Lynxmotion Tech Support

unfortunately...

i've built two arms with the SES pretty much the same as those shown on the product page of the lynxmotion site apart from that they have 6" tubes between servos and no grip attachment, having a very very light scoop instead. the arms are controlled by a very old SSC-12 v1 powered by a 7.5V 4A supply. i'm using 645MG servos for the swivel and two main motors and 625MGs for the other two further out.

after about 20 mins of use with a heavy load on the servos a couple of times and periods in a 'rest' position some of the servos get very very hot. in particular the exterior one of the two at the base and the one at the first joint. i accept that these are the servos that are working the hardest (though why just one of the base ones heats up i don't know...) but i'm surprised they get so hot - too hot to touch actually. i've already burnt one out unfortunately.

does anyone else have over-heating problems? is it unreasonable to expect my arms to work every 15 or 20 mins for 6 hours per day?

they are going to be in a space a long way from home for 6 weeks so i'm wondering about solutions. maybe i should put in a timer that powers the whole thing down for 15 mins every half hour? but then i get the big jumps when they power up again. why can't robotics be really really easy??

any views on this would be very much appreciated.

cheers

You're probably running into the standard problem of too many volts over too much time.
Servos weren't made for the type of pain which we put them through.

If you want to improve the longevity of your servos, keep them down at 6.0V, which is what they're rated for.
If you do chose to overvolt the servos, be prepared for the types of things that you've experienced.
Even at 6.0V, you'll find that you still run into the same sort of problems (although less frequently, and less drastically).

Again, it comes down to the servos not being made for this type of wear and tear.
If you have the kind of money to spend on two servo arms, you might think of purchasing one arm with better servos, rather than two weaker ones.

When you overvolt a motor of any type, you emphasize both it's strengths and it's weaknesses.
Any flaws in the manufacturing of that motor will become more prominent with the more voltage and strain that you put on them.

If you're going to continue to overvolt them, I'd suggest frequent periods of rest.
You can also try cooling them down every so often with ice-packs from your freezer (the soft, non-leaking, variety, of course).
Be sure to wrap the ice-packs in a paper towel just in case of minor condensation.

To be honest, overvolting should really only be done with toy RC car motors.
Those are a couple bucks a pop.

_________________
Disclaimer: The above person does not claim to have knowledge pertaining to the following subject: anything. Thus, said person may not be held liable for any mishaps/explosions that his advice incurs.

hey nick

thanks for following me around the forum answering my questions!!

i'm going to go and hunt a 6V supply with a decent amounf of amps and see how i go from there. maybe a combo of that and working out how i can rest them now and again is the 'answer'. maybe i'll even hack a computer fan onto the arm base and aim it at the lift servos...

i needed two arms for the project i'm working on so there was inevitably a trade off with the quality of servos i could get - even having said that the 645MGs aren't exactly bargain basement. not in europe at any rate!

ok - one last question about my servo/heat/ssc/power/volt/amp/robot arm obsession...

so i'm going to get a new 6V power supply tomorrow which'll hopefully diminish my heating problem somewhat. the 7.5V one i have now is 4A which seems to be plenty of juice for 10 servos, but i've just been wandering about the forum and reading of people using 1600mA for 12 servos.

how many amps would people say i'm going to need here? or thinking ahead, is there any rule of thumb, equation, mystic process that will enable me to calculate what i'll need for however many servos in whatever configuration?

thanks again.

It depends on what servos you are using. I am using the Hitec HS-475s. If you do a bit of searching, you come up with the spec sheet.
http://www.hitecrcd.com/Servos/spec_sheets/HS475HB.pdf
It states there that the stall current for the 475 at 6.0V supply is 1100 mA. So, if you are using 10 of those, you could draw as much as 11A if all 10 servos are stalled out.
I know that for my projects, I tried to use a 1.6A 6.0V supply to run my 18 servos, and it just couldn't handle it. For your 10 servo setup, you should probably go with at least .5A per servo for a total of 5A.

- Stephen

You may have been getting confused with mA and mAh.
I know that I personally have used 1600mAh batteries to run twelve servos, and I plan upon using 1600mAh batteries to run nineteen, as soon as I get the rest of the servos that I need -- at least, until I get the power supply and the LiPo batteries that I am looking at getting.

I think that mAh has little to do with the max amperage that can be drawn from the battery at any time.
I have no clue what the discharge peak of the lynxmotion NiMh batteries that I use are, but I would imagine that they're a high discharge battery, as I've had no problems with them.
I guess that would put them at around 5-10 max amp discharge.

Hopefully, I haven't lied in the above text, but it's very possible.

_________________
Disclaimer: The above person does not claim to have knowledge pertaining to the following subject: anything. Thus, said person may not be held liable for any mishaps/explosions that his advice incurs.

You may want to add some load balancing springs to help relieve some of the load. If one of the shoulder servos is hotter than the other, that means it's doing more work, or you don't have the two servos aligned properly. Both servos have to be either mechanically aligned, or you have to upgrade to the SSC-32 for its pulse offset (PO) command to correct for the missalignment in software. To do this mechanically you directly connect the right servo (the robot's right) to the "C" bracket then power both servos up and position them to center (1.5mS). Then drill out two holes in the left servo horn to allow connecting the "C" bracket to the horn with 2-56 hardware. It may be necessary to pull the round servo horn off the final gear and rotate it to find a clear place to drill the holes.


Strong servos have to be used with a duty cycle when under load. A powerful servo under significant load will overheat in as little as 10 minutes. You mentioned they would be used every 15 to 20 minutes for 6 hours a day, but you didn't mention how long they will be working at a time.

_________________
Jim Frye, the Robot Guy
http://www.lynxmotion.com
I've always tried to do my best...

well i've replaced the power supply with a 5V 5A one and the servos don't heat up so much, although eventually they get pretty toasty...

the arms will probably be working for about 2 minutes every 10 minutes or so. i'm trying discover a position that will enable them to rest when not working and have set this so far as the position the servos go to on power up. its better but i feel there must be more i can do. i've been thinking about powering them down completely every 20 minutes or so for 5 or 10 minutes but powerig them all up again at the same time causes the inevitable and possibly damaging jerking about.

i thought i'd done pretty well about lining up the two shoulder servos but i guess they must still be a tiny bit out.

Well, as for that inevitable jerking, it looked like they were getting somewhere in this thread:
http://www.lynxmotion.net/viewtopic.php?t=236

But, then, it trailed off into the realms of cyber muck...
Perhaps we can revive that topic?

_________________
Disclaimer: The above person does not claim to have knowledge pertaining to the following subject: anything. Thus, said person may not be held liable for any mishaps/explosions that his advice incurs.