Using VFDs on old motors

[Andrew SkinnerParticipant @andrewskinner94774]

As mentioned in a recent thread, the ‘spiky’ output of VFDs can damage older motors whose insulation was never designed for them. I wonder what cheap and practical steps you can take to mitigate this, given that dedicated sine wave filters are quite expensive. I think most of us quietly ignore the problem and hope for the best.

Apparently, the motor windings themselves act as quite good low-pass filters for the current, and that it’s actually the voltage rise time that causes the problem. The high dv/dt pulses result in reflected waves sloshing back and forth due to impedance mismatch. These waves add together in the cable between VFD and motor, leading to high voltage transients that damage old motor insulation.

^^ I’m sure this is a poor summary, and also sure that you could write a whole monograph on the subject. But, down to brass tacks, what practical steps could we take in a hobby workshop with several old motors on VFDs?

1. In-line chokes, such as old fluorescent light ballasts, providing they could take the rated current?

2. Keep the VFD-motor cable short, and manipulate the VFD’s carrier frequency based on this length?

3. MOVs, or something, to clip the transients? I looked into the surge protection devices (SPDs) now becoming more readily available because they’re mentioned in the new UK wiring regs, but they are apparently not sensitive enough.

4. Some sort of snubber network at the motor terminals?

[Pete RimmerParticipant @peterimmer30576]

Here’s my solution. I just put a VFD on it and stop worrying. I have VFD’s on two pre-70’s gear hobbers, 1960 milling machine, Mk1 Clarkson grinder, Mg9 cylindrical grinder and several machines no longer owned. So far I’ve had exactly no motors fail on any of them.

My 1942 lathe which is on it’s original DC motor uses a parker electronic drive (like a VFD but for DC motors) and I have spend the last 12 years switching the power on and off with the drive run-enabled – something else I was assured would kill the drive in short order, and that’s still going strong too. I bought a spare drive just in case and it’s still in it’s box.

I’m not saying that it’s not a problem, but it sure as heck isn’t so much of one as the people who keep parroting about it would have you believe.

[renardiere7Participant @renardiere7]

I’m with Pete, same experience here. Probably jinxed myself now!

[Andrew TinsleyParticipant @andrewtinsley63637]

In a previous thread, the Steinmetz connection was referred to as a bodge. Looking at the original Steinmetz papers, I think that this is somewhat unfair! At least the Steinmetz solution produces sinusoidal voltages which are what motors were intended to run on.

The hugely popular VFD solution, utilises a high frequency switching solution which is far from the ideal sinusoidal voltage. An oscilloscope shot of the current “waveform” is just a square block of high frequency pulses. I would venture to suggest that this is far more of a bodge than that of the Steinmetz connection!

However I am reluctant to push this further as the topic has been well aired recently.

As to running older motors on VFDs, I am not aware of any methods of reducing the high frequency voltage spikes associated with VFD outputs. Running motors from the 1980s such as the Crompton Parkinson and Brook Gryphon, is probably relatively safe in an amateur workshop setting (relatively low usage). Motors from say the 1960’s and earlier are the ones that are problematical. I have some pre war kit fitted with integral motors that cannot be easily replaced. I would not dream of running them on a VFD. Have you seen the prices that rewinds command?

If you do your homework, you will see that some motors are advertised as suitable for VFD operation. Mainly TEFC motors. I am not sure if this is a marketing ploy, but definitely there are some motors that are not suitable for VFD use!

Sorry if this is not more useful,

Andrew.

PS If you have an older motor that you intend to run on a VFD, I would urge you to check the insulation resistance  with a Megger, preferably with a 1000V output, but 500 V is better than none. If the motor has been out of use for some considerable time or is new to you, then put it in a warm environment (airing cupboard if you have one, for a week or two). If the insulation resistance test is not looking too good, this treatment can dramatically improve the insulation resistance. Always run the motor with an RCD protected supply. If the leakage current goes up then the RCD will cut and save the motor. Without this protection the motor will probably burn out a set of coils.

PPS I note a couple of contributors say just fit a VFD and forget it. I certainly have some sympathy with that view, motors of say up to 60 years old are relatively immune and failures are unlikely. The older motors are a different kettle of fish. I have had a couple of failures with old BTH motors on a VFD, that is why I used the Steinmetz connection for many years.

 

[Andrew SkinnerParticipant @andrewskinner94774]

Pete – yes, that’s what I do at the moment. I’ve got a few machines on a rotary, mainly woodworking, a coolant pump on a Steinmetz and Tom Senior mill with a new, inverter-duty TEC motor on a VFD.

But, the Boxford AUD/VSL, Beaver VBRP and an old morticer have their original motors on VFDs. These motors are either old imperial or custom-fitted ones so would be a nuisance to replace, and uneconomical to rewind. If I could knock up a simple snubber circuit for them, that would be nice. I could even get a run of PCBs made by one of those online companies, and sell ‘peace of mind’ kits for a reasonable fee!

