Rack-shelf metal is great for this type of experiment. You can bolt it
together easily, make lots of adjustments, and with some cross-beams, a very
strong structure can be build. The even spaced holes let you easily make
accurate symmetrical constructions.
I used 30x50mm, 2mm thickness material. It was quite cheap in Taiwan, about
US$ 1 per meter. You can get it up to 2.5m long. In Europe, check for "dexion
slotted angle"
For the cables I used 2.5mm steel cable, that easily holds the total platform
weight of about 120kg (including me). Cable tensioners and fasteners are
standard hardware stuff.
The chain-sprocket wheel drive generates quite some noise. A
much more silent drive can be obtained by using belt-pulley. I used "A" type
belt, 2" pulley on the motor that drives a 6" pulley mounted on the drive
axis via #22 belt. Then a 2" pulley on the same drive axis moves a 2nd belt
that is connected to the cable. The mechanical drawings of the pulleys
are shown on the side. I had these made in Taiwan, hence the Chinese
comments.
The motors are real industry DC servo motors. I was lucky enough to find
them on the junkyard one day. They are specified as:
Typical power 230W, current 4.4A, 2100rpm@ 52V, torque 1.05Nm. I used this
motor with +/- 24V drive voltage. When I increased the voltage to +/- 36V,
the platform drive forces got somewhat scary. These motors could rip the
whole structure apart.
You really can get away with much less, depending on the gearing you want to
implement. Some alternatives could be electric gate openers, electric
screwdrivers, car wiper motors.
The roll-bearing wheels came from old production line conveyer belt rollers.
They are aluminum with V-groove, and inside have double ball bearings with
15mm inner shaft.
For the end-stop switches I used micro-switches, with extended lever, that
is pulled down by a rope and a spring.