Suppression capacitors reduce effectiveness of PWM

I am trying to control a small DC motor for an 00 Gauge model train powered from a 1S LiPo.

I have an optical detector that measures the speed and allows PID control. I had installed the usual 3 x 0.1µF capacitors on the motor terminals to eliminate electrical noise that was causing spurious signals on the optical detector input to the Arduino.

That all seemed to be working nicely. Recently however, as the project has come closer to completion, I have been examining the behaviour more closely and it is obvious that the low speed is not low enough.

Then I discovered by accident (because I used another (identical) motor with no capacitors and no PID system) that with high frequency (31k) PWM the motor would perform very well at low speed.

Then, when I added capacitors, the low speed ability disappeared - presumably because the capacitors are smoothing the PWM. This happens even with just a single capacitor across the motor terminals.

I have Googled (well, DuckDuckGOed) but have not found anything that addresses this problem. I wonder if anyone here has any bright ideas.

I guess it is possible that the PID system would be unaffected by the un-suppressed motor power system at high PWM frequency as I have not yet tried that - doing so is very inconvenient and it would be nice to get some guidance to save a day or two of "suck it and see".

Hope this makes sense. Thanks.

...R

If you used shielded wire to power the motors, that may provide enough capacitance to eliminate the noise and then you would not need the caps.

Paul

Thanks. I am not dismissing the idea but I don't think I have space for shielded wire. The insulated wire I have been using is just 0.6mm diameter. Also the wire length is short - about 4 or 5cm - so manipulating it for soldering is not easy.

...R

An 0.1 uF capacitor has impedance of about 50 Ohms at 31.5 kHz. Try 0.01uF or even smaller -- that will still reduce RF radiation.

I now have it working nicely without any suppression capacitors. I twisted the two wires together that bring power to the motor but I don't know if that made any difference as I have not tried it with non-twisted wires.

I'm sure the lengths of different wires and their positioning is very different from the situation (months ago) when I originally solved the interference problem by adding capacitors. There are too many potential variables to waste time trying to find out what was the critical change that has enabled it to work.

Thanks for the input.

...R

Robin2:
I now have it working nicely without any suppression capacitors. I twisted the two wires together that bring power to the motor but I don't know if that made any difference as I have not tried it with non-twisted wires.

I'm sure the lengths of different wires and their positioning is very different from the situation (months ago) when I originally solved the interference problem by adding capacitors. There are too many potential variables to waste time trying to find out what was the critical change that has enabled it to work.

Thanks for the input.

...R

Ancient history: we used to add two wires to a plug-in crystal for our ham radios and twist the wires to lower the frequency just a bit. More twist, lower the frequency. Called a "gimmick" capacitor!

Glad you found a solution.

Paul

Paul_KD7HB:
Glad you found a solution.

I suspect that is a rather generous description. More like the solution found me :slight_smile:

...R

10nF ceramic is going to knock a whole load of RF, no need for 100nF for most of the RF spectrum
(in fact 10nF will do better for VHF and up). Its vital the cap is right on the motor terminals or even better
inside the motor directly across the brushes (not always possible).

FYI

Also, as mentioned later, use hot air to set the twists.

.

LarryD:
FYI

Thanks. I have bookmarked that Thread.

…R