Andrew I do megger newly-acquired motors, mainly with a Fluke MFT but I have a collection of old meggers, including a wind-up Bakelite one. I quite like them, for some reason. Usually, I change the bearings and while in there, pull out the star point.

[duncan webster 1Participant @duncanwebster1]

When I converter a Ferret grinder using Steinmetz, the motor manufacturer strongly advised against using a VFD. I didn’t pursue it as I wasn’t intending to, but the motor man had no axe to grind, so is worth listening to. Of course loads of people will get away with it, but a few won’t.

The grinder worked a treat by the way, slight loss of power didn’t matter as the motor was well up to light domestic use.

[Pete RimmerParticipant @peterimmer30576]

I have a Steinmetz connection on my Christen drill sharpener, and had one on my Union Tool & Cutter grinder and both worked just fine. I did once have a Mk1 Clarkson grinder which was a bit slow to spin up but otherwise worked just fine. I just choose a capacitor that gives the best 3rd-phase voltage and doesn’t make the motor run too hot after half an hour of use.

I once put together a control box for a lady who was building a ring roller powered by a 3 phase geared motor using a Steinmetz setup. It used a timed relay to add in an extra capacitor for 1 second to give a startup boost. It was required because if you stopped the roller under load it would struggle to start off again. Worked great.

Steinmetz is used in industry too. I had to remove a rising traffic barrier from a building access road one time. It featured a 415v 3 phase motor (wired in star), a capacitor and a 2-way switch (for manual operation, plus control stuff for automatic). It was very simple, they put the capacitor across 2 of the phases, neutral to the remaining phase. The switch applied 240v single phase to one end of the capacitor for up and the other end for down.

[Robert Atkinson 2Participant @robertatkinson2]

To protect your older motor against the small possibility of damage due the the PWM output connect a inductor (aka choke, reactor) in series with each of the output connections. The manual should give information on suitable values..
It’s not just VFDs that produce spikes. The 3 phase mains has plenty of it’s own and a VFD isolates the motor from those.
It’s not much of a problem in reality. Do an insulation test on the motor and if it passes just use the VFD.

Rant mode on. Why whenever a question about VFDs come up do some members just have to hijack the thread to discuss Steinmetz? You know who you are.

Robert G8RPI.

[Andrew SkinnerParticipant @andrewskinner94774]

Thanks Robert. Elsewhere, it’s been suggested that I use three inline chokes together with a delta of pulse-rated capacitors at the VFD output, as an LC filter.

The VFD/old motor problem is described here:

https://www.mtecorp.com/wp-content/uploads/SP-006-E_Web.pdf

So, it is a thing, maybe not significant for hobby use, but if I can whip up something simple and cheap, perhaps worth a go if it saves a custom or imperial-fit motor, and the attendant hassle and cost of replacing it?

At the moment, I’m not sure what values to use. AI suggests 5mH and 0.47uF.

[Robert Atkinson 2Participant @robertatkinson2]

You need to know the PWM frequency to decide on a inductor value. Generally capacitors are not required and can give issues due to resonances.
What make and model is the VFD. The manual of the better makes will give guidance on output inductors.

Robert..

[Andrew TinsleyParticipant @andrewtinsley63637]

Again I am most amused that Robert states there are plenty of voltage spikes on the mains. This is quite true, but nothing like the quantity of spikes that a VDF produces. Seems that selective use of information is at work here.

I am not anti VFD or pro Steinmetz. BOTH systems work well, even if certain individuals have bees in bonnets about the respective systems.

Andrew.

[not done it yetParticipant @notdoneityet]

I’m like most above.  Always converted to delta and used a VFD or simply driven the motor on star.  Never considered any bodge, which provides only one frequency (motor speed).  I do have a Transwave converter but don’t use it and will eventually (as soon as possible) get rid of it!

Capacitors can give up with long-term unused VFDs, I believe – but that is mostly likely of cheap chinese origin, I expect.

[Andrew SkinnerParticipant @andrewskinner94774]

The carrier frequency on the LS M100 can be adjusted from 1 to 15 kHz, default 3. I wasn’t sure why you’d need to change it, so left it at factory. I assume you’d want it as low as possible, without being annoying to listen to, if trying to reduce voltage spikes. But is whatever I choose the problem frequency, or is the motor damage happening at higher harmonics?

I guess the choke/cap values can be a broad range, given the desired frequencies will be under 100 Hz, and the problem ones an order of magnitude higher.

[John HaineParticipant @johnhaine32865]

Voltage spikes shouldn’t be much affected by switching frequency. Thinking in terms of harmonics is not very useful for this kind of problem, it’s caused by current or voltage rate of change. As I think Robert said above some small inductors in series with the motor connections  should be enough and there may be a selection guide in the vfd documentation. I’ve had an old Brook motor, probably from early 60s, on my mill with a vfd and it has worked perfectly with a direct connection for over 10 years, though I have had to replace the inverter because of a control fault.

