hmm a black cylindrical thing with a red band and a stepped end.
looks a little like a land mine but probably isnt.
My current setup can't handle something that short. The current roller is
rated at 120V 0.15 amps and can easily handle a larger load then the drum in
that picture. That's what 18 watts. The difference is the bigger setup turns
two rubber wheels that the big drum sits on. The gap between the wheels is
too big for a small drum. I'm likely going to try and drive the short drum
using the magnet fitted to it's base.
Just spinning it then reversing direction.
it appears to be reasonably well balanced so there's not going to a lot of
torque needed.
I'd bet even a 10% variation wouldn't be a deal breaker. It might actually
help by introducing some random action into the process. OTOH I'd like
something that wasn't too random -)
most DC motors will meet that 10% figure quite easily if not loaded heavily.
Worse case is 30 minutes of steady turning. It won't be used much just
when I need the smaller drum.
I'll assume that's typical... is that slowest or fastest?
my first idea is to make some sort of adaptor to let you use your existing
machine (or add another roller to it etc...)
Otherwise it seems to be a matter of taking a small DC motor and gearing it
down from the 10000+RPM they typically run at to about 0.04 RPM for your
worst case.
I'd suggest getting the largest motor you can power easily, the bigger the
motor the less its speed will fluctuate if there's a rough spot on one of
the gears etc.
small gear sets boxes are available from hobby model making places.
I've seen some that can be "stacked" to give any degree of speed reduction.
I expect the motor will run smoothly enough so as to not need any regulation
of its speed.
For permanent magnet motors with brushes the speed can be reduced quite
effectively by reducing the voltage to the motor.
the next problem is the 520 degree turn and then the reversing at the end.
and coming back (and then stopping?)
It seems to me that a disk with two features that activate a sensor and a
system to count the passing of these features would be the way to go.
start at one of the features, count three new features on the way out,
and three new features on the way back.
a setup like that can easily get 5 degree accuracy, and with a bit of care
could exceed 1 degree accuracy.
a small circuit board with a few dollars worth of parts.
another idea is that a small three phase AC motor could be used they give
an lmost perfect constant speed (but only one speed) and are about as easy
to reverse as DC motors. (they ca be driven off a single phae AC supply
with a simple circuit)
If you need an "industrial strength" solution you could spend a little
more money and do it "properly" with a PLC (programmable logic controller
- basically a small computer thingy designed for process controll I expect
that even the smallest (cheapest) PLC would be capable of tasks many times
more comple than this one) one of those motor controllers you saw earlier
and sensors from the same (or a similar) supplier... this has the advantage
that you don't need to solder it together (and once setup it'll probably be
more reliable) and the disadvantage that you'll need to program the PLC to
get the behavior you want. (this is done by hooking it up to a PC and
running the manufacurers software and entering the program aftwer which
the device is disconected and runs autonomously)
Disclaimer: I don't know a whole lot about this sort of stuff. so research
everything for yourself, consult real experts etc...
Bye.
Jasen
