Marty stumble and seized hip servo

I took Marty to work today for a day out. I was doing a few demonstrations of wiggle and kick in Scratch, I probably did that about 20 times and he then surprised us by toppling over. After a bit of investigation it turned out his right hip servo (K-Power DMM016 Servo) got stuck at an approximately 45 degree position backwards. I've just disassembled the top of right leg to check and the servo and it does appear to have seized. I've not tried anything too forceful, is there anything you can recommend to recover this? Do I need a new one?

Does Marty auto detect failures like this? I was in a noisy room so may not have heard any beeping. It would be useful if he had a distinctive, documented special beep for detected servo failures either during operation or start-up.

What would the current graph over time look like for a servo stuck?

Discussion

Leave a comment and join the discussion here


Please Log In or Sign Up to write a comment response

5 Responses

kjw OP

Yes, in my case it was one silver and black servo.

15:18:57, 14th December 2017   |   Permalink   |   View Source
angus STAFF

@kjw just to double check, this is about the silver and black metal geared servos, correct?

We did see quite a few issues with these when we were testing and calibrating them, I think we caught most of the defective ones but it does look like a couple slipped through. We did indeed give the manufacturer some feedback, and the new load of the motors we've got in right now for our current run have a new, chunkier limit nub in there to (hopefully) prevent stuff floating around inside from causing jamming

14:58:09, 14th December 2017   |   Permalink   |   View Source
kjw OP

My Marty is still ok post servo surgery.

Looking back at other forum posts I see similar problems: Leg servo refuses to move

For mine it could have been a manufacturing problem with either too much torque being used for the screws which hold the servo together or weaker than expected plastic. Either of those could cause the thread to be stripped from the hold and that swarf might then end up bouncing around gearbox til it seizes the gears? If you are getting more than the expected failure rate here it would be worth opening them up for close inspection and giving some feedback to supplier/manufacturer.

14:52:23, 14th December 2017   |   Permalink   |   View Source
kjw OP

I removed servo from Marty and then removed the horn screw and the two thing long ones that hold it together. The insides are roughly what I expected, the gears didn't want to move and then did move after some wiggling. Whilst I was examining it a small piece of black plastic did fall onto my finger, like a tiny lasso with three rings but I had not noticed that jamming anything. I noticed the thin long screws went back and didn't seem to find much resistance, perhaps the two things are linked. Anyway, I think it'll be ok now as re-assembled servo can be moved with normal pressure. I'll put it back together and see how it behaves.

21:11:40, 22nd October 2017  *   |   Permalink   |   View Source
angus STAFF

@kjw Marty does have some degree of detection for over-current detection and prevention built in. If a motor is drawing too much current, motor has seized or is being asked to push against something too hard, Rick makes a sort of ringing-bell noise, and will cut power to the motor after a couple of seconds.

If you've got a seized motor then some wiggling with a servo horn can help. If that doesn't fix it, we certainly can send you a replacement :)

18:44:08, 21st October 2017   |   Permalink   |   View Source