17
$\begingroup$

For my robot, I am using two continuous rotation servos to spin a threaded rod. I am trying to make this project as cheap as possible. Here are the servos that I can find:

  • Servo #1: This is a very cheap option and it has half of the torque I need.
  • Servo #2: This has all of the torque my project requires, but it is much more expensive that two of servo #1.

Can I hook up two of servo #1 to each end of the rod and have them move synchronized? I can spare a few extra pins on my microprocessor that I am using; that isn't a issue. I know hooking two together will increase torque, but I don't want 75% of the torque I want in this situation. Also, I don't care if I only have 98% of my torque "goal" with the extra weight (which probably won't happen) but I don't want to, like I said earlier, have 70, 80, 90% of my "target goal" of torque if possible.

Any help appreciated. Thanks in advance.

$\endgroup$
10
  • 1
    $\begingroup$ yeah, this is why I quit hanging around. All pilots who get paid to fly by the RC manufacturers as professional representative, use double servos. It increases efficiency, reduces the chance of failure, and improves the life of each servo. Just saying. !enter image description here Double servo trays $\endgroup$
    – Spiked3
    Feb 24, 2013 at 9:54
  • $\begingroup$ Are there any specific manufacturing tolerances etc on those servos to mitigate the issues mentioned in the other answers? $\endgroup$
    – Joe Baker
    Feb 24, 2013 at 19:02
  • $\begingroup$ I never knew something like that existed. I wonder how you could build that... huh. It doesn't look that hard once you get all of the parts. I'm assuming this will work with continuous rotation servo. Thanks and awesome idea. EDIT: Would this work?: youtube.com/watch?v=jqsmai2Nafk The only thing I don't understand is that in Method 1 [of 2] it said "Use only if the servos travel the same distances. (Most don't!)" What does that mean? The reason I ask is it must have some benefit and it seams a little easier to build... but I wouldn't mind if it was better (reliability, etc.). $\endgroup$ Feb 25, 2013 at 0:53
  • $\begingroup$ Okay... I just read the description... not good. "3 servo failures in two weeks." Oh well. Back to square one. $\endgroup$ Feb 25, 2013 at 0:55
  • $\begingroup$ This fits with ApockofFork's answer, specifically not using a hard linkage between the two servos. Relevant: youtube.com/watch?v=3fDyULSO8KU $\endgroup$
    – Ian
    Feb 25, 2013 at 18:16

4 Answers 4

18
$\begingroup$

In theory, you're correct. But in practice, slight differences between the two motors will make them fight each other instead of working in perfect harmony:

Even the smallest mismatch in wire tolerances, lengths etc. will create different back emf characteristics due to different impedances.

So both will work separately, but the actual characteristics of the 2 motor within the very small timeframes of a "minimum" move at relatively high speed will not be in synch .. and thus the 2 motors fight each other.

For this reason, your plan to save money by connecting a second servo to your microcontroller will probably cost you money. Not only are you shortening the life of the servos, you're doubling your chance of a servo failure. Worse, the failure of one servo may cascade to the other unless you are using a mechanical linkage (as ApockofFork's answer recommends), e.g. the one in this video.

With that said, connecting multiple servos with a fixed linkage is done all the time, e.g. RC plane rudders, RC plane ailerons. But to avoid the problems cited earlier, it requires an additional control component that can compensate for the differences in the electrical characteristics of the servos. These are commonly called "servo equalizers", "matchboxes", or "servo synchronizers".

$\endgroup$
2
  • 1
    $\begingroup$ I guess saving money doesn't always work. Thanks! $\endgroup$ Feb 18, 2013 at 16:01
  • 3
    $\begingroup$ This answer only concerns putting both motors on the same shaft. ApockofFork's answer might be relevant to you if you can use a more complex linkage. $\endgroup$
    – Ian
    Feb 20, 2013 at 20:26
11
$\begingroup$

What Ians answer says about the two motors fighting is true however, I wouldn't go as far to say that you can never put two motors together for more power. In general if the two motors are identical and you are hooking them up and controlling them identically then you can probably get away with it but I would expect some wasted energy and thus extra heat from the two motors fighting a little bit. You will also not get exactly two times the output.

Worst of all it will probably hurt the lifespan of the motors. In other words it will probably do in a pinch but is definitely not ideal. If you really do, do a good job syncing them up then you can theoretically keep the losses to a manageable level.

If you are going to use two motors on the same output you will probably want to gear them together or use some other linkage that isn't fixed (I.E. don't attach them both to the same shaft). That way there is some looseness between them so they will fight each other a little less.

$\endgroup$
3
  • 6
    $\begingroup$ +1 for reduced lifespan. Also, I'm curious whether running a differential gear setup in reverse (motors where the wheels should be) would be a possible solution here. $\endgroup$
    – Ian
    Feb 20, 2013 at 20:23
  • $\begingroup$ @Ian yep a differential connected in that way should work great $\endgroup$ Feb 21, 2013 at 3:59
  • $\begingroup$ I believe a planetary gear system would also work for outputting a combination of two rotary inputs. $\endgroup$
    – Robz
    Aug 4, 2014 at 0:42
4
$\begingroup$

Put some value on your time and on system reliability. You're increasing design time & failure points with this approach. Good small gear motors are pretty cheap to start with, it's a lot more valuable to have a working robot that's reliable than a broken or unfinished robot on the shelf.

$\endgroup$
1
$\begingroup$

This guy put a spring in between his 2nd servo to protect them from any slight differences... Maybe this is what you're looking for? http://youtu.be/jqsmai2Nafk?t=1m13s

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.