Usually a frequency above the range of hearing is used, 24 to 35KHz bein
g
typical. However I used a frequency of 1KHz with a 6" diameter PMI pancake
motor with good results; the sound level was fairly low (ie it did not make
a good speaker).

        This allowed me to sample the back EMF from the motor through a 4066
analog switch feeding a .22uF cap followed by an opamp voltage follower.
The dc permanent magnet motor acts like a generator when no power is
applied and the sampled voltage is a good representation of the motor
speed. This allowed me to control the motor speed to about 1% despite load
changes (robot wheel drive, steering through differential motor speed).

        The pancake motors (flat printed circuit board armature) have low
inductance and the "kickback" is minimal at the end of the power pulse but
you should wait as long as possible before sampling the "generator"
voltage. The ideal time to sample is just before applying the next power
pulse. I used a 1uF mylar cap and a tranzorb across the motor leads (at the
motor) to reduce RFI and to absorb the kickback to protect the 4066.
However this may not work too well with other motors which have higher
armature inductance.

*****

At 11:09 AM 5/23/99 -0400, you wrote:
>   Is there any rule of thumb for the frequency one uses
>to run a (small in my case) DC motor at various speeds?
>It seems obvious that the duty cycle determines the
>speed.  What's not so obvious is what range of frequency
>& why.
>
>  Thanks
>
>_________________
>
>Bob Russo
>Guilford, CT
>brusso@iconn.net
>_________________
>
>