Motor Kv - the truth

[Chris Bott - Moderator]

I think that the ESC is not setting the rotational speed, with the motor following. The ESC is in fact reacting to feedback it gets from the motor. It "sees" a magnet section being pulled through the field that was generated by the current. As it "sees" the magnet leave that field it switches so the field is again ahead of the magnet. Thus generating a rotating field always slightly ahead of the rotor. How much it is ahead is set in the timing setting.

 

The speed of rotation is set by how hard the magnet is pulled, and this in turn varies with voltage. As well as keeping the field pulling the rotor, the ESC also switches on on off much more rapidly. The mark/space of this switching setting the effective voltage. The speed if this switching is set in the frequency setting.  

[Tim Mackey]

I plan on flying the model again tomorrow with the smaller wooden 15 x 8 prop, so that I can compare the data as well as the flight performance. I will be using the exact same battery as with the 15 x 10 APC prop, and plan for a similar flight pattern and time.

 

To recap and summarize the average high figures obtained so far.

 

  4 minute       STATIC                Amps        Watts          RPM

 15 x 10 APC e                             47               650           6750

 15 x  8  wooden                          56 +            800 +       6500 -

 

 

10 minute      FLIGHT  

 

15 x 10 APC e                              44               625           6700

15 x 8  wooden                            50               725           6550      

 

 

Edited By Timbo - Administrator on 15/05/2010 14:27:24

[Erfolg]

Timbo

 

I assume that the measured voltage, is input to the ESC. I guess that the voltage to the motor is lower?

 

Is this drop enough to have a significant to the KV measured? Which I guess you obtain by dividing the revs by volts.

 

If the voltage loss is measurable I would expect that the KV would apparently be higher than the true value.

 

Again I am assuming that the revs are directly proportion to volts, so supplied voltage is irrelevant within viable  range?

 

With respect to how a ESC operates, we have had another very full discussion in an earlier thread, where it was established how timing was relevant to the motor and how and when the current was applied. It did resolve the discussion was a DC or AC current applied to the motor, by reference to published oscilloscope traces.I seem to remember that the method of throttling was achieved by a simpler process, than some had conjectured. Just cannot find the thread.

 

I do not know any answers and am just curious

 

Erfolg

 

 

[Tim Mackey]

Yes, the input voltage is to the ESC not the motor, but then surely the specs should allow foir that anyway, as you cant run the motor without an ESC. We did cover this point earlier in the thread. Anyway, being not quite so curious or bothered about it as some folk, I am happy to conclude that the stated Kv of motors is non- gospel, and unsurprisingly, some are closer to spec than others. I have added in the data from this mornings flights using the lighter Zinger 15 X 8 prop as opposed to the APC 15 x 10, and in line with the results I got in static testing....the higher pitch APC is definately pulling LESS current and producing less electrical power in Wattage terms, than the smaller Zinger beech prop.

I would say ( no onboard airspeed measuring device ) that the model flew just as well with either prop, and no discernable difference in speed was observed by either myself or photographer. It seems that for whatever reason - efficiency, stalled blades, blade shape or the price of fish....the heavier, higher pitch prop produces less actual power but the model is still flying the same, albeit with LESS power consumption.

[Erfolg]

Timbo

 

I did read the rest of the thread. These days i do not find it that easy to remember all what was said. Or even recognising what is relevant. I think part of the problem is the twists and turns in the discussion.

 

With regard to how the manufactures undertake their measurements, it would not surprise me if this was not done on the input to the motor. I remember reading that brushed motors were tested by using a constant voltage source. At face value this does not seem as easy with a brushless motor.

 

I find it ironic that I know we have a number of Electrical Engineers, who choose not to contribute

 

Is it possible to produce a simple table of results, perhaps with deviation from stated Kv. Just curious how great the deviation from manufactures reference value actually is

How did you measure the none load revs. A disc or just marked the shaft, or some other method?

 

None of this matters, although I am curious.

 

Erfolg

 

[Chris Bott - Moderator]

Timbo, excellent logical testing.

