Expected Resistance in Drivetrains

We are currently finishing up our drivetrain for this year’s game, and the question that I had was what was the expected force loss for a standard drive train? As in, calculating the expected force that the motors should output if the motors were free spinning, and then comparing it to the actual output of the motors when running with normal load. Right now we have about a 49% power loss, which may or may not be reasonable. I guess in my mind 50% power loss is probably the maximum I would expect, and 25% is probably the minimum power loss that I would expect.

If anyone has any idea what the standard power loss is approximately, it would be much appreciated.


After testing the power loss with a dummy bot (direct driven 1:1, 4 393’s) it seems that the dummy bot loses about 21% of the expected power… If it were heavier I guess the percent loss would be more around 25-30% loss, just going now to test it by putting sacs onto the dummy bot so that it has equal weight to our actual robot (13.2 pounds) and we shall see how that changes things.


After testing with a heavier dummybot, it seems that it now loses about 71% of the expected freespin power. So it would seem that the weight is our main problem here… Any comments would still be appreciated.

How are you measuring power output?

I assume they’re measuring in In./Lbs.

uhh i read in some article (long time ago)
that gears transmits 90% power
and chain 80%
and bevel gears 70% :confused:

if you get 50%, then theres DEFINITELY something wrong…
(either excess friction or measuring it wrong)

Is that with 2 gears/sprockets? or does that scale with compound gearing?

not sure, ill see if i can dig up the article again

It scales with compound systems, e.g. 90% * 80% = 72%.
This is likely to be only a small part of the power loss, other things like friction on shafts and friction from the compressibility of the foam will mean it all compounds together. Still, 50% sounds too high a loss to me. I’ve never done any measurements though.

It sounds like you are mixing two or more things.
If you have IME or optical encoders, you can measure the free (unloaded) speed of just a bare motor shaft. That’s one baseline. Whats that number?

Then you can put your robot up on blocks, and measure the free (unloaded) speed of the entire drive train, including spinning the wheels in the air.

Then you can check the loaded speed of the robot while driving on the ground, under various conditions.
Some of the speed loss is in the drive train.
Some is lost in friction with the Vexfoam flooring.
And some of the initial power is used to accelerate your robot, but I wouldn’t call that “lost”.

Most posts here report speed based on gear ratios times the Vex speed chart baseline.
Few people report actual max robot speed in fps, and fewer or none under repeatable conditions, so your data may be new and useful if you post the details. Disco-bots had an old report that they had optimized their drive train to better than expected to be possible by Vex speed-chart data, as measured by wall-to-wall times.