Motor Kv - the truth

[Myron Beaumont]

Peter

I see you've sined off so I won't go off at a tangent anymore . SO,

Doesn't a low KV figure equate to inefficient (insufficient) windings or should I say not realising the potential of loads of windings to utilise the power available or is my concept wrong & really we are talking in the realms of electronic gearboxes so to speak allied to a suitable ESC ? As you've probably gathered ,I havn't a clue what I'm talking about 'cos I'm more confused than ever

[Peter Beeney]

Myron,
I think I’ll wait and see if there are any other postings with any ideas on why the motor has the Kv that it does; plus how the motor torque and emf equate to the current consumed and speed. Then I shall be able to see if my guess as to how all this goes together has some semblance of credibility.
Coincidentally, at the strip yesterday my friend that operates very small ‘hurry up‘ type electric models had a small Komet, the Messerschmitt ME 163. This is powered by an inrunner, I have little other detail at the moment, other than that he casually mentioned that the motor speed would be in the region of around 40,000 rpm. I’m guessing here but the prop looked like it was about 5 by 6. If this is the case then 40,000 rpm is 227 mph. Even after allowing a 30% drag factor then it’s still 160 mph. If noise is any indicator of speed then that’s right on the button!
Anyway, the speed controller is a CACC, (Cheap As Chinese Chips) model, no name or indeed any other information, at least on the unit, just a very impressively large chunk of nicely fluted aluminium heatsink, anodised in a dark red matt finish, as it should be, and a few components wrapped in heatshrink plastic. At 40,000 rpm the motor is revolving 1.5 times every one thousandth of a second. If the ESC is switching a few times in every revolution then that is switching at.. well.. quick time!
Both the motor and the ESC seemed completely unfazed by this dramatic display of converting watts into adrenalin, but it will interesting to see just how long the battery, and indeed the model, last. It’s usually his ESC’s that go u/s first, just melting into a lump of ballast. On at least one occasion one exploded, departing through the side of the foamy model in a long shower of sparks and a plume of black smoke that descended vertically. Thus again proving Newton’s First Law Of Model Demolition remains firmly intact, at least.
From our point of view, also interesting is working out exactly what in going on in both the motor and the speed controller.
So now I’m still confused, but at high speed!     PB

[Myron Beaumont]

Something to do with"Spin"I think

[Peter Hepworth]

Slight change of subject peeps. I would like to know the relationship between length of leads from battery to ESC, and length of lead from ESC to motor as I am told motor performance is affected. we are getting conflicting reports. One report says that length of lead from ESC to motor is critical as the impedance is altered and as I recall from my engineering days a loonngg time ago says that this could be so because frequency is related to impedance plus a 3 phase signal going to motor. If memory serves a high impedance output is affected by length, low impedance not so eg., length of leads to HIz mics is restricted whereas low impedance mics are not affected within reason. Hopefully someone can give us the defnitive answer.

 

Cheers All

Edited By David Ashby - RCME Administrator on 11/06/2010 11:11:10

[Chris Bott - Moderator]

Peter the conventional wisdom is that it is far better to add length to the wires form esc to motor, than to add length to the wires from battery to ESC. 

[Klippy]

Hi Peter, when you put the wiring in the model try to keep the the motor wires as far apart as possible, do not tape them together! Parallel wires in close proximity will create additional capacitance, coupled with the Inductance of the motor could give rise to unforeseen circumstances.

[Peter Hepworth]

thanks for the info on motor leads etc., It gets a lot more complicated the more you learn.

 

thanks again

 Peter

 

 

[empeabee]

I tried a post, having logged in, but when I said send it said Login & lost the post,

Is this normal operation?

Mike

 

[Sparks]

Tony, I'm afraid I can't agree with you about separating the motor leads. The capacitance between wires will be minimal (a few pF) so would have negligible effect on ESC capacitive loading.

 

In fact, it would be better to twist the wires together to reduce radiated emissions from the motor/ESC wiring.

