Two Brushless motors on One ESC….YES!

[Ub Hauled]

Joe, I have a wrench to throw on your gears... how does the timing on the ESC act on the motors? Will it help with the sync?
And, how about the SENSORED ESCs/MOTORS, have you done any testing with those?
How are you?

Thanks

[questtek]

Originally Posted by keithbradley
"Ahh...I watched the video and a couple of things became apparent.
First, I think you would have much more accurate results measuring the actual thrust of the props rather than the lift created. The dual motor rig has the props out to the sides, where as the single motor rig has the prop in the center with the scale directly below it. The difference in available area behind the prop can have a drastic difference on the lift created. I think you should try a rig more like this where the thrust is measured at the scale rather than trying to lift an object off of it:"

Keith, you are an absolute wizzard and figured it out first! The area behind the prop makes an incredible difference, expecially if that is the scale. My test set up for this configuration was totally wrong and to prove it I did the testing to prove it. The following video shows what happens with the motor off to the side of the scale and then directly above the scale. I included a plot of the data at the end and comment on it.

http://www.youtube.com/watch?v=YpOq8ET_Gps&feature=youtube_gdata

I will take your excellent advice and go 2 motors 1 esc and then, IN THE EXACT SAME MOUNTING CONFIGURATION, test 2 motor each with their on ESC. As I now expect, the values should be much closer. This should go a long way to help determining the practicality of using two motors on a single ESC. It can be done but may not be feasible until after these latest tests are completed. Please keep your comments/suggestions coming!

[antslake]

That test does not prove the point we are trying to make. You added another variable by"twisting" the scale. You don't even needed the scale to determine effceincy. The props are a constant load, just measure that rpm vs input watts.

[questtek]

That test does not prove the point we are trying to make. You added another variable by"twisting" the scale. You don't even needed the scale to determine effceincy. The props are a constant load, just measure that rpm vs input watts.

Here is the data from two motors 1 ESC:
RPM WATTS
12800 229.7
14000 287.3
14700 349.3
15900 418.9
18000 539.0
18600 639.9 (This is a 600 Watt outrunner I am using)

Here is the data from one motor 1 ESC but with the motor improperly over the digital scale I use for thrust measurements:

RPM WATTS
10700 76.5
11900 93.6
13200 121.8
14050 147.0
15100 179.6
16000 192.2
17800 274.5
19500 356.1
20900 439.6

Can you make anything out of this data?

[antslake]

Yes I can! I just realized one other thing we need to know. We are assuming the second motor and prop has the same exact efficiency as the first one. If you could run a single test on the second motor and prop, and post those numbers, I can show you the math. I am typing from my phone, and will elaborate later when I get home.

[antslake]

Ok, here is how I see it. I am no expert on this, but I do know a lot, but I don't do it frequently enough and I maintain the right to be wrong, lol.

In automotive, and what I've learned from my Tekin dyno, and my medusa research dynos we can measure and compare hp/torque/efficiency of motors if the load is constant. The prop is a constant load (provided the load isn't changing when moving from 1 to 2 props due to placement on the motor mount) I think that affect is nil in the lifting setup you used. You are using 2 props of the same brand and size, but they may not be exactly the same, or are the motors exactly the same. That is why I requested the second motor test, to see if it draws around the same as the first one, which I think it will be very close.

Ok, having said that..we can simply take some numbers from the list you provided. It takes X amount of watts to turn Y amount of rpm's. We can get a number, rpm's per volt (N). Simply divide the rpm's by the watts.

2 motors, 1 ESC:
14000 / (287.3/2) =97.45 rpm's per watt (average for two motors)

1 motor, 1 ESC
14050 / 147= 95.57 rpm's per watt

At a higher rpm (which would be less accurate due to stalled props):

2 motors, 1 ESC
18000/ (539/2)=66.79 RPW

1 motor, 1 ESC
17800 / 274.5= 64.85

You are getting slightly more rpm's per volt with 2 motors. This percentage is almost negligible due to inaccuracies in the test equipment, and possibly using one battery pack, and having different input voltages. Not to mention, motor #2 may be more efficient or the prop may have less drag due to slight imperfections, or balance issues, etc.

What I see is they are even, and work fine under a constant equal load. How about trying one with a much larger prop to simulate a exaggerated uneven load?

I admit, I don't know enough about ESC and exactly how they work to give some better ideas as to why we shouldn't use this in a real setup. I do have a question about ESC's, do they sense rpm's on the poles that are not energized to know when to apply current to the energized poles? Could this be the reason why if one motor is severely loaded up different that the other, you could have major issues? I suspect the efficiency of the 2 motor setup would go down as the loads change. Or, as the load on one motor increases, the sensing part of the ESC will slow both motors down. I realize there are sensor, and sensorless ESC, but I do not know the specific differences.

[antslake]

Actually I think I found the answer:

http://www.rcgroups.com/forums/showpost.php?p=13687977&postcount=2

This would explain why one motor would have a hard time starting if the other was already spinning, and the ESc was past it's start up phase. It also explains why in a real world application this may not work. It may also just work for a while, then the FET's will give out.

[LiPo Power]

Great tests Questtek!!! I love to read your findings.
The way you do it is straight forward....
The rest is rocket science to me.....
Cheers
Robert

[questtek]

Yes I can! I just realized one other thing we need to know. We are assuming the second motor and prop has the same exact efficiency as the first one. If you could run a single test on the second motor and prop, and post those numbers, I can show you the math. I am typing from my phone, and will elaborate later when I get home.

