Motor Kv - the truth

[Brian Parker]

My tu’pence worth on Zinger verses APC.
Its not in the wood its all in APC's distortion and drag.
 The drag of a rotating prop resists the rotation of the drive shaft, producing a torque reaction.

 Maximum rotational speed is reached when torque reaction equals shaft torque.

 Lower the drag (through distortion) and rotational speed will increase for a given power input.

This may or may not increase the efficiency (Thrust *Velocity/airflow factor).

[Roberto Gava]

Hi Timbo,

 

This is my first reading of this thread.

Just a comment on your first measurings:

A 15x10 prop is stalled in an static test. A 15x8 is close to stall, but it isn't.

This should be the reason why you got more revs on the first one.

[John Privett]

That was my thinking too Roberto.

 

I think I posted that on the "cloned" thread Timbo mentioned a couple of post ago.

[Danny Fenton]

I am not sure that 7000 rpm on a 15 x 8 would be near the stall but raises an interesting point.

 

[Tim Mackey]

Me neither.

Even if it was "stalled" that surely only affects its thrust /performance...rather then the RPM that the motor is able to spin it at.

[Brian Parker]

Technically, are not all props stalled during static testing at whatever the RPM? (Unless the tests are conducted in a wind tunnel).
 In flight, at a given air speed the Prop will 'unstall' because the lift/drag ratio of the prop is increasing in the moving air.

[Roberto Gava]

Some specialized softwares such as Extended Prop Selector or eMotor tells you if the propeller is stalled or near to stall.

@Timbo: McLaren F1 F-Duct technology is based on the drag reduction of the rear wing of the car due to the fact that the injected air stalls it. So, a stalled propeller will consume less power for the same revs that one that is not stalled.

That's because stall reduces the induced drag.

@Brian: stall occurs at a certain angle of attack. In a propeller if the pitch angle is less than the critical stall angle it is not stalled on static test. When the pitch angle is higher than the stall angle, then the displacement speed reduces the effective angle of attack and, being stalled on static it is not at certain speed.

I.E. according to EPS, that 15x10 prop is stalled between 0 and 2.14 MPH, it is probably stalled (depending on its profile) between 2.15 and 7.38 MPH and is close to stall between 7.39 and 12.8 MPH.

[John Privett]

I was looking for more info on this last night.  There is an engineering page on the APC props website (here) that states that "During static operations, the blade is operating under dominantly stalled conditions."  As Roberto states,  it's to do with the effective angle of attack,  and near the hub that will be 30 or 40 degrees on a typical prop.

[Danny Fenton]

Interesting stuff, but doesn't prop speed also have a bearing on the stall of a prop. I would certainly expect the prop to stall at high rpm.

 

I tried using the e-motor calculator you mention and fed it numbers for the setup in my Chipmunk, it basically told me that it wouldn't work. Good job I don't put much faith in calculators. It is an interesting area to investigate though.

 

Cheers

Danny

[Tim Mackey]

Hmmm, I never get on with these theoretical calculators either - I prefer to do real time actual experimentation.

Now as to the stalled prop thing.  I understand ( I think ) what a stalled prop is, and how obviously the stall is affected by both RPM and speed through the air. What I still cant quite get my head around is how stalled Versus non-stalled blades can effect the RPM of a static mounted electric motor. Is this simply to do with the stalled blades affecting the drag - and consequently allowing a higher RPM?  I can see how it may affect power consumption due to less / more drag but could this affect things so much as to actually slow /speed up the electric motor considerably ?

[John Privett]

The idea is that a non-stalled blade is accellerating air through it and producing thrust - obviously puts a load on the motor (can't get owt from nowt!)  A stalled blade however is just thrashing a limited amount of air around and the load on the prop is less - so for the same power from the motor it runs faster.

 

I've noticed a similar effect with a PC cooling fan.  In normal operation it's working reasonably well accellerating air into the PC to help cool it.  But if you partially block the intake then it's not doing the same amount of work (as you're restricting the airflow to it) and the speed increases as there is less load on it.

