http://web.mit.edu/cmse/educational/Beauvais_DC_Motor_LP.pdf @archive.orgThink how to improve it - e.g.
* Add two rightangle plastic flaps cut from a blister-pak, not quite touching the wire ends to stop the coil lead 'shafts' from wandering sideways in the 'bearings'.
* Does lubricating the 'bearings' with graphite from a carpenter's pencil make it go faster?
* Does increasing the contact pressure with a nylon bristle pushing down on the shaft on the part stripped side, next to the bearing on the inside, make it run more smoothly?
* How would you measure its speed?
N.B. the battery is being virtually short-circuited by the coil every time it makes contact, so it has a really hard life. They are using a 9V Duracell PP3 because its reasonably safe if short-circuited - it will get a bit hot but is unlikely to cause burns or explode. However dollar store Zinc Carbon C cells would give you far more runtime and also let you vary the voltage - get a 4 call battery holder and solder a wire to a flat strip of copper or brass so you can insert it between a battery base and the contact to tap off various voltages.
Then there's all sorts of traditional stuff you can do with torch bulbs, home made switches and buzzers etc.
Or if you are looking for something advanced, grab a long roll of CAT5, a dozen cheap novelty magnetic compasses, a reel of thin magnet wire, soft iron wire (florist's wire) to build the narrow U shaped cores for the coils, to get a predictable deflection angle, and to create a magnetic shunt so the compass returns to center, two dozen microswitches and two battery holders and build a two station, six needle
Wheatstone telegraph. With sprung return pivoting pointers to operate the microswitches (wooden, pivoting on a screw, with a small washer as a spacer, with a pin through a slot in the base board to operate the microswitches which provide the spring return), it could be very simple to use.
i]Edit: Killjoys at M.I.T. have taken the Beauvais DC Motor PDF down
but its fortunately still on Wayback![/i]