Great....I will run the tests in the AM pomorrow and immediately post the results. Mark F and Tony are coming over tomorrow to run but my 16' electric recovery boat...that is about 20 years old............took an untimely motor fart so I have to do a motor exchange. The 1 HP brushed moter that powers this boat weight about 30 lbs! Ugg.

[questtek]

Great tests Questtek!!! I love to read your findings.
The way you do it is straight forward....
The rest is rocket science to me.....
Cheers
Robert

Thanks but it is the guys in the forum providing the good ideas. I just have the hardware handy and love to experiment.

[questtek]

Ok, here is how I see it. I am no expert on this, but I do know a lot, but I don't do it frequently enough and I maintain the right to be wrong, lol.

In automotive, and what I've learned from my Tekin dyno, and my medusa research dynos we can measure and compare hp/torque/efficiency of motors if the load is constant. The prop is a constant load (provided the load isn't changing when moving from 1 to 2 props due to placement on the motor mount) I think that affect is nil in the lifting setup you used. You are using 2 props of the same brand and size, but they may not be exactly the same, or are the motors exactly the same. That is why I requested the second motor test, to see if it draws around the same as the first one, which I think it will be very close.

Ok, having said that..we can simply take some numbers from the list you provided. It takes X amount of watts to turn Y amount of rpm's. We can get a number, rpm's per volt (N). Simply divide the rpm's by the watts.

2 motors, 1 ESC:
14000 / (287.3/2) =97.45 rpm's per watt (average for two motors)

1 motor, 1 ESC
14050 / 147= 95.57 rpm's per watt

At a higher rpm (which would be less accurate due to stalled props):

2 motors, 1 ESC
18000/ (539/2)=66.79 RPW

1 motor, 1 ESC
17800 / 274.5= 64.85

You are getting slightly more rpm's per volt with 2 motors. This percentage is almost negligible due to inaccuracies in the test equipment, and possibly using one battery pack, and having different input voltages. Not to mention, motor #2 may be more efficient or the prop may have less drag due to slight imperfections, or balance issues, etc.

What I see is they are even, and work fine under a constant equal load. How about trying one with a much larger prop to simulate a exaggerated uneven load?

I admit, I don't know enough about ESC and exactly how they work to give some better ideas as to why we shouldn't use this in a real setup. I do have a question about ESC's, do they sense rpm's on the poles that are not energized to know when to apply current to the energized poles? Could this be the reason why if one motor is severely loaded up different that the other, you could have major issues? I suspect the efficiency of the 2 motor setup would go down as the loads change. Or, as the load on one motor increases, the sensing part of the ESC will slow both motors down. I realize there are sensor, and sensorless ESC, but I do not know the specific differences.

Good idea..............I will go over to the local hobby store and get a 7" prop. One motor on 6" and one motor on 7" props. I will use the Eagle tree but only have one RPM sensor. Which otor would you like to see that on? Or, as an option I could use my optical Cen-Tek hand-help optical tach and compare the RPM,s on each motor............but I am sure the speed control will keep them the same. Or am I sure of that?

[LiPo Power]

That is what I am saying....
You have enough motivation to keep it up with all the quastions and concerns....
Just wicked!!!!
Thanks

Thanks but it is the guys in the forum providing the good ideas. I just have the hardware handy and love to experiment.

[antslake]

Good idea..............I will go over to the local hobby store and get a 7" prop. One motor on 6" and one motor on 7" props. I will use the Eagle tree but only have one RPM sensor. Which otor would you like to see that on? Or, as an option I could use my optical Cen-Tek hand-help optical tach and compare the RPM,s on each motor............but I am sure the speed control will keep them the same. Or am I sure of that?

I guess use the optical tach, and measure both motors. Also test each motor individually with it's respective prop, so we can see the difference.

I've been doing a little reading on how ESC's work to try and understand this whole thing. It is as I assumed. I wanted to understand how the ESC knows how many pulses to send to the motor at a given throttle setting. Only any two of the three wires going to the motor are energized at a given time. The third wire is sensing the inductance created by the magnet going over the coil. Except on start up where the ESC just sends a generic signal to get the rotor going. The magnet passing over the unenergized coil does create enough current for the ESC to sense until it reaches a certain rpm.

So knowing that, everything would probably work fine if both motors were turning the same rpm(we've already kind of seen that). But if one were to get way out of phase, it would be receiving wrong timing signals from the ESC. What the affects of that would be, I don't know. Possibly higher amp draw, more stress on the FET's inside the ESC, maybe less amp draw, and a motor that starts freaking out. We already know that sometimes a stalled motor in a dual setup might not start sometimes, and you have to "reset" the throttle to zero to make them both start again. Maybe in a dual motor setup in a boat, the force of water going over a stalled motor might make it turn again?

The next step after this would be to make a setup that varies the load on one motor (variable pitch prop?) Because right now we are dealing with static setups. Just to further create a real world environment.

Great work, this thread is so interesting.

[sundog]

The next step after this would be to make a setup that varies the load on one motor (variable pitch prop?) Because right now we are dealing with static setups. Just to further create a real world environment.

Yup, that's what I was getting at on page one when I said "Try loading the unloaded motor (pinch the shaft?) while performing the same test. Will the submerged prop motor respond by slowing down to match the problem (pinched) motor?" I'm waiting for someone to try this theory in a boat. Could be tested much quicker. But it might fry an esc or motor (or two) just as quickly!

[antslake]

Maybe it would fry it quickly, or over time it would wear out the internals faster than using two speed controls.