Measuring the efficiency of a motor ???

With great difficulty.

If at all.

The classic engineering way is a dynamometer, to measure output torque: that times the rpm with appropriate constant is the output power.

It tells you nothing about propellor efficiency though, and must perforce include ESC losses and possibly battery losses as well as motor losses.

If you want a a really quick and dirty approximation, though, which for quite subtle reasons works fairly well in MOST cases, just take the prop off, measure the RPM, and then put the prop back on. The efficeincy is very close to the ratio of the on load RPM to the off load RPM, if the battery doesn't sag too much.

The better approach is to do some fairly tedious mathematics involving measuring the current drawn under zero load conditions as well, and that way you can get a better figure.

That will tend to give you the motor + ESC efficiency if you have a volt/ammeter (whattmeter) between the battery and teh ESC.

You have to do more maths to separate out the controller losses..

ACTUAL power OUT of the motor AND prop is a function of the amount of air being moved by the prop, and the speed its accelerated to, and is a fiendish thing to measure.

and almost useless when you do, since it changes according to the models airspeed.

You best bet is to gear the motor, and use a big coarse pitch prop. The quieter it is, the more efficient is is.

If you are used to glow engines, forget everything you thought you knew: very little of it applies with electrics ;-)

Ah the speed and volume of air moved, good! What about simply hooking the prop/motor to a scale to measure how many ounces it pulls? I am guessing that this will only measure thrust, and not Watts because it does not take into account the speed of the air.

Maybe my measure above combined with some way to measure the speed of the air (little paddles turning a wheel hooked to a gear) would derive a reasonable approximation?

"The Natural Philosopher" wrote

I'm very curious to know what your method is for measuring the RPM of a motor without a prop on it.

"Peter Olcott" wrote

Ahh, but what factors or formulas or constants would you use to turn this thrust into meaningful watts output?

As a way to measure relative performance, it is good, but even better is to measure the thrust created with the prop-motor combination in a wind tunnel, so you are not measuring static thrust.

Unless you are wanting to know how well it accelerates from a standing start, of course.

I just figured out how to measure the airspeed, something like measuring the minute electricity generated by a brushed ducted fan motor that is placed in the airflow, once this has been calibrated to airspeed. There would probably already be devices that operate on a similar principle.

Measuring air volume is another matter maybe you could put the motor/prop in a duct and measure the airspeed at the end of the duct, by knowing the diameter of the duct you could figure air volume moved per unit of time. This kind of measure would be an ultimate measure inherently factoring in the total efficiency of the motor, prop, gearbox (if any) and anything else.

I am not sure whether or not I erred in any of my analysis, I am a software engineer, not an aeronautical engineer.

"Peter Olcott" wrote

Several problems, there. I'm not going to take them in any particular order.

If you use a calibrated airspeed by generating electricity, the airflow is going to be non-turbulent while you are calibrating it, (before it hits the measuring prop) and straight flowing against the prop.

If you used this rig to measure airspeed coming off of a prop, the air is going to be cork-screwing, which will have an undesired effect of the measuring prop. To be accurate (more so, at least) you would need to use straightening vanes to straighten out the airflow. Ooops, now we have added drag which will slow the airflow coming off of the powered prop.

If you put the powered prop in a duct, you still have the corkscrew problem, and increased drag from the sides of the duct. That will be quite significant. Then, you might need to factor in increased powered prop efficiency, _if_ you maintain very, very small clearances between the power prop tips and the duct walls, as you would with a good ducted fan unit. You also have engine mount and duct support (possibly not the second) drag to consider throwing off your measurements.

You have too many unknown variables, and unlike units that need to be solved, and as everyone knows, you have a big problem trying to solve more than one unknown variable at a time.

Long to short? Use the ballpark figures for efficiency of the props at the observed RPMs, and with the ballpark efficiency for the different types of motors at the amps and speed you are going to be running. Subtract out the loss for the type of gearbox (if any) you will be using.

Then there is the very best way to figure it all out.

It seems overly simple, but the answer is.....

Put something on a plane, and go fly the sucker ! ! ! ! ! !

Plus, it is more fun than bending your brain like the above would require. (at least it is for me-I hate math, especially ugly math)

I want to optimize my purchase of brushless motors. In the ideal case the companies that make these motors would provide these figures in advance. How exactly can the efficiency of a motor/prop combo be precisely measured, disregarding the cost of this measurement?

I am guessing that the expensive part of this may have already been done and that all that I need to know to determine how much power that any given motor is generating is to know the RPMs that the motor is providing to a specific prop, and then look up the Watts per RPM in a precompiled table for this specific prop. I already know how to determine how much power the motor is drawing.

you can try...

very simple. thrust times velocity=power.

google 'convert lb feet per second to watts' for the constant..

put two bits of black tape on the motor rotor and use a tacho.

I tend to agree.

Mmm. 50/50 here.

That's easy. you need a whattmeter, a tacho and one of the free calc programs.

Feed in no load volts amps and RPM, then on load volts amps and RPM, and the calcs will tell you what the KV, Io and Rm figures are.

