For instance, if I had a drivetrain gearbox, and I put a brand new CIM, and a CIM that has been used on a drivetrain already, to what extent would the old motor hamper the performance of the gearbox? Would it cause the newer motor to wear out faster? Thanks for any replies.
When you have two gearboxes with motors, each on one side of a robot, the motor/gearbox setup that has run for less time will run slower.
If the left side has a new gear box with all new motors while the right side has broken in motors/gearbox, the robot will have a left drift because the right side runs faster. The solution to this would be to run the left side at full speed for a bit, then full speed in the other direction for the same(ish) amount of time. Just something to look out for when doing this. 1257 has assembled practice chassises with old/not-same-age motors/gearboxes and this is the only trouble that we have run into.
These motors are not precision instruments. There are many variables that will affect how much power and torque a particular motor puts out. Some examples are;
- friction in the bearings
- wire size tolerances
- strength of the permanent magnets
- contact pressure of the brushes
- contact area of the brushes **
I don’t recall ever seeing any sorts of tolerances on the motor output power and torque figures in the datasheets. Nor do I see any specification for matching between any two motors of the same type. My guess is that you might find 5-10% variation in output power/torque between two new motors from the same batch. Unless a motor is really worn out or damaged, the variability is such that an old one might actually put out more power and torque than a new one.
Wear out of the motors comes from time running so an older one will wear out before a new one if they are mounted side by side but it is rare to hear about these motors really wearing out.
Damage can occur if they are stalled for long enough. You would then damage all the motors connected to the same gearbox, regardless of their age.
** If a motor is made using carbon block brushes, a new motor will have less brush contact area than an older one that has been “run in”, leading to the older one being able to draw higher currents and putting out more power and torque, all other things being equal (but of course they are not).
I concur with both above. Unless a motor has been abused or is choked with dust or some such, putting an old and a new motor of the same type on the same gearbox should not cause any undue stress or wear.
It is more important to have balance in the total power and speed output between the two sides of your (tank) drive train; in my experience, this difference is more often due to friction and other issues in the gearing than differences in the motors themselves. In six years, we have had (as far as I can remember) five failed motors. Four were due to improper load/gearing*, and only one (a drive system CIM this year that flaked out during competition) was an apparently random failure – and that was a CIM with about a dozen hours of use, so more likely “infant mortality” than “died of old age”.
- The team’s rookie year (when I wasn’t a mentor, just a parent of a student who came early to do pickup 'cause this was fun stuff, and no one else was doing the math or properly recreating the game elements), we used a window motor on about a 28" moment arm to tip the bridge, missing by about two orders of magnitude. Two window motors disintegrated at Bayou. The next year, we tried using a BaneBots motor (well reduced, I thought) for the climber. On reanalysis, I found where I slipped a digit. This year during practice, we burned up a CIM due to some bad code (it was supposed to be running at about 10-20% but was running stalled at 100%).
Oh yes, for the 2014 season (Aerial Assault), Discobots built a butterfly drive train with a CIM and Mini-CIM on the same gear box at each corner. The rated free speeds for the two motors are different by almost 10% and the rated torque are different by about 70%. We didn’t have any problems with the drive motors.
It depends how much strain and wear the old motor had on it. If it’s really warn out I wouldn’t use it. If they show a difference when you are driving it ( Drifting to one side ) you can use PID control to even out the difference in most cases.
If a motor is made using carbon block brushes, a new motor will have less brush contact area than an older one that has been "run in
Man, I misread “motors” for “mentors” in the subject line. That’s an entirely different discussion… :eek:
The best course of action here is to use empirical data. Because, as it was stated before, motors are not precision controlled machines, you can see varying performance differences between two motors from the same lot (spinning different directions, different internal resistance, etc.).
But that’s all fine, as long as they both contribute generally equally to the gearbox in all cases.
The best course of action, when building a gearbox, is to bench test it by driving it with the same input voltage, and read the current across all the motors individually, and when running together. Unloaded, and under load.
This will give you the data you need to say “yep, all motors are performing equally, or nope one motor is doing more work than the other”, and chase down if it is a friction issue, resistance issue, or a motor issue.
Generally we like to keep a 5% margin between motors.
The monitoring and logging can be done by your own software, but the Driverstation logs already monitor current on every channel, so all you need to do is run your robot, then view the logs.
In my honest opinion, this is the best way to be reasonably sure your gearboxes are performing as intended.
Also check these logs periodically, you can determine failures before they occur if you see over time the current draw is increasing, or even identify failures like tripping snap action breakers (current, then sharp no current, then current).
Comparing two gearboxes in this way, i.e left and right side of a drivetrain, or a replacement spare help identify if one was built better than another.
Also Comparing these values, to any gearbox calculator you may have used is also good to do, so you can find assumption issues with your calculator to help make them more accurate over time, with real world data.
Hope this helps,
Kevin