeing is the inverter internal power ground to chassis potential with the output stages are not switching or active. If you checked with a scope, it would be abundantly obvious if they were switching (xxkHz square wave with ~340V p-p).

So all the values you measure make sense for a single phase inverter with a powered off output stage, your DC meter would sort of average the half wave voltage seen between the internal ground/negative and chassis/earth due to the rectifier.

No, neither L1 nor L2 is a neutral--I live in an area with an IT system [1] and so our domestic outlets have L1 and L2 in addition to (safety) earth/ground. When I looked at the VFD output on a scope prior to the original post, the U, V, and W waveforms (referenced to mains earth) were in-phase 50Hz sinusoidals with a slight dink at one point and an amplitude of roughly half the input or thereabouts. It looked nothing like e.g. the waveform in [2]. I should probably have included this information in the OP, sorry about that.

[1] https://en.wikipedia.org/wiki/Earthing_system#IT_network
[2] https://3l4sbp4ao2771ln0f54chhvm-wp...tent/uploads/2015/02/dc-voltage-sine-wave.jpg

 

[essenmein]

Weird, never heard of this IT system before... Learned something new!

Testing with that kind of system I imagine demands an isolating transformer, well testing rectified mains equipment in general does anyway.

Your link [2] is for multilevel inverter (usually reserved for proper sized equipment), but basically the idea is the same, it would be multi kHz square waves, not 50/60Hz sine waves if the thing was running.