I think EMC having direct control over the motors is a much better 
setup. Step/dir drivers have numerous disadvantages as you have listed, 
plus they can lose position when you estop. With EMC controlling the 
motors, it won't complain if you move the motors while in estop. It will 
simply keep track of the moves. When you power up the next move will be 
from that position. Most step/dir drives will either lose position, 
fault or violently move to where they think the motor should be.

Electronic handwheels are great. I use them on my lathe (2 wheels) and 
mill (3 wheels) and would not be without them. It is pretty easy to add 
a scale switch so you can go from coarse movement to rapidly get from 
one point to another to very fine movement for accurate positioning. I 
found that 250CPR (1000 counts/rev) to 500CPR(2000 counts/rev) works 
well. All you need is a couple of bearings, an encoder and some sort of 
brake.  My handwheels use a couple of skateboard bearings to support the 
shaft. On one end is a cheap HEDS encoder and on the other end is the 
wheel. The shaft between the bearings is a larger diameter (about 1") 
and I use a spring loaded pad pressing on this part to give the friction.

Les




On 10/02/11 19:08, Kirk Wallace wrote:
>
> I agree. It just depends on what the budget is and the expected
> performance. I'm just exploring the options and issues.
>
> Personally, on a machine with existing servos, I would not use step/dir
> drives. But the original Mach doesn't support analog amps(?), so I
> understand why this machine has them. An option is to sell the step/dir
> drivers and use a hardware signal generator and PWM input amps, which is
> what I have on my HNC lathe. I think it would be harder to use the hand
> wheels this way (DRO mode). Two separate configs might be needed.
>
> I haven't tried it yet, but I would prefer to remove the hand wheels and
> make MPG's for each axis for "manual" hand wheel mode. Theoretically, it
> should be the same as real hand wheels, but with benefits. This requires
> the integrator to be fairly creative and might not be worth the effort.
>
> Oops, I just realized, in my last post, to use the hand wheels the
> drives will have to be disabled. This may make the drives lose there
> position or there might be issues when they are re-enabled (big
> difference between encoder position and step/dir input position).
> Hopefully someone with first hand experience will shed light on this.
>
> Thinking aloud -- with the position loop in EMC2, EMC2 would fight the
> change or would produce a following error. With MPG hand wheels, the
> hand wheel position, commanded position, and encoder position would
> normally be in sync in auto and manual modes. With a step/dir servo, if
> one tries to move the hand wheels the drive will try to correct the
> position back to the original position. The drive might alarm, or if
> EMC2 reads the encoders, EMC2 might alarm. Either way, in order to use
> the manual hand wheels manually, the loss of position will need to be
> considered.
>
> Another thought, I believe some resolution could be gained by going to a
> hardware signal generator/encoder counter. It depends on the current
> step/dir configuration. A step input might invoke a move of, let's say
> ten encoder counts. With a PWM (analogish) system the position could be
> controlled down to an encoder count or two, a least doubling resolution.
> Of course it would be better if the mechanical system were tight enough
> to take advantage of the increased resolution.
>


------------------------------------------------------------------------------
The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE:
Pinpoint memory and threading errors before they happen.
Find and fix more than 250 security defects in the development cycle.
Locate bottlenecks in serial and parallel code that limit performance.
http://p.sf.net/sfu/intel-dev2devfeb
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to