Calculating efficiency at a given current draw and voltage is then trivial.

Don't even need that. For most motors efficiency is almost exactly :-

on load RPM/off load RPM.

Its not quite that simple, but the motor that holds up best RPM wise at a given power draw is nearly always the better motor efficiency wise.

However, efficiency isn't everything.

You need to spend some time hanging out here

easy, measure the amount of energy used to make noise the amount of energy given up as heat, subtract the two numbers from the total wattage, and whatever is left must be the amount of work done.

Or, get MotoCalc, and accept that their estimations are within a usable range, and go have fun.

Life is too short to get all bogged down in absolute efficiency numbers on something that is designed to have fun with.

I don't know about TNP, but I use a painted disc.

So the free calc programs convert specific Prop RPMs into Watts? Can you send me a link to one of these?

well download the motocalc 30 day demo from

Or hit the power forum on

its not that easy to take and arbitrary prop and say how much power it takes to drive it: the more telling measurements are to be had by working out how much the motor slows down under load, and how much current it draws on no load.

Then the calc programs can estimate the drive train efficiency.

Actually I think that I found the answer:

I have found plenty of help on remote control sites, since many RC planes use electric motors.

There is a commonly used formula by Bob Boucher in his book "The Electric Motor Handbook". It is:

Power = k * P * D^4 * N^3 where:

Power is in Watts P = prop pitch in feet D = prop diameter in feet N = RPM in thousands

The k is a coefficient that depends on the propeller type. There are websites that list tons of numbers for all known RC props. However, I don't know how that would relate to a real prop. I've been looking for that info for a while.

That is the formula that underlies motocalcs behavior certainly, BUT there is a big gotcha: even props of the same 'type' do not display a particularly good constant between sizes.

Also, the graph is non linear at high pitches or high RPM where turbulence and or transonic effects lead to increased drag.

In real life, however, this is arsey versey for what you are normally trying to achieve: Namely for a given airframe, what motor/(gearbox) pack and prop to stuff in.

My general approach is to select a power level - that might be anything from 40W/lb for a light slow scale or vintage model, through 60-80W/lb for a spirited sport plane, 100W/lb for either full vertical or high speed, and 150W/lb for both..through 200W/lb for an EDF with a decent rate of climb, or 300W/lb plus for an EDF with full vertical capability.

Note that watts per pound OUTPUT translates exactly into drag times airspeed plus rate of climb. Its a terribly ubiquitous formula. A full size cub is about 50W/lb as is a model cub.

Drag is a function of weight and the lift to drag ratio of the model, or in fact at around cruise speed, the tangent of the glide slope times the weight.

If you tune the model up right, you are likely to see an overall drive train efficiency of around 50% - prop, motor controller etc.

Anyway, applying the above as a first step allows you to determine final model weight and pack size for a given duration .

The final steps are about selecting a prop size and motor and possibly gearbox.

In general the bigger the prop the more efficient it is: right up to the point at which the torque effects make the model unflyable. 33% diameter to span is a fairly decent upper limit..15% is a reasonable lower limit. EDF of course is far less, and much less efficient as a result.

The next stage is to determine the models stall speed..there are formulas for that. It relates to the square root of wing loading.

Then you need to arrive at a pitch speed - static pitch speed should be at least twice the models stall speed. For high speed planes 3 times or more.

THEN you can look at that equation and try and work out what RPM will be needed to suck the sort of power that you have calculated you need, and multiply that by the pitch to get the pitch speed.

Finally you have a prop size, RPM needed and pack energy all settled.

THEN you need to select the motor. If its a brushless of a cooking variety, assume about 70% efficiency. Take the loaded prop RPM and divide by that sort of figure, to get the putative 'unloaded' RPM.

Divide that by whatever pack voltage you have selected, to get the motor KV, and go looking for a motor that will do that kv at the sort of currents the power needed divided by the pack voltage gives you. If te KV is too low, you need less voltage or a gearbox. Or you will have to settle on a smaller prop.

Or buy Motocalc for $30 and just enter the data in. :-) with Motocalc, all you really have to do is come up with the motor constants..which a reasonable whattmeter and tacho can help you define, and a reasonable figure for the prop constants, which is generally arrived at by feeding back real world data into it and adjusting the constants till the simulation matches reality.

The key thing is that motor efficiency really doesn't come into this much at all. Its generally between 55% for a rubbish brushed motor, and about 80% for a good brushless, with 70% being an average value for budget motors. What is more important is to get the motor big enough so its not running on the back end of the efficiency curve and getting dangerously hot, and to get a decent prop size.

And use the watts per pound rule. IF you have the right watts per pound, you can fiddle with the prop to get the performance you like, but without the right watts per pound, you wont get anywhere.

"The Natural Philosopher" wrote >

Yeah, I guess that _would_ work on an outrunner, pretty easily. It would be a bit more difficult on a regular brushed motor, I would think.

I guess you could put a small rotor on a regular motor, and it would not change the RPMs very much.