[Chris Bott - Moderator]

You may all be right but I'm still not totally convinced by this stalled prop stuff. If a prop was taking less power because it was stalled, wouldn't we see an increase in power as the model accelerated and the prop moved from a stalled to an un-stalled condition?

I have never seen any evidence of this in an Eagle Tree graph. 

 

The other evidence I have that Tim's phenomenon may be down to prop design, particularly the tip shape, is that I've flown models with both APCE and  "I/C" props and notice a marked difference in both performance and flight times. This indicates to me that the electric props are noticeably more efficient.

 

 

[Chris Card]

Tell ya what - this is all getting so intriguing I think we need to get up a petition to demand that Timbo does some further experimentation to get to the bottom of it all!

[Chris Bott - Moderator]

It is intriguing and I look forward to more test results.

 

The Pragmatist in me though says that the "budget" motors I use have a variation in their kV's even if I buy the same spec motor in different months. I decide to live with that, they are close enough not to worry a sports flier.

For a new model I have to half calculate half guess which motor I need, once I have it I do static tests with a fist full of props to get close to the size I need and then I tweak the prop size with flying tests. Even then I've been known to change horses to a different motor.

After quite a few years of trying other makes, I stick exclusively to APC-E props (or sometimes props that look very similar).

What it means for me is that I can't entirely predict exactly what is needed for any particular model. Getting that last few ounces of performance or minutes of flight time is all part of the enjoyment of this fascinating hobby. 

 

 

 

[Myron Beaumont]

Probably a stupid comment ,but when you block the pipe on a vacuum cleaner , it speeds up considerably doesn't it ?.Now then ,is this because the airflow is nil ,or the blades are stalled ,or what ? Just an analogy that I'm sure an expert could put me right on .

In the case of a prop or for that matter an EDF -the higher the air speed flow -the faster it spins

Edited By Myron Beaumont on 13/05/2010 08:44:30

[Brian Parker]

Under the static tests we have airflow over the blades but we have no forward velocity.
The lift vector is tilted against rotation and the thrust force is transferred to resisting rotation rather than providing forward motion.

 Is that not a technically stalled situation?

 (The APC Engineering site seems to confirm this).

 Also ref. the Eagle Tree Graph above it will show no change as we are simply transferring forces.

[Tim Mackey]

Posted by Chris Card - Moderator on 13/05/2010 08:27:07:
Tell ya what - this is all getting so intriguing I think we need to get up a petition to demand that Timbo does some further experimentation to get to the bottom of it all!

 

Oy you.... watch it !

 

Actually I am later today going to be flying the exact same model and powertrain with the smaller lighter Zinger 15 x8 prop, so we will have actau flight data to compare against the APCE 15 x 10

[Roberto Gava]

Resuming the original topic of motor Kv.

Can someone explain me why a brushless motor changes its rotating speed according to the applied DC voltage to the ESC?

Correct me if I'm wrong. But I undestood that Brushless motors are similar to three phase asynchronous AC motors.

If it is right, they create a rotating magnetic field by the combination of the three phace AC over different windings in the stator and the rotor magnets follow that rotating field. In this case, the speed is defined by the AC current frequency created by the ESC.

Are the ESCs changing their working frequency according to their feeding DC voltage to emulate the behaviour of a DC brushed motor?

Edited By Roberto Gava on 13/05/2010 20:44:56

[Myron Beaumont]

This is getting interesting y'all

[John Privett]

OK - I'll stick my head above the parapet,  and prepare to have it shot off!

 

In a brushed motor, the commutator switches the power to the right windings at the right time to make the motor rotate.  The speed at which the motor rotates (when no load is applied to it) is proportional to the supply voltage.  The design and physical characteristics of the motor determine the 'revs per volt.'