 

As I have experienced whenever I've strayed away from APC-E props, we seem to get better performance or duration using  them.

 

I think the best way of looking at different props is not that APC-E's PRODUCE less power, but that they USE less power. I.E they have the edge on efficiency.

As you say - who cares how or why! 

 

 

 

 

[Tim Mackey]

Posted by Erfolg on 15/05/2010 15:16:02:

Timbo

 

I did read the rest of the thread. These days i do not find it that easy to remember all what was said. Or even recognising what is relevant. I think part of the problem is the twists and turns in the discussion.

 

With regard to how the manufactures undertake their measurements, it would not surprise me if this was not done on the input to the motor. I remember reading that brushed motors were tested by using a constant voltage source. At face value this does not seem as easy with a brushless motor.

 

I find it ironic that I know we have a number of Electrical Engineers, who choose not to contribute

 

Is it possible to produce a simple table of results, perhaps with deviation from stated Kv. Just curious how great the deviation from manufactures reference value actually is

How did you measure the none load revs. A disc or just marked the shaft, or some other method?

 

None of this matters, although I am curious.

 

Erfolg

 

 

Hi Erfolg...no worries mate, I wasn't criticising, just reminding peeps that this particular aspect had been discussed already

I am sure you are correct in the assumption that manufacturers use volts at the windings ( or as someone else suggested, quite possibly its all calculated theoretically rather than actual measured and logged ). My reference would be for them to state the actual results as used - EG: via an ESC which after all,  we all need to use to even get the thing to work!

Of course, this then starts to bring in factors such as timing and so on.

It would be possible to tabulate some results, but from the 4 or so examples tested so far its obvious that large deviations exist, and I am not sure that averaging the results would be much help. I simply use the stated Kv as starting point, and often find that I need another cell adding or maybe one removing as well as trying a few props either side of the planned one in order to hit the spot. No biggy.

The RPM is measured electronically by a sensor which connects to any two of the three motor wires - this is connected to the eagle tree logger and RPM is displayed either real time on the PC screen, or captured along with other data and then uploaded to the PC later, where the "flight" can be replayed on the PC.

[Tim Mackey]

Thanks Chris.... and I think you are dead right.

I have to say though, the wooden prop does look right when she is parked

[Chris Bott - Moderator]

Yes I agree Timbo depending on the plane, wood is the only thing that looks right.

 

I've bought a couple of these, could be interesting. Although they will be going on my two moths where performance really isn't an issue anyway.

 

They are here Turnigy wood electric props 

 

[Tim Mackey]

Interesting... I only bought one wooden prop from our hobbyking friends, and broke it on its maiden

[Myron Beaumont]

Going back a bit ,referring to ic engines (& yes this is not really off topic )do you remember when we had tests of engines (Chinn for example ) quoting bhp output ? I could never work out what the power output had anything to do with anything as long as the rpm was there ?

Can you see the analogy I'm getting at ?

[Erfolg]

Timbo

 

I take it that the Kv, values are obtained under no load conditions?

 

I have some reservations with regards props. It seems possible to produce a prop, that has a lot of mass, which takes proportionally more power to accelerate to full speed, then commensurately more power to maintain momentum, essentially a flywheel. Yet thrust could be zero

 

The answer seem to be to measure watts in and plot against kg thrust. I know Biggles the Elder will not like this and suggest Newtons are more appropriate, as grams are mass. But scales read Kg, irrespective of the ideal. So a ratio of force per watt seems OK.

 

As with many things It will not be perfect but indicative for any dia. prop.

 

Erfolg

[Roberto Gava]

Posted by Timbo - Administrator on 14/05/2010 23:47:08:

I plan on flying the model again tomorrow with the smaller wooden 15 x 8 prop, so that I can compare the data as well as the flight performance. I will be using the exact same battery as with the 15 x 10 APC prop, and plan for a similar flight pattern and time.

 

To recap and summarize the average high figures obtained so far.

