[FTC]: Experience with new REV servos?

We’re trying to run a pivoting arm with a REV servo with a ServoCity servo gear driving a larger gear at a 10:1 gear ratio in continuous rotation mode. According to the REV web site, its stall torque is 187.8 oz-in so we’d thing we’d have 1878 oz-in to work with. However, the arm stalls and we’re afraid we’ll kill the servo.

We were powering it with a Neverrest 40 motor at a 3:1 gear ratio, and it has 350 oz-in of torque (so that’s 1050 oz-in). It worked fine when set up this way.

We think that 1878 > 1050, so we aren’t sure why the geared servo didn’t work as a replacement for the motor.

We powered the servo for the test with the REV servo programmer, so maybe that is an issue, but we’re not sure. We ran in it “test” mode for continuous rotation, and it couldn’t fully lift our arm.

Has anyone else tried these servos? (Can the GDC make an exception so we can use 9 motors instead of 8? :slight_smile: )

I don’t see anything obvious, so let’s start with the basics.

  • Double check your math. If you give us some model numbers and such for the servo and reduction system, we can do this as well. Also compare the peak current draw of the servo against the peak supply of the programmer.
  • Check the batteries in the servo programmer. You never know.
  • Run the servo with no load (not connected to the arm) using the servo programmer. Is the top speed about right? If you can check it easily (a load on an arm) is the top torque about right?
  • Check the 10:1 gear reduction for binding or other inefficiencies. Can you operate it manually, and if so, does it go smoothly? The loss in a 10:1 gear box should only be 10-25% depending on type and quality (better for simple spurs or planetary, not so good for bevel gears or worm).

Sure, they can. But I’d bet you lunch they won’t.

It’s a 12-tooth gear (https://www.servocity.com/html/32p_futaba_metal_gear.html) meshed with an 84-tooth, so I mis-typed (it’s a 7:1 gear ratio – 1314 oz-in total) mounted with an Actobotics channel and ball bearings for the larger gear.

REV Servo: http://www.revrobotics.com/product/smart-robot-servo/

REV Programmer: http://www.revrobotics.com/product/srs-programmer/

Thank you for the advice, and we’ll check the speed and torque (and swap the programmer batteries) tomorrow!

We are thinking about adding a 2nd servo to the other side of the large gear to double the torque. Servo City is closed this week, though, so we can’t get another one of those small gears, yet.

Yes, that’s a lot closer to what gear inefficiencies could cause a problem. A second servo would put you back into solid territory assuming the NeveRest numbers were right.

Here I found an issue. The Servo wants 2A near stall (that is, when delivering maximum torque). While the programmer does not list a maximum current, it is powered by 4 AA batteries, apparently in series, as the output is nominally 6V. This means that the output current of the programmer is limited to the current of one AA battery. When drawing 2A, a fresh alkaline AA battery drops to an average of about 1.25V. This will reduce the peak torque. Using a second servo (with a second programmer) should resolve the issue. Alternately, you may be able to figure out a way to use D cells or parallel AA cells with the programmer to bring the current draw for each cell down to about 1A, which will give 1.4A for 2x AAs, or probably a bit more for D cells.

Also check the current capabilities of the 6V circuits on your regular robot control system. (I’m not familiar with the FTC systems.) Is it up to delivering 2A here, plus whatever else you will need at the same time?

As I stated above, this might work if you also use a second programmer. The key is probably getting 2A at a full 6V total to the servos.

It will be attached to a Modern Robotics servo controller (http://www.modernroboticsinc.com/core-servo-controller-3) which says 5A max, so maybe we just need to test it under actual conditions with the robot electronics instead of just with the programmer?

Yes, That would certainly bypass the 2A problem quickly.

I just purchased two of these servos and the programmer for them as well. Glad to read the items here - you’ve probably already solved a few issues teams I would with might’ve encountered!

Really looking forward to swapping between continuous rotation and angular mode.

We checked our math again last night with weights of everything on the arm, and it should have taken less than 200 oz-in to pivot. So, we put fresh batteries in the programmer and it worked perfectly. Funny since the first batteries were new but I guess we had a bad batch.

Thanks again for the help. We think the programmable servos are pretty cool and will be writing the software for them next week.

You’re welcome, and my pleasure. I always like when I have to learn a bit to answer a question.