 

The brushless motors that we use are really just brushed motors that have had the brushes and commutator replaced by a box of clever electronics.  The electronics switch the power to the windings just as the commutator does on a brushed motor.  So the speed is still proportional to the voltage.  For a given voltage the ESC is simply incapable of making it spin any faster than the kv rating dictates either by warying the timing or by any other means.  

 

To reduce the motor speed, I believe the ESC reduces the effective voltage supplied to the motor by very rapidly switching the supply on and off.  So 'on' for half the time and 'off' for half the time is effectively half the voltage,  'on' for 10% of the time and 'off' for 90% is effectively one tenth of the voltage and so on.

Right,  fire away...  Corrections and/or refinements to the above are welcome!

[Tim Mackey]

Seems a nice simple ( I like simple ) explanation to me John

[Myron Beaumont]

John

What you are saying bears out perfectly the infamous water analogy ie pressure -not flow (current)-Over to you

[Myron Beaumont]

John

X -posted Timbo didn't I

He said what I said ! (didn't he ? )

[Roberto Gava]

Posted by John Privett on 13/05/2010 21:45:32:

OK - I'll stick my head above the parapet,  and prepare to have it shot off!

 

In a brushed motor, the commutator switches the power to the right windings at the right time to make the motor rotate.  The speed at which the motor rotates (when no load is applied to it) is proportional to the supply voltage.  The design and physical characteristics of the motor determine the 'revs per volt.'

 

The brushless motors that we use are really just brushed motors that have had the brushes and commutator replaced by a box of clever electronics.  The electronics switch the power to the windings just as the commutator does on a brushed motor.  So the speed is still proportional to the voltage.  For a given voltage the ESC is simply incapable of making it spin any faster than the kv rating dictates either by warying the timing or by any other means.  

 

To reduce the motor speed, I believe the ESC reduces the effective voltage supplied to the motor by very rapidly switching the supply on and off.  So 'on' for half the time and 'off' for half the time is effectively half the voltage,  'on' for 10% of the time and 'off' for 90% is effectively one tenth of the voltage and so on.

Right,  fire away...  Corrections and/or refinements to the above are welcome!

 

I follow your explanation and I've a couple of doubts:

1 - In a brushed motor conmutator is mechanically driven by de shaft, so the conmutation speed is proportional to the rotation speed. So, in this case, the higher the voltage, the stronger the magnetic field and more revs is the result.

2 - In a brushless motor conmutation speed is defined by ESC (I don't know if it is variable or not)

At this point I see two posibilities:

 a - ESC functioning at a fixed conmutation speed and reducing efective voltage as you said resulting in higher diference (slip) between magnetic field rotating speed (fix) and rotor speed.

 b - ESC is varying conmutation speed proportionally to the throttle stick position and maybe reducing also efective voltage to reduce motor speed and reaching a maximum conmutation speed and voltage at top throttle that, according to motor Kv rating results in actual maximum revs (always affected by slip proportional to the resistance of the prop to spin and inverselly proportional to the effective voltage supplied to the windings.

Edited By Roberto Gava on 14/05/2010 21:38:23

[John Privett]

I think the definitive answer would have to come from somebody who designs ESCs and understands their workings inside-out.

 

However, my belief is that the ESC does two things. 

 

1. It switches the power to the windings at the best time to run the motor efficiently - mimicing the work the commutator would do.  If the timing is wrong then the motor would run badly,  or not at all,  or maybe backwards...

 

2. It regulates the effective voltage to the windings - as commanded by the throttle position - to permit the motor to run at a higher or lower speed.

 

I think  this is subtley different to either of your options a or b.  The "commutation speed" is, I believe, determined not by the throttle position but by the motor - exactly as a mechanical commutator would do.  We know that early brushless motors had position sensors to feedback the position of the motor to the ESC and that modern ESCs manage without these by sensing the motor position via the back-emf generated.

 

I feel that artifically altering the timing of the ESC to vary the motor speed would be a bit like advancing or retarding the ignition on your car engine to vary its speed.  It might work, to some extent,  but would probably cause the engine to run very badly.