Odd Article in ME 4559 28April - A Variable Frequency Drive

I get my ME from the local newsagent in Sydney, so always a bit late.

The referenced article was eagerly read as I spent my career in heavy variable speed drives, initially servicing thyristor drives and finishing with applying and using VFDs.

But, the referenced article is either:
- a spoof initiated by the editor on the reader,
- a spoof initiated by an author on the editor, and thus also on the reader,
- or an abominably written document mauled beyond recognition by the editorial and style conformance systems.

Where can I find a discussion on this article ? - there must be something out there by now.

I checked the date, but it is not April the 1st.
A puzzled John.

I'm not sure quite what you are getting at, but he author seems to have created a hybrid between a BLDC and a true inverter.

I think he has mis-analysed the reason he gets more torque at low speed but the principle of varying the voltage:frequency ratio is used to provide torque boost at low speeds by many inverters.

Neil

I can't find it now, but I recall it. I also recall a lecture at Sandown Park a few years ago, regarding turning a car alternator into a variable speed motor. I think it used a power oscillator, but this isn't my field.

Geoff

Neil,

A cursory review of the technology will reveal that the very best one can achieve is 100 % rated normal speed torque down to standstill with very sophisticated microprocessor control of a standard induction motor.

The established laws and general theories of physics, thermodynamics and electrical engineering demonstrate that torque multiplication of an induction motor at low speed is a nonsense.

The ME does no service to the hobby in promulgating such a monstrous lie. There may now be numerous model engineers out there that have read the opening paragraphs of this article and, lacking critical knowledge in this area, believe it because it was published in the ME.

The article makes no effort to demonstrate its hypothesis of torque multiplication, the final paragraph claiming witness of successful operation smacks of a snake-oil salesman without any actual test results.

I suggest you (ME) refer the article to a competent electrical expert for review.

Not only puzzled, but now disgusted,

John.

OK, found it now.

Bear in mind that conventional electric motors lose torque as voltage is reduced, leading to thyristor control as mentioned, which cuts off more or less of the AC mains waveform reducing the speed but not the torque by reducing the time that AC power is 'ON' during half the 50 Hz supply waveform. Modern versions use Triacs, to control both halves of the AC waveform. I built such a device for my single phase power drill. I now have a pulse width modulated speed control to control a Gauge 1 locomotive, running at a nominal 12 volts, which works the same way, with DC.

I find no problem with the article by Mr Gawthorpe, except that he states that no flip-flops or gates are used in his device. I would say that there are, but built from discrete semiconductors, not integrated circuits. It doesn't help that a mains motor is rewired for 12 volts, and an inverter built, this just confuses the matter. An inverter would be used with a AC mains motor, not with a 12 volt car alternator.

Geoff

To a first approximation in an induction motor torque is proportional to current. So if you double the current you get double the torque. Below base speed less voltage is needed to drive a given current. That's why the simple VFDs use a V/f curve to reduce applied voltage as the speed decreases. The simple equation tends to go wrong at very low frequencies where the resistance of the windings become significant. Some VFDs allow you to boost current at startup, and hence low frequency, to give extra startup torque (above 100%) for difficult loads.

Of course there may be I²R heating problems with high currents at low speeds, if a cooling fan is tied to motor speed. But that's an implementation problem, not one of basic physics.

To control an induction motor at zero, or very low, speed you use vector control. Basically this is a transform from a rotating three phase space to a two dimensional time invariant space. The two dimensions are normally labelled d and q, direct and quadrature. This sort of control is often implemented on a DSP, as it has an architecture suited to vector, ie, MAC, operations.

None of the above means that I necessarily agree with the referenced article.

Andrew

MAC = multiply-accumulate, a fundamental operation in a wide range of signal processing algorithms

Thanks John, your post did me a huge favour. During a quick search for 4559 I found my missing Lathe Manual! Now I'm having a glass of wine to celebrate. Of course there's no sign whatever of the offending magazine...

Dave

Blimey, before 6pm? That's serious!

