1 to 3 phase inverter

[robgraham]

I bought one of these some time back for a project which never materialised, but am now experimenting with it to drive a 3 phase powered wood turning lathe (ex school hence the 3 ph motor probably connected in star).

I am looking for some input from someone who has used one of these before and done some measurements.

I didn't realise to begin with that it converted 240v single phase to

240v 3 phase and that the motor should be connected in delta format to get full power - I thought it would have been 440V 3 ph and continued to think so as my DVM was registering 600v ph tp ph. I put the 'extra' volts down to DVM anomaly ! Anyhow a bit more reading and then experimenting with an old AVO and a more up to date DVM gave 320v ph to ph.

This is a bit nearer 240v but far enough off to puzzle me somewhat - the fact that both the AVO and the modern DVM read the same would suggest that this is a true RMS reading but at some 30% above 240v, that does seem a bit odd, and should I be concerned about it with respect to the winding insulation.

Thanks Rob

[fred]

I probably have this wrong but IIRC, for 240v 3 phase, if you had a neutral you would measure 240 between any phase and neutral, but ~350 between phases.

[fred]

In article , fred writes

240V phase to neutral is 415V between phases, sqrt3 being the multiplying factor.

[Bob Minchin]

Rob, You should have no need to be concerned. The inverter you have does not put out any higher voltage than you put into it. I rectifies the mains directly to peak mains voltage and uses this to synthesize three phase waveforms by pulse modulation. The resultant waveform is tolerably smoothed to be sinusoidal using the inductance of the motor. Your Avo is a peak reading device with a weighting factor to read RMS assuming a sine wave. Any other waveform will give erroneous readings. Most DVM/DMMs work much the same. Top of the range models claim to offer true RMS via signal processing. I suspect they could also get it a bit wrong when pulses are present.

Out of curiosity, I have just made some quick measurements on a Siemens inverter running a 3HP motor in my workshop. Nominal single phase mains in and delta connected motor on the output. Voltages measured from phase to phase. My cheapy DMM max range 700 vac indicates off scale. My Avo 8 indicated about 700 v on 1000vac range My Tektronix true RMS DMM reads between 235 and 240 vac

I have run many motors on these converters with no signs of insulation breakdown or inverter failure. The worst problem was a stress fracture in the lead of the crystal on a control micro. Even that failed safe and a replacement part rendered the unit working again.

HTH

Bob

[Bob Minchin]

True but inverters like this produce 240 phase to phase. Equivalent to about 140v phase to neutral if there was such a thing.

Bob

[dave]

I don't pretend to know what one of these beasties is, but are we talking sine waves here? If not then sqrt3 does prob not apply. ie waveshape accounts for the difference?

[robgraham]

Many thanks, Bob - that was the answer I was hoping someone would come up with. Taking your point about peak rectification, that would give

340v pk, so although the logic isn't perfect, that makes 320v on my meters understandable.

Rob

[Bob Minchin]

Yes. The waveform is the core of the perceived problem. All Ac voltmeters aim to give correct readings for sinusoidal waveforms at typical mains frequencies but use differing methods to achieve this. These methods lead to errors when the waveform is non sinusoidal or has a higher frequency.

Bob

[Bob Minchin]

Rob, My point on p-p rectification was that with internal supply rails no greater than input mains p-p values, the inverter could only generate voltage pulse magnitudes up to the supply rails. These pulses are quite likely to upset readings on many meters however which is probably what you have found. I suspect that my cheapy DMM gets it so wrong because the high frequency components of the waveform is not being attenuated in the meter due to stray capacitance in its input circuitry. To be fair it was free promotion with an order from Rapid a few years back. Bob

[John Evans]

The moving coil AVO is an Average reading device, not peak.

[Bob Minchin]

I'm not totally sure but, I suspect it is a mix of the two actually. Peak detecting due to the diode and averaged/integrated by the inertia of the mechanics of the movement. It certainly does not like the pulse output of my inverter.

Bob

[John Evans]

The diode(s) conduct when the input is the appropiate polarity presenting half of the waveform to the meter movement. If it is a full wave - or bridge - circuit then both halves of the waveform are presented. There is no peak action as there is nothing stopping the diode(s) conducting for the full half cycle - ignoring the small offset. The mechanical inertia then averages out the pulse levels.

John

[Bob Minchin]

John, I don't disagreed with the intended operation but something, presumably the waveform or possibly high frequency components present, causes my avo to read too high on an inverter. It could even be due to capacitance which is irrelevant at 50hz but no so at inverter pulse frequencies. One day I'll get a scope out into the workshop to try and understand more fully what is going on. Fortunately my true RMS Tek DMM seems to be telling the the truth. Regards Bob