[Dave HalfordParticipant @davehalford22513]

Old motors suffer from failed insulation anyway, even single phase ones like mine, who can tell what causes it?

[Andrew SkinnerParticipant @andrewskinner94774]

https://industrialmonitordirect.com/blogs/knowledgebase/eaton-reflected-wave-trap-for-vfd-motor-protection

This link suggests that switching frequency can affect reflected waves and voltage doubling at the motor terminals. The impedance mismatch from VFD -> cable -> motor is the root cause. However, the problem seems to exist only for longer cables, >15m, which are unlikely in the hobby workshop.

 

So, I think we’re left with the sharp rise time of the voltage, perhaps with some ringing, which could be damped either with the L(C) filter at the VFD, or an RC snubber at the motor terminals. I have a couple of scopes, but only ordinary probes, so I’d be interested if a better-equipped member could try to capture the problem and show which measures, if any, solve it.

[Julie AnnParticipant @julieann]

The incoming spikes on the mains are generally of short duration (microseconds) and high voltage (several kV) and are generated by substation switching or external interference. They are dealt with using MOVs or transzorbs.

A VFD output does not inherently produce spikes. The output is a PWM signal where the pulse width changes on a cycle by cycle basis. The fundamental of the PWM is a sine wave. The pulse edges are slow by electronics standards – around 100ns to a microsecond. The carrier frequency is a trade off between harmonic distortion of the sine wave (low frequency and hence low resolution) and low distortion of the sine wave but with harmonics of the switching frequency spreading higher in frequency. Generally use the lowest frequency possible.

A H-bridge has diodes across each switching device that limit the output voltage to DC link voltage plus a diode drop and 0V minus a diode drop. These diodes also provide a path for the circulating current in the motor winding when the associated switch is off. The motor winding inductance, and the inertia of the machine tend to mean that the motor essentially sees only the fundamental of the PWM signal. Of course the harmonics need to go somewhere and it is this that may cause problems with older motors. It is not correct to call these harmonics spikes.

Once the PWM signals leave the VFD then the external circuit can cause issues with resonances and di/dt effects. Hence the recommendation to keep output cables short if at all possible. Adding filters may well make the problem worse; experimentation is usually needed.

Sine wave filters for VFDs are expensive as the inductors need to be fairly high inductance but also need to carry the full load current without the core saturating or the coil overheating. This means large cores and thick wire; equals expensive.

Julie

[Robert Atkinson 2Participant @robertatkinson2]

Julie beat me to it. I was trying to stay away from the details.
I’m biting my tongue on a couple of comments….

Robert.

[Mark RandParticipant @markrand96270]

Julie has it concisely.

 

I’ve made my position clear in (too?) many posts on similar threads. But for information:-

1. If you are worried about the insulation condition of an old motor and want/need to run it with an inverter; These folk make and sell L-C filters. They even publish the technical details so you can brew your own if you can get your own inductors and capacitors.
2. The dates on the motors on my 3 phase machines fed from the central inverter are 1952 (Hardinge), 196? (Beaver milling machine), 3X1976 (J&S grinder),  19?? (Meteor drill grinder), 19?? (Brierley drill grinder), 198? Wolf pedestal grinder. 19?? (BCA Mk111 jig borer)
3. A significant, but not vital, advantage of a filter after an inverter is that it eliminates the carrier frequency whistle/whine from the motor windings.

Generally speaking, with the hours us folk tend to put on our machine tools, we aren’t going to see any insulation failures from partial discharge erosion in any motor/VFD combination we can dream up.

[Dave Halford]

On Dave Halford Said:

Old motors suffer from failed insulation anyway, even single phase ones like mine, who can tell what causes it?

Very true when you’re talking about the insulated leads. I’ve had motors which had perfect-looking wires in the control box that shed their crusty rubber coating the moment you touched them. The parvalux work-head motor in my MG9 was one such example.

I dont think I’ve seen a motor with disintegrating winding insulation though.

[Dave HalfordParticipant @davehalford22513]

3/4hp motor sucking 10a continuous and too hot to touch with the ambient at 5C and a working centrifugal switch and good plastic insulation.

[Pete Rimmer]

On Pete Rimmer Said:

On Dave Halford Said:

Old motors suffer from failed insulation anyway, even single phase ones like mine, who can tell what causes it?

Very true when you’re talking about the insulated leads. I’ve had motors which had perfect-looking wires in the control box that shed their crusty rubber coating the moment you touched them. The parvalux work-head motor in my MG9 was one such example.

I dont think I’ve seen a motor with disintegrating winding insulation though.

That is probably because they were “burnt out” before you looked.
The old enamel insulation was prone enougth to breaking down that failures supported at least one rewinding compny an any decent sized town. The modern insulations are much better and combined with high labour costs have put most of them out of business.

Robert.

[Author]

Posts