Andrew

Got carried away: after drinking the wine I moved on to anti-freeze. I wish I hadn't finished off on superstore turps though. I enjoyed the aroma of wild benzene with a hint of goat urine but overall it was too round-shouldered for this connoisseur's palette.

May I politely suggest you catch up on modern VFD technology.

what-when-how.com/motors-and-drives/torque-speed-characteristics-constant-vf-operation-motors-and-drives/

If you would like, I can email you a copy of the manual for my IMO CUB manual which explains how to configure toque boost (of about 6  - 7% at starting).

Neil

Edited By Neil Wyatt on 10/07/2017 20:40:48

Neil,

I reviewed the link you supplied and, in Fig 8.5, the areas of ac motor operation are denoted as "Constant Torque Region" up to base speed. The regions above base speed are not the subject of this discussion. The diagram is clear, the torque at base speed is essentially maintained constant down to very low speeds.

If you drive an automobile in top gear at speed on the level and are confronted by a steep hill, stopping to tinker with the carburetor to get 7% increase in torque will not get you over the hill. Hill climbing requires a gearbox.

The last manual car I owned had a first gear 4:1 compared to topgear at 1:1, producing 4 times the torque at the wheels with a corresponding speed reduction to 1/4. The torque boost by the gearbox is 300%.

The author of the article used the term "mechanical gearbox" and "electric motor exchange revs for torque" in the opening introduction and in the closing paragraphs, claiming he had demonstrated electronic equivalence of a gearbox, without offering any actual numerical values.

My central concern is the mechanical gearbox analogy, and the danger of leaving the wrong impression in the minds of our fellow model engineers.

An ordinary V/f control, where the applied voltage is varied with the frequency, will work perfectly Ok for hobby lathes, mills and locos, with perhaps a 10% droop in speed as the load is increased. Industrial applications requiring accurate speed setting and quick response to varying load need the vector control. I have simple V/f inverters on my lathe, mill and surface grinder.

Of interest, as the cost of microprocessors comes down, the V/f technology may soon become superseded by the universal use of vector control.

I applaud any amateur playing around with electronics to see what can be done, we need more of them. If the author should read this, I am happy to discuss via email and even build an equivalent unit here in Sydney for test.

I thank the other contributors to this discussion.

John

I'm not defending (or condemning) the article, which I had overlooked until I read your first post.

My response to your comment was aimed at joining a civilised discussion, and also in the hope that you might be a bit clearer about what you were criticising.

In your response to that, you said "A cursory review of the technology will reveal that the very best one can achieve is 100 % rated normal speed torque down to standstill with very sophisticated microprocessor control of a standard induction motor."

And

"The ME does no service to the hobby in promulgating such a monstrous lie."

Your first statement is demonstrably wrong, simply because a VFD can apply over-voltage a stationary or slow-running motor to deliver above nominal torque, and tapering this over-driving down as speed rises. I have a setup on my lathe able to deliver 107.1% of rated torque from a standstill.

The second statement that ME is promulgating a monstrous lie, is therefore inaccurate, so it's rather unfair to express 'disgust' at my response.

Claiming '7%' isn't torque boost because it wouldn't work in a car is nonsense, actually your analogy is even worse.

Electric cars have far more starting torque (without gearboxes) than conventional cars. The gearbox is a kludge because IC engines, unlike steam and electric power, have horrible lumpy torque curves.

In practical terms you don't lose as much performance with a VFD lathe compared to a gearbox as you expect, as the amount of torque on a convention gear/belt drive bench lathe at low speeds is usually far more than it is practical to make use of.

Neil

Edit - Just to add, most inverters used on hobby lathes already have vector control, my IMO Cub does albeit what they describe as 'simplified vector control'.

Edit - my mistake, 7.1% is the default value, you can program low speed torque boost on the IMO Cub in the range 0-20%.

