Heating issues (Locked-Antiphase) with RX-28 and 18v25CS

Hi all,

I’m having some heating issues that I’m hoping someone could shed some light into.

I’m currently powering a stripped down Dynamixel RX-28 actuator (i.e. barebones DC motor) via the Pololu 18v25CS driver.
The PWM is running at 32MHz using the locked-antiphase method connected to a 16-18V battery (NiMH, 8.4V 7-cell 4500mAh x 2).

I’ve noticed that after about 3-5 minutes of run-time, the actuator shell is quite hot and thus, the motor is extremely hot. If I use the signed-magnitude method, the actuator runs just fine with no heating issues at all.

Is this expected? Thank you.


Did you mean to type 32 kHz, or are you actually using a 32 MHz PWM? If you just supply it with a 50% duty cycle PWM, is the motor completely stationary, and does it still get hot?

- Ben

Hi Ben,

Yes, my bad. Its 32 kHz PWM that I’m supplying.

At 50% duty, the motor is stationary. However, it still gets hot eventually.


That makes it sound to me like your PWM frequency isn’t high enough for your motor, though I’m not sure the driver you are using will work well at much higher frequencies, so you might need to go with sign/magnitude control. In locked-antiphase, you are rapidly switching between driving the motor full-speed forward and full-speed reverse, and you are counting on the inductance of the motor to average this out so you get a relatively constant current (at 50% duty cycle, the motor inductance should keep the current at zero). If your switching rate isn’t fast enough for your motor inductance, the inertia of the motor might still average the motion out and give you the desired motor speed (so things look like they’re working fine), but the current draw could end up being closer to the stall current of your motor, which isn’t good. Does this make sense?

- Ben