Applying the JVN calculator using multiple motors and Mecanum

[Ether]

Originally Posted by pntbll1313

Can you just add the stall currents together for the CIM and mini CIM and use that number as if it were 1 motor or will that not work?

The stall torque of the combination is the sum of the individual stall torques.

The stall current of the combination is the sum of the individual stall currents.

But the free speed of the combination is not as simple. Follow the link in this post for the formula for free speed:

http://www.chiefdelphi.com/forums/sh...8&postcount=64

Note: since the free speeds of the CIM and miniCIM are so close, the free speed of the combination will be roughly the average of the two.

[Nemo]

One can put, for example, an 11 tooth pinion on the Mini-CIM and a 12 tooth on the CIM to even out the speeds a bit. Then you have to adjust for that in the math before you combine them. If you're saying the pinion gear is a 12 tooth in your gear ratio math, Multiply the Mini-CIM's stall torque by 12/11 and its free speed by 11/12 before combining them with the CIM stall torque and free speed.

[Chris is me]

Originally Posted by Nemo

One can put, for example, an 11 tooth pinion on the Mini-CIM and a 12 tooth on the CIM to even out the speeds a bit. Then you have to adjust for that in the math before you combine them. If you're saying the pinion gear is a 12 tooth in your gear ratio math, Multiply the Mini-CIM's stall torque by 12/11 and its free speed by 11/12 before combining them with the CIM stall torque and free speed.

From my experience, the motors actually seem to run a bit better with the same pinion on each motor. Matching the free speeds is less essential than common wisdom would have you believe - at operating speeds / torques the motors match quite well with the same gear ratio.

[Dad1279]

For a simplistic, proof of concept ballpark, you could run the spreadsheet with two CIMS, then two minis, and average the results.

Our last experience of combining motors was with 1CIM+1 775(with cimulator) per trans, worked well. I'd agree with Chris above. We considered testing and programming different values to the two motors, but in the end just sent them the same values.

[Ether]

Originally Posted by Chris is me

Matching the free speeds is less essential than common wisdom would have you believe - at operating speeds / torques the motors match quite well with the same gear ratio.

Yes. True for the CIM and mini CIM.

But to be clear, it's not true for motors with substantially different free speeds: the higher-speed motor may be at risk of overheating.

[Knufire]

Paul Coipoli did write a spreadsheet with the motor combination equations in the thread that Ether linked earlier:
http://www.fightingpi.org/Resources/...culations.xlsm

His MSC seminar also discusses the topic (link skips straight to the motor combining section): http://www.youtube.com/watch?feature...xpYnqJ8#t=1479

[Caleb Sykes]

It seems that question (1) has been definitively answered. Add the stall currents together to get the stall current of the 2-motor system.

I am curious about questions (2) and (3) though. Is that the value of CoF that is needed for this problem? On the AndyMark website, the forward/backward CoF is listed as 0.7 for mecanum wheels. Is that implied to be on FRC carpet? None of the plaction wheels have static CoFs listed, but all of the Rubber Treaded Wheels do.

If 0.7 is correct, and if you are not worried about drawing too much current, I think that you should gear it to be faster since you're not gaining any more pushing power.

On the same topic, does this mean that the max pushing force of ANY normal (i.e. with wheels) robot is ~155 pounds, assuming 155 lb robot, 0.93 carpet efficiency, and 1.07 CoF? I understand that many teams use their low gear so that their robot does not draw too much current, but having that upper bound on pushing ability is interesting.

[Ether]

Originally Posted by inkling16

...assuming 155 lb robot, 0.93 carpet efficiency, and 1.07 CoF?...

To avoid misunderstanding, what does "carpet efficiency" refer to in this context?

[pntbll1313]

Originally Posted by Ether

To avoid misunderstanding, what does "carpet efficiency" refer to in this context?

In my head that number was from rolling resistance because of carpet compression, but that shouldn't affect max tractive fore then should it? Maybe it wasn't supposed to be in that calculation. I found it in another paper too called wheel/floor efficiency http://www.chiefdelphi.com/media/papers/2750

Does that mean both of my calculations in parts 2 and 3 of my question are wrong?

[Ether]

Originally Posted by pntbll1313

In my head that number was from rolling resistance because of carpet compression...

That would have been my first guess. But I prefer not to guess. I'm trying to trace the etymology of the phrase back to its source.

I found it in another paper too called wheel/floor efficiency http://www.chiefdelphi.com/media/papers/2750

I did a quick search of both documents there and couldn't find it. Does anyone know: Who coined that phrase, and when and where? And in practical application, how is it supposed to be different from the widely used term "rolling resistance"?

In the drivetrain acceleration model located here,
rolling resistance is modeled with a constant parameter Kro and a speed-dependent parameter Krv.

[pntbll1313]

That .93 number is straight from the JVN spreadsheet. It is discussed here but again the OP in this thread is not entirely sure where it came from.

Originally Posted by Nemo

Right now I have a 0.85 gearbox efficiency and a 0.93 carpet to wheel efficiency built into the spreadsheet. What other terms eat up power and slow down the robot? I lifted that 0.93 directly from JVN’s spreadsheet, and I have no idea if it’s anywhere close to a realistic number.

http://www.chiefdelphi.com/forums/sh...d.php?t=110382

[Michael Hill]

Sorry for resurrecting a dead thread, but I've been curious how the motor combinations in the JVN spreadsheet (CIM + MiniCIM, etc.) were derived. They don't quite match up with Paul's spreadsheet. Unfortunately, there aren't any formulas in the JVN spreadsheet.

[JVN]

Originally Posted by Michael Hill

Sorry for resurrecting a dead thread, but I've been curious how the motor combinations in the JVN spreadsheet (CIM + MiniCIM, etc.) were derived. They don't quite match up with Paul's spreadsheet. Unfortunately, there aren't any formulas in the JVN spreadsheet.

Use Paul's numbers.
I actually meant to delete those out before I published it. Those are "close enough, but not very good" numbers.

[Paul Copioli]

Originally Posted by Knufire

Paul Coipoli did write a spreadsheet with the motor combination equations in the thread that Ether linked earlier:
http://www.fightingpi.org/Resources/...culations.xlsm

His MSC seminar also discusses the topic (link skips straight to the motor combining section): http://www.youtube.com/watch?feature...xpYnqJ8#t=1479

I am very confident in the combination equations in this spreadsheet. I developed it for the exact same reason the OP asked the question. These combinations and current draw calculations match up very well with my multiple motor dynomometer testing.

It does 2 and 3 motor gearboxes and also allows you to take a snapshot at any particular speed condition. One other subtle thing it does is allow you to put your entire gearbox gear ratio into the input section so the combined motor equation is your actual gearbox output equation.