 

  4 minute       STATIC                Amps        Watts          RPM

 15 x 10 APC e                             47               650           6750

 15 x  8  wooden                          56 +            800 +       6500 -

 

 

10 minute      FLIGHT  

 

15 x 10 APC e                              44               625           6700

15 x 8  wooden                            50               725           6550      

 

 

Edited By Timbo - Administrator on 15/05/2010 14:27:24

For me, those figures support my theory about props.

The 15x10 prop lost 50 rpm between static and flight test meanwhile the 15x8 prop rose 50 rpm on the air because of the unloading by fwd speed.

This can only be explained by the decrease of resistance to spin of the prop in the static test due to the stalled condition of the blade that reduced its induced drag.

[John Privett]

Posted by Myron Beaumont on 15/05/2010 23:00:24:

Going back a bit ,referring to ic engines (& yes this is not really off topic )do you remember when we had tests of engines (Chinn for example ) quoting bhp output ? I could never work out what the power output had anything to do with anything as long as the rpm was there ?

Can you see the analogy I'm getting at ?

 

I remember those tests too.  Not quite sure exactly what you mean when you say you "could never work out what the power output had anything to do with anything as long as the rpm was there ?"

 

You could have two engines that can both rev at 15,000 rpm.  Are they equal in power?  We don't know! 

 

But if we know that one can do it on a 12x6 prop and the other can only do it with a 6x3 now we're getting a better idea of the relative power of the engines.  

 

From the power quoted we are in a better position to compare different engines - though this is hampered a bit by most engines producing their peak power at revs somewhat higher than most of us are comfortable running at.

 

Knowing that engine A will do 14,000rpm on a 12x6 and engine B does 13,000rpm on an 11x7 makes it difficult to know which is the more powerful.

[John Privett]

Posted by Erfolg on 17/05/2010 13:59:47: 

I have some reservations with regards props. It seems possible to produce a prop, that has a lot of mass, which takes proportionally more power to accelerate to full speed, then commensurately more power to maintain momentum, essentially a flywheel. Yet thrust could be zero

 

A heavier prop will certainly accellerate more slowly.  But at a constant speed it does not require more power to "maintain momentum" - not unless it also has more drag.

 

[Brian Parker]

John is correct.
 An accelerating flywheel (prop) is storing kinetic energy. A decelerating flywheel is releasing energy. At a constant speed a flywheel is neither storing nor releasing kinetic energy.

 If we accept that the brushless motor, at a constant speed will maintain (unlike an IC engine) a constant velocity throughout each revolution the prop will behave as a flywheel at a constant speed.

 Just how much kinetic energy can be stored in a prop given its small mean radius, low mass and the fact that mass is not concentrated at the prop tips (flywheel rim?)?

Not a lot I expect (but am not quite geeky enough to work it out for myself).

[Erfolg]

Brian

 

 

As the flywheel has an angular acceleration, energy is required to keep it going.

 

Even if you take conservation of energy as your basis, you know that a spring with a weight released from height, will not in practice return to the original height. I used to wonder as a teenager, why if I went down a hill, i could not coast up to the top of the next similar hill, based on conservation of energy (we all know why and that the concept has limits.

 

But as you say or indicate the most energy is required in the acceleration phase.

 

I was pointing out that a heavy prop takes more energy, particularly as it moves through the air, as the models orientation to the earth will change, altering the g force on the prop, these changes will again use some energy. How much? proportionally probably quite low, yet the greater the mass, the more energy will be used during the acceleration due to +ve changes.

 

Erfolg

[Sparks]

I often use Colin Dedman's method for estimating motor Kv. This involves spinning the motor in a pillar drill and measuring drill speed and peak voltage across each of the three phases (these are then averaged). Apply the following formula to get motor Kv

Kv = (RPM/VOLT.pk) / 0.95

or, if using a digital volmeter to meaure RMS Voltage

Kv = (RPM/(VOLT.rms*1.41)) / 0.95

These are the results I got from testing a range of E-Flite motors

Motor  Published Measured
No       Kv                Kv
480   1020          1038
10     1100           1156
15        950             951
25        870             902
32        770             788
46        670             667
60        400             394

 

Probably not an exact method but, as you can see,  good enough for practical purposes.

