Hard-coupling PF motors experiment

[nerdsforprez]

Hello All -

As Technic MOCs get bigger and badder, with improved controllers and
battery sources powering them, I have noticed that hard-coupling motors together in most supercars and large construction-themed Technic MOCs is becoming more and more common.  Now, Philo has shown that there really is no difference in RPM and power between hard-coupling motors or coupling them through an adder http://www.philohome.com/motors/motorcomp.htm  Despite this, the assumption that Lego motors run at similar RPMs is still important to consider when hard-coupling them because if one is slower than the other it can degrade performance over time, and perhaps even permanently damage a motor (another good source of information is Sariel’s book, 1^st version, The Unofficial Lego Technic Builder’s Guide, Chapter 18).   

During a recent  build of mine I was challenged to really think about how consistent motors are when I accidentally noticed how inaccurate the RPMs were when running two M motors off a single PF battery pack.  I noticed that even with the naked eye I could detect difference in their RPMs.  This led me to
ask "How much difference is there really?" and to wonder if more damage than
previously thought is caused by coupling PF motors.  This led to the little
experiment below.

(Note: caution when setting the volume; set it fairly low and adjust from there.  It changes greatly during the film as I slowed down, then ran it at regular speed.  Usually I run a song or something over the audio but not here, I felt the audio an important part of the actual content of the experiment). 

 

 

Little information of the experiment.  4 M motors, and 4 XL motors were compared against each other for speed.   10 seconds each trial.  Trials compared each motor, one against the other.  Example:  1 versus 2, 1 versus 3, 1 vs. 4, 2 vs. 3, etc. (you get the picture).  RPMs for the same motor, but different trials slightly differed because although I tried to run each trial for exactly 10 seconds, when I examined the film with digital software I noticed that there were slight deviations in the trial time (10.4, 9.6 seconds, etc.).  These deviations are so small, but when you are running motors at several hundred RPMs then even  fraction of a second can result in several additional turns.  These little deviations did not matter because trials were independent.  Winners in each bracket (M motors, XL motors) beat all other motors in independent trials. 

As can be seen in the video, the differences were not huge.  M motors and XL motors were 100% similar in their average inconsistency (5.5 percent).  To summarize, I think this is little variability and likely really does not lead to much damage or decrease in overall output (same conclusion made by Sariel and Philo; referenced above).  Two things to consider however, and I think are new contributions not made by others, is that although there really is only 5.5% difference (average) between motors, when they operate at like 405 RPMs (M motor, Philo for reference) then we are still talking about quite a bit of difference, at least in revolutions per minute between differing motors (~23).  When you put it that way, the difference seems  a little more significant.  Lastly, if one is going to hard-couple two motors together, he/she may want to examine the speed of both independently first.  Although average difference is not that bad, at the tail ends of the distribution large differences can exist.  The largest difference in my little examination in the M and XL motors was around 9%.  For M motors that is around 40 RPMs different and for XL motors that is around 20…. Which really sound bad, especially when considering over the long haul!  So, check motors for reliability first before hard-coupling them, you don’t want to get combinations such as 1 vs. 2 in my experiment, where unreliability between motors is fairly high!

[mocbuild101]

Hard-coupling motors is not really that bad, because the speeding up and slowing down of the motors is nothing different from what they go through when used separately - consider the motors in a car or truck, they are slowed down by the weight of the vehicle when accelerating, and the are forcibly turned by the momentum of the vehicle when stopping. 

So, there is nothing to worry about when hard-coupling motors - all you need to do is make sure they are the same type of motor, and that they are geared the same - nothing else!

[nerdsforprez]

5 hours ago, mocbuild101 said:

Hard-coupling motors is not really that bad, because the speeding up and slowing down of the motors is nothing different from what they go through when used separately - consider the motors in a car or truck, they are slowed down by the weight of the vehicle when accelerating, and the are forcibly turned by the momentum of the vehicle when stopping. 

So, there is nothing to worry about when hard-coupling motors - all you need to do is make sure they are the same type of motor, and that they are geared the same - nothing else!

Not entirely true.  As demonstrated in the video, the forces on hard-coupling motors is much greater than when in regular use, or even the forces from regular stopping models.  Stopping models does not constitute whole rotations, let alone 40 in  a minute.   Not to mention hard-coupling motors causes resistance on top of the resistance from regular use (like those forces you mentioned, weight of vehicle etc.) which can degrade performance over time.

I admit, as mentioned in the video though, the effects are likely very small, but not non-existent.  These are essentially the same conclusions arrived at by others.  But I do think the main point still stands; when using two motors hard-coupled together, I would check first to see how similar they are in RPM.  I know in my project, I will be using the XL motors that were 100% in agreement one with another.  In cases such as that, I can be reassured there is essentially 0 problem because they run at exactly the same speed. 

[BusterHaus]

Very well thought out experiment. What power source did you use? I believe @Philo uses a regulated power source instead of a battery box to avoid voltage variations. 

[mocbuild101]

4 hours ago, nerdsforprez said:

the forces on hard-coupling motors is much greater than when in regular use

Yeah, I guess so, but it's still not that different...

Also, the different speeds of the motors are most likely caused by the age difference of the motors - I would be good to know the production codes of the motors, and what RPM those motors turned at, to see if I'm right.

[nerdsforprez]

9 hours ago, BusterHaus said:

Very well thought out experiment. What power source did you use? I believe @Philo uses a regulated power source instead of a battery box to avoid voltage variations. 

Good point.  I don't own a regulated power source, so just a regular pf battery pack was used but I actually think it better that way, as results are more generalizable that way. When any of us build a moc, we don't use regulated sources, just good ol' battery packs.  Therefore, results more applicable to what we actually experience when building a moc.

[BusterHaus]

I agree that the power source is irrelevant - you set up the experiment in such a way that apples got compared to apples. I was just curious about the power source. I made a DC motors vs servo motors video a while back and it also shows the same discrepancy between DC motors, minus the detailed analysis of freeze-frame. I also used a battery pack - it's good enough to drive home the concept. 

[Captainowie]

No, the power source is not irrelevant - some types of batteries have a significant voltage dropoff as they start to lose charge. When you're measuring such small differences between motors, that might make a measurable difference, especially between the first and last tests.

In any case, an easy way to test two motors against each other would be to have each motor driving one side of a differential - the speed of the diff cage will be proportional to the difference in speed of the motors.

Owen.