 

Richard 

 

 

[Peter Beeney]

Tony, Can I say this with the greatest respect here, please, and I certainly do not mean to oppose or dispute your post in any way, but are you talking about just the three motor wires or the motor wires plus other electrical wires? If this is the motor wires plus others (servo?) then I’m in agreement, keep everything separated if possible but I think I would tend keep the three motor wires close together, to try and avoid any possible interference being emitted. I think my main concern would be, if I were extending the motor leads, to bump up the size a bit if possible. That’s the cross-sectional area. I would consider that we need to keep the total impedance as low as possible. Lets say the resistance of the extended part of the wires is maybe 0.04 ohm, and if our motor is consuming 50 amps, V = I times R, then we would drop a further 2V across our leads. Maximum power transfer takes place when the resistance of the load equals that of the source; the resistance of the motor is very low, as is that of the battery and the speed controller, we need to keep the bits in between low resistance too!
I’m wondering that if when Peter is talking about the impedance varying with frequency he’s thinking abut resonant circuits, these are circuits which consist of capacitance and inductance subjected to a frequency such that causes the circuit to become a high or low impedance.

I have a little more info on the Komet. It’s motivation is a 4S, 2.2Ah, 40C lump of polymer and lithium ions. The prop size is 4.75 by 4.75. It weighs just on the 2lb mark. When running the current quickly settles to about 70 amps; thus we’re estimating the watts/lb to be in the region of 450 to 500. However, I believe small props to be notoriously inefficient, so it might not be quite what it seems; plus the fact that there are also losses across the ESC and motor. A 4Ah pack will fit in, so at that point I’m wondering if it really will be a battle as to who can last the longest per flight, the pilot or the model!
I’ve yet to clock the prop speed, but at 40000 rpm the theoretical speed is 180 mph. It would be interesting to get the actual speed on a radar gun and try and calculate the efficiency. All that remains is for me to now attempt to ponder on the reasons why the motor turns at this speed.

How long before this sort of performance is the norm? PB

[Klippy]

No apologies needed guys, this is a Forum, a platform for testing ideas, whether right or wrong they all have a place in improving our understanding of what's going on inside these little marvels.

Thanks to Richard I've revised my position on the motor wires. Twisting the motor wires together would be better, perhaps even put a sheath of cooking foil around?

As regards servo wires et al, given the relatively low frequencies and currents at which they operate (relative to the motor) I don't think we need worry, other than for neatness and security, the last thing you need is a loop of wires catching on a servo arm or Z bend!

[Sparks]

Tony, if you want some light (!) reading on good EMC practice have a look at DEF-STAN 59-411  DEF STAN 59-411 Part 5    (it's a free download)

 

It's intended for Military applications, which have strict EMC requirements, but the recommendations are equally applicable to any EMC sensitive application.  

 

For cable shielding/twisting recommendations see section 14.6 page 137

 

regards, Richard

 

 

[Klippy]

That's a very large document! However page 141 would appear to recommend twisted, or twisted screened, depending on which part of the table you look at. I think in future I'm going to start screening my twisted motor wires. Not that I seem to have any problems, but as we know they can spring up at any time. And probably will now I've said ti!

Isn't it amazing the things you find on the internet? I envy todays youngsters, they have access to so much first rate imformation.

[Brian Parker]

Twisting three cables together will only be beneficial (in respect to EMI) if they are in balance ie. the sum of the three voltages is at 0volts at any instantaneous measurement.

[Sparks]

Brian, you will see from my earlier posts that only two phases are energised simultaneously (third phase is feedback) so twisting the three wires together will, in effect, provide a 'twisted pair' for each commutation step.

 

Richard

[Brian Parker]

Sparks,
 Just re-read through the thread.

I note your point, but, if twisted, would that not the un-energised wire produce a couple in the phased energised wires as it moved from energised to un-energised throughout the sequence?

Better to leave untwisted I would have thought.

[Peter Beeney]

Please forgive me for asking, but are any emissions from the motor leads really going to be that significant? I’ve done some practical bench tests, admittedly a little haphazardly, trying to get the motor to cause trouble, but I’ve never been able to pick up any interference from anything very much. I’m slightly disappointed, really. I was looking at it from the other way round, I just wondered how good is the rejection of the receiver was, and I assumed the motor/ESC could give rise to spurious rf transmissions. I don’t have any means of knowing if in fact there were any such interfering emissions or not, anyway, perhaps there weren’t any, and so the rx wasn’t tested, difficult to say. From practical experience, I’ve have thought that there was a better chance of picking up induced pulses etc. from long paralleled leads. This can happen, but perhaps only under more unusual circumstances. One good source of interference was the brushed motor commutator, when there was some arcing and sparking going on; this was usually very good at generating local, short range, rf rubbish. Particularly when the suppression capacitors on the motors starting playing up, I’ve seen a few of those; one broken leg was the first thing to look for. But I’m counting these motors out now, I’m not sure that I’ll be going back to the brushed variety!

