I'd read somewhere on the net that you can increase the performance of a servo by upping the command frequency and also by increasing the supply volts. It's logical. This could be especially important to a walker with potentially high torque requirements. I'd already killed one servo whilst trying to get the bit-banging code right, but I thought it would be worth looking into. I best point out that my equipment is not calibrated, so the results are approximate, but these are comparative tests anyway.
I took an economy GWS servo and clamped it, with a spring balance hung from the arm and hooked to the clamp base. The distance from the servo pivot to the spring balance was 13mm (a 13mm lever). Using my Servo Checker (Milford Instruments) and a supply set to 4.8V, I powered up the servo, read the pull force and checked my meters.
I repeated the test at 6.0V and recorded the data.
Next, I hooked up the servo to the PIC that would over-clock it at 97.8Hz (the servo didn't seem to like 100Hz - I'll have to try higher frequencies again sometime). The PSU is set to 4.8V and the servo re-tested. Then tested again at 6.0V. Obviously the speed had increased noticeably with the higher command frequency, but it is difficult to measure.
I then tried over-volting to 7.2V but there was no noticeable performance increase, just more current drawn. See table for results.
Not bad, I thought, for a quick investigation. The results are quite surprising. Maybe I'll try pushing it a little further. Anyone know how far you can push a hobby servo? Email me if you do.
