Re: [CAD_CAM_EDM_DRO] Timing Belt speed reducers

Dan Mauch wrote:

> One of my customers wanted to use one of my MCG ID23005 nema 23 servo motor on an axis that was not suited for a direct drive. I calculated that to use 80% of the output power of the motor that he would need a 5.4 to 1 reduction. This reduction would require use of a jack shaft between the motor and the output shaft.
> He did not have the equipment to make a speed reducer so I offered to build it for him at a nominal price.
> I designed a base plate and a motor plate that would fit his nema 23 mounting flange. Using a .1875 end mill I bored the base plate and the motor plate to fit .250 ID ball bearings. I also bored two other .625" OD holes for two more bearing to support the jack shaft.
> In the design process I used the center distance calculator at stock drives Inc. When I received the parts and install the .080 pitch pulleys on their shafts and into the base and motor plate I found it quite difficult to insert the main drive shaft bearing into the base plate with the timing belt over the jack shaft pulley and the output shaft pulley but after several attempts finally got it assembled.
> The question I have is the center distance calculator designed to give the belt a really tight fit over the pulleys? It sure seems like the belt is overly taunt but I double checked my drawing against their calculator dimensions and it right on. The base plate and motor plates are also right on. Other wise the speed reducer works as it should.

Sure, assuming the belt and sprockets will eventually wear, the
belt is going to start out quite tight. You really need to
leave some slack somewhere in the mechanism so you can tension
the belt AFTER the components are assembled. The only way I can
imagine that in this particular job is to have the whole
assembly slotted so the jackshaft will pull the final belt into
tension, then slide the motor to tension the primary belt. It
is not a real optimum scheme, and I think this all goes back to
Mariss Freimanis recommending a particular ratio to optimize
what may have been a non-optimal motor selection. You don't
need a perfect match between motor and leadscrew to get optimal
performance, you need to have sufficient torque available at all
speeds to avoid stalls. Sometimes a bigger motor directly
coupled can be better than a smaller motor and complex belt
drives. The jackshaft is the result of trying to use the wrong
motor, and the customer's refusing to admit that he made a bad
selection there. I have NEVER seen such a scheme in a
properly-designed motion control system. Adding needless
complexity to a system is just going to be trouble in the end.

(I'm not faulting you, Dan, you gave the customer what he
requested, and I'm sure it will work to some extent. Putting
too much reduction on steppers is a disaster due to the
torque/speed curve, but it often is not so great with servos,
either. It greatly magnifies the rotational inertia of the
motor, and double belts might vibrate more than a single belt,
leading to some servo stability problems.)

On the other hand, I am using some small servo motors (both
brush and brushless) on a minimill with 4:1 reduction, and it
works pretty well.

Jon