Regarding the problem of increasing the length of the battery leads, I once had a hotliner with an Aveox motor and ESC. The ESC had a large size capacitor on it’s battery input, I assume to smooth out the vigourous pulses resulting from the current being constantly switched on and off. This capacitor would get extremely hot, even with the standard length of battery lead, perhaps pointing to a combination of inadequate cooling, (difficult to improve), and the fact that the ESC generally was also being pushed to it’s utmost limit. However, it was the recommended unit, and it never faltered, ever. If by increasing the lead length we increase the resistance, does this make the capacitor work even harder, to the point where it can’t cope, and the components on the ESC are affected? If so, and it’s imperative to increase the battery lead length, would this notion be worth considering? If we double the unit length of wire we double the resistance, but if we double the cross sectional area we halve the resistance. By trimming the existing wire length back as much as possible, and replacing with a longer but heavier gauge wire we could at least keep the total resistance more or less the same. If the self inductance of the wire is a cause of the problem, and increasing the wire length increases the inductance also, it may be possible to stitch some more capacitance onto the ESC. Although I have a feeling that increasing the cross sectional area reduces the inductance, too. If customising the supply lines like this results in being able to keep the ESC in a cooling area perhaps, it may be worth doing.

Sounds a bit like a cue for some more experiments coming up………. PB

[Klippy]

Hi Peter, I understand you can buy a gadget that allows you to lengthen battery leads to silly numbers. It looks to me like a bank of capacitors. If you liken the flow of electricity to the flow of water (I know you shouldn't, but lets do it) Consider a 1 metre pipe with say 6bars of pressure in it, with a valve on the end. Open the solenoid valve, water flows, close the valve, water stops, suddenly. Now keep the pressure the same but now use a 10metre pipe, when you close the valve that pipe clangs! and how. You now need a small accumulator to absorb the energy of the flowing water. I can imagine something very similar happening in long battery leads, and the input side of the ESC doesn't last very long.

You can't lengthen the motor leads, but you can lengthen the battery leads a bit (I would guess about double without running into too much trouble.

Any comments?

[Peter Hepworth]

Hi Tony,

              Thanks for input, it seems to me that if it doesn't need fixing don't fix it. Apparently a length of  of 18" is the absolute max. when I joined the electronics game long ago we only had valves to cope with Oh and the first transistors OC72 if memory serves. After all this it seems the ESC was faulty??

  Thanks to all...

Peter

 

P.S. You aren't by any chance the Tony Prince who worked on Radio Caroline &  Luxembourg...? 

Edited By Peter Hepworth on 16/06/2010 10:38:44

[Myron Beaumont]

Peter,I think you mean OC 71

[Brian Parker]

Peter B,
 The emissions from the motor leads are probably not an issue.

 My posting was in response to the ‘twisted pair’ and ‘twisting the three motor wires together’. Apart from anything else they would probably get rather warm if twisted tightly together.

 

On the subject of cable length between the battery and the ESC.

Consider..., as the capacitor is charged and then discharged the current will be out of phase and leading the voltage. When the capacitor is fully charged, voltage is maximum and current is minimum with minimal power to dissipate.

 However with inductance, voltage leads current, (the exact opposite of the above). Increasing the cable length increases the inductance and so alters the circuit impedance.

 Also by increasing the cable length/resistance (current in phase with the applied voltage) the impedance of the circuit will again be altered.

 

The capacitor will then have to sink more current.

Edited By Brian Parker on 16/06/2010 11:45:33

[Peter Hepworth]

Thanks Peter, Yeah right Myron,  told you it was a long time ago

 

Cheers

 Peter H

[Chris Bott - Moderator]

Guys I go back to the AC126 and AF116 - is that earlier or later?

Edited By Chris Bott on 16/06/2010 17:07:52

[John Privett]

Posted by Myron Beaumont on 16/06/2010 10:54:04:

Peter,I think you mean OC 71

 

Going off on a bit of a tangent...   I've just found a site dedicated to Mullard transistors. The OC72 was a high-power version of the OC71, and the OC70 was a low-noise version of the OC71.  The R.F. transistors OC44 and OC45 followed a couple of years later in 1956.  Apparently Mullard's first transistors in 1952 were the OC50 and OC51 - which I'd never heard of until now!

 

[Chris Bott - Moderator]

Amazing John well found. Isn't the www wonderful? 

Edited By Chris Bott on 16/06/2010 17:37:51