OpenServo.com

m_kanter · Joined: 19 Apr 2006 Posts: 31

Hello,

i'm trying some things out with the h-bridge and micro servos.
I've removed the electronics of an micro servo (TOP-LINE Mini Servo ES-05JR from Conrad Elektronik) and put an controller and h-bridge onto it. It moves correct and the measurement of the position works too.

The low-side of the h-bridge is connected to normal ports and the high-side is connected through an mosfet driver to the OCx pins.

My problem is that the motor becomes very hot in a short time, even with no load on the gearbox. Can this be caused by the switching of the mosfets?
Or can is the motor too small and overpowered with the 5v (in the original servo specs it is rated to 6v)?

Did you encounter such problems with other servo motors?

Best Regards,
Marcel

jharvey · Posted: Fri Aug 28, 2009 12:22 am Post subject:

It's hard to say with the information you gave. It's possible that it's switching at a very rapid pace, forward then reverse, causing it to simply dissipate heat. I'm not familiar with the servo you noted, I'd guess it's OK for 5 V operation.

I know the OSV3 will hum if you don't have it spot on. Does your's make a noise when it's unloaded? If so it's likely switching very fast F vs R.

m_kanter · Joined: 19 Apr 2006 Posts: 31

the software currently let the servo move from 0x1000 to 0xf000 (the poti position is left adjusted) this is nearly 180° and then it makes a short pause of ca. 400ms and moves to 0x1000.

the pwm has an deadband of 30/255 and the motor brakes (both high side mosfets 100% time on) within the deadband. so there is no noise or fast switching from f to r.

the switching frequency is at 15kHz which is no problem for the mosfet driver (it is rated up to 200k). maybe for the motor itself? its an ca. 6mm diameter motor like from old fashioned vibration of mobile phones (it looks like that).

ginge · Posted: Fri Aug 28, 2009 9:17 pm Post subject:

Hi m_kanter,

This sounds like you are switching the FETs too quickly. Do the FETs get hot?

I would look at changing the switching delay in the codebase. Start high and work down until the temperature is acceptable.

You might also want to check flyback conditions and possibly add some hefty flyback diodes so you dont dump the current back through the bridge.

Without more information it is hard to say for certain.

Cheers
_________________
http://www.headfuzz.co.uk/
http://www.robotfuzz.co.uk/

m_kanter · Joined: 19 Apr 2006 Posts: 31

The mosfets get not hot, when the servo is not blocked.

Ok, guys the servo is crap. the motor seems to be rated for 3v or so only. i've intalled the circuit onto an hs-65 and let it run a bit. the motor runs very quickly and does not get warm.

i'll put all together into the case and will report later. (maybe with current measurements)

thanks for the tips