 

 

[Tim Mackey]

Ah, but herein lays the rub surely. These are theoretical figures, and what I wanted to do was see the actual RPM of the shaft when supplied with a specific voltage - as surely for our intents and purpose this is what matters. It would be interesting to see what results were obtained on these same motors using my test method and comparing against this theoretical method. If I find myself inclined, and with the spare time to do so, I may well do this calculated method on the 3 motors I tested here.

[John Privett]

I'd not heard of the Colin Dedman method, but it makes sense.  I'm just not quite sure where the "/0.95" comes from! 

 

The method is measuring the back-emf from the motor,  and that is precisely the reason that a motor has a maximum speed per volt.  As the motor accellerates (powered by the battery) the back-emf increases.  When it reaches the battery voltage then no current will flow from the battery to accellerate the motor, so if nothing is slowing it down (like a propeller!) then the motor will maintain it's current speed and can turn no faster.

 

Timbo,  I would expect the Colin Dedman method to give results closely matching the actual speed you find in your tests.  It would be interesting to see!

 

Erfolg - no energy is required to maintain the speed of a flywheel,  a propeller,  or anything rotating,  other than to replace energy lost through friction in the bearings,  air resistance etc.  Just the same as in your spring or bicycle examples.  Individual parts of the prop are accellerating towards the centre and internal forces in the prop, towards the centre, stop the prop flying apart.  But there is no net force maintaining the motion.  Have a look also at conservation of angular momentum.

[Sparks]

OK Timbo, I shall have to drag out my PSU and do some tests on a couple of the E-Flite motors to see how the no-load Kv compares to the generator Kv method. Will report back later.

 

John, the 0.95 is an empirical value - see the following link for further explanation/links

 

Measuring Kv

 

Richard

 

 

[John Privett]

Interesting link - thanks Richard!

[Erfolg]

John

 

Your are making my point for me, as there is a centripetal acceleration which generates a force, these forces act as you say on bearings etc. Requiring an input of energy, all be it small. Even precision centrifuges (which may operate in a partial vacuum) take some energy to keep them spinning. In the case of one industry requiring a power station to keep them going, not just at start up. So we both recognise some energy is required, Or we would have a perpetual motion machine. Yet I get the impression that we are arguing about angels and pin heads, as we are debating can we detect the scale of the additional energy to keep it spinning, between a light and heavy propeller. I thought I had made it plane, that I do not know if it is significant, but it is a variable.

 

I think the word commensurate, could be unclear. It was my intention, to indicate, some power to maintain the velocity. Which i suspect was taken to mean a greater amount. Which was not my intention. 

 

I do not know the Colin Debman method, yet from the description by John, seems to have a good scientific basis. Is this the process that the motor manufacturers/designers use.

 

For me the principle issue is there a recognised standard method of determining Kv.

 

Erfolg

[John Privett]

Erfolg - that link in Richard's (Sparks) post a couple of posts up was interesting.  Colin Debman, it appears, is a professor of electrical engineering - so I think that lends even more weight to his method.

 

And also from that link it seems that different motor manufacturers get their Kv figures by different means or even by plucking them out of thin air.

 

Anyway, we'll agree about the "perpetual energy machine" impossibility even if we might quibble slightly about some of the fine (and irrelevant to this thread) details... 

[Erfolg]

I have just finished sweeping up a mess from my home improvements. At this point my mind drifted back some 40 odd years. As a student we had a lab work, which at first sight, seemed easy, no work in a lab, no results to take down. Yet it soon became apparent that it was was one of the the most demanding. It was identifying where the flaw lay in a series of systems, which mathematically appeared to require zero energy to maintain them. Many were based on the concepts of  conservation of energy. It was tough, as invariably an assumption was made somewhere that some expression expressed the process completly, where as it was the ignored force, reaction, or partial mathematical model, that was the issue.

 

I think it is this aspect of establishing how Kv is derived and its basis that is of great interest. Also understanding the limitations of the methods.

 

Why? it could help in understanding which of the attributes should be considered in the selection of a motor, for a particular duty.

 

Erfolg