Edited By Neil Wyatt on 11/07/2017 15:50:14

Edited By Neil Wyatt on 11/07/2017 16:47:51

The discussion, kicked off on another forum, prompted me to re-read the article. My initial comment would be that the author was confused between "Converters" and "Inverters" (aka VFDs) as applied to driving 3-phase synchronous motors. ​I also found it hard to understand the points he was trying to make, However, he obviously put a lot of work into rewinding the motors for 12v operation and building the electronics. Shame his conclusions were muddled.

As I have fitted a number of VFDs to Lathes and Mills and fully understand the principals of V/f and Vector Control - as well as being able to design and build such units if necessary - the article was not of much interest to me and was, therefore, only scan read back in April. I was, however, intrigued to find out why electrosteam had posted the comment about it being a spoof.

​Neil's comments above would be mine, only he got there first.

Must try harder then.

When I machined the flywheels for my traction engines I had quite a lot (~3/8" to take off the diameter. If I remember correctly I was running at 85rpm, with 50 thou DOC and 10 thou feed per rev. Given the flywheel is about 16" diameter that gives 2.1in³/min. Assuming removal of about 1 in³/min per horsepower that's about 2/3 of what the lathe could manage, albeit with not the most rigid of setups. The swarf was flying off:

Personally I'd miss the power at low speeds.

Andrew

Andrew makes the case eloquently.

When I started this thread, I thought the article would end up being a spoof, but, from the responses so far I am forced to recognize it as the third option in my original post. The article is poorly constructed, impossible to understand and fails to make its case.

Along the way we have seen strident claims that VFDs can supply torque boost with apparent acceptance that a 7% boost in an area of motor operation that cannot be used for continuous operation is in some way equivalent to a gearbox that provides continuous delivery of torque boosts 100's of times greater. My geared head light industrial lathe has a speed range of 37:1 or 3600% torque boost (less due to friction losses).

Reasoned argument fails me.

To anyone reading this thread in the future, read the literature, check the data published for motors and VFD's and make a rational decision.

To aid anyone considering a VFD for a lathe, I use a simple VFD on my lathe using predominantly two speed settings, 1100 and 348 RPM. With a 3:1 turn down ratio I easily handle all my light hobby requirements. Other speeds to 3000 and 80 RPM are available when the cut or load requires it.

John

And he did so without using words like lie, spoof, abominably, mauled, monstrous and strident. Pejorative language doesn't fit well with a claim to be making a 'reasoned argument'.

Dave

The problem is you made your case with statements like "A cursory review of the technology will reveal that the very best one can achieve is 100 % rated normal speed torque down to standstill with very sophisticated microprocessor control of a standard induction motor."

Whether or not the actual figure is 101% or 3600% is irrelevant. If you make sweeping statements that are demonstrably untrue whilst rubbishing other people's work, expect to be challenged.

Continually moving the goalposts with extreme comparisons in an effort to undermine other people's statements isn't helpful either.

Neil

Neil,

As an engineer, I would have thought the difference between 101% and 3600% was significant.

The language I used to describe the article was simply my opinion, you may have a different opinion.

I haven't moved the goalpost at all.

The subject of Vfd torque boost was introduced by you.

Comparisons introduced were analogies to demonstrate the effects of torque boost versus gearbox multiplication.

I did not rubbish the work done, I congratulated the author for doing it.

The problem is the article as published, and the conclusions reached within it.

The author has probably over enthusiastically described his endeavours, misstated the engineering use of gearboxes and confused a demonstration of operation with proof of performance.

The editorial team should have picked up the poor writing and the erroneous statements before publication.

Did this article get any peer review ?

If not, then I respectfully suggest that the editorial team consider this in future.

We are all being inundated with wild claims in various media, and we want the ME and MEW to provide a solid view of the hobby as it is, and how it could develop.

John

So this isn't rubbishing the article? :

" ...the referenced article is either:
- a spoof initiated by the editor on the reader,
- a spoof initiated by an author on the editor, and thus also on the reader,
- or an abominably written document mauled beyond recognition by the editorial and style conformance systems."

[ Just trying to understand where the goalposts are in your world ].

Edited By Bandersnatch on 24/07/2017 00:55:30