In article ,
Frank Erskine writes:
On Sun, 16 Nov 2008 08:30:06 +0000, Derek Geldard
wrote:

On Sat, 15 Nov 2008 23:59:04 +0000, Frank Erskine
wrote:


If you can get hold of Wireless World magazine for (ISTR) September
1967 there's an article on speed control for small dc motors.
The idea is that you apply variable pulse-width full voltage to the
motor, and between pulses measure the back-emf of the motor to measure
its speed. The b.e.m.f. is used to control the speed to pretty fine
limits, taking into account varying load on the motor.
The idea was based around a controller for model railways :-)

Ah, yes I remember it well.

The problem was trying to overcome stiction so that a model train
would pull away smoothly as opposed to sullenly sitting there, then
hareing off without warning and coming off the tracks at the first
bend which is what you got with a traditional (Hornby) rheostatic
controller.

I built one, and it worked very well, although at 'minimum' speed
there was still a tiny pulse from the controller, which made the train
creep very slowly at a 'station', especially if it didn't have many
wagons/coaches.

Everyday Electronics did one sometime in the 1970's.

Nowadays, you can buy a chip which does it all for you.
A couple of years ago, I designed a circuit to drive a
motorised potentiometer and used the A3953 which is a
full bridge PWM motor controller. (Only need full-bridge
if you want to have reverse control too.)

However, as I said in another post, PC fans are not
simple motors, and they aren't designed for PWM control
and certainly not reversing. (I did permanently reverse
one once, by carefully disassembling and swapping around
the motor field winding connections.)

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]