OT - smooth motor rotation

[elevmike]

DickDV,

Are you famillar with gearless elevator motors? Imperial makes them down to 4 HP. Couple one of these baby's to his "large" cylinder and run it at 10 rpm all day long.... (I wish I knew more about this app...it's starting to bug me...)

 

[katratzi]

wow, thanks for all ideas/websites. I'll be doing some
research over the next couple days. I'm not sure what a film-
casting application is rdrast, but I've been cleared to say
that a substrate will be traveling, its speed determined by
this cylinder. a material will be applied to this substrate,
and the major factor that determines the amount of applied
material is substrate speed.
The Danaher site seems a very good place to start, and
Mike, I hadn't thought of the flywheel, but that could really
help in some situations.
Dick, by surface driven were you meaning using a motor
similar to the pancake motors that were mentioned, and have
the cylinder, in essence, be the shaft? no gearbox backlash,
no coupling misalignments or shifting, this sounds good!

DEmotivation? thats one funny site, allan. I'd love to post
a few of those at work. Anyway thanks for the replies, guys,
I've got a lot to think about. But for now, its almost time
for Battlestar Galactica!

 

[elevmike]

How big is the cylinder dia? 10" or 60"? That would make a difference too.

 

[kamenges]

You essentially have two issues. The first is your actuator (motor) needs to be able to deliver the smooth motion you desire. This tends to lead the discussion toward high pole count motors. The only real reason for this is that small errors in sine wave reproduction at the power section of the drive will have a lesser effect on shaft motion at the motor. I believe the motors that jdkasper referred to below fall in this class. Some permanent magnet servos can have a problem with torque ripple, which may cause you a problem. Make sure you bring up torque ripple and how this in mitigated with whoever you talk to.

The second issue is feedback resolution. If you are looking to go 'extremely smooth' 1024PPR probably won't cut it. Someone previously mentioned Indramat EcoDrives. One of the reasons they produce such smooth motion is the feedback resolution. They produce in excess of 2 million counts per revolution. These are sin/cos encoders with a high speed serial link from the motor feedback to he drive. A 'standard' sin/cos package will work well also as long as you babysit the leads. The feedback signal is usually 1Vp-p so noise can creep in pretty easily.

Indramat (Bosch/Rexroth) has a line of high pole count pancake motors that operate very smoothly. You may want to look that way. But however you look, high resolution feedback is a must. Getting the highest resolution feedback your drive can handle won't hurt you and will most likely help.

Keith

 

[rsdoran]

rdrast said film but it could easily be a printing process, a large drum would imply something similar to a flexographic press which uses multiple plates for printing.

The drum aspect implies that it is not a cutting or stamping process because a drum (in most situations) would need to be perfectly smooth or the imperfections would be transfered into the process. It could be an engraving process too if the substrate has a thicknees or backing that would prevent imprinting on the drum.

That said then the speed should not be the concerning factor, registration (alignment) would be the key). The process with things like a flexo press, which uses a very large drum (4 to 6 foot diameter) will appply the process (the plates are moved against the drum) and registration (alignment) is then checked.

In this type application the drum speed is usually matched to "line speed" ...ie if it is paper the line speed could be 1500 feet per minute therefore the drum rotational speed has to match what is needed for 1500 feet per minute to travel across the used surface. Depending on the application this usually requires multiple encoders and/or tachometers because the drum alone does not "transport" the paper.

I am no expert in any field but I do not see enough information to offer an answer. The application may be able to use stepper, servo, DC or AC but if high torque is required at low speeds then I would not eliminate the use of a gear drive system. There are numerous gearboxes available that have backlash measured in 1000ths (.001) of an inch.

 

[katratzi]

from what I've found out over the last couple of days, keith,
you are right on target. were going to be looking at motors with
high rez feedback, high pole counts , skewed slots, anything
that helps to minimize velocity ripple. you seem to have worked
with these types of motors before.
For those still following this thread, heres one more analogy
that may explain this a little better. Imagine you and a friend
(to make it even better, imagine a good looking friend of the
opposite gender), have a flat substrate stretched between you
and you are walking toward a device that can apply a material
with an absolute constant rate onto this flat substrate. Now,
if the application rate is constant, the only thing that can affect how much material is applied is the substrate speed. The slower the speed, the thicker the application layer, and of course if you and your friend just stop moving, the material
being applied will just build up there. So if you can just keep
a perfect constant speed, you will have applied same amount of
material the length of your product, which is the idea.
I hope this explanation gives those who may still be wondering a clearer picture of why such uniform speed is
needed.

- charles

 

[kamenges]

Originally posted by katratzi:

you seem to have worked with these types of motors before.

The company I work for manufactures custom web handling equipment. I've run across things like this before.
One machine we did a couple of years ago had to run a web as slow as 1"/minute around a 36" diameter drum. You really had to watch closely to see the material move.

Keep up the research and don't be afraid to ask questions. And make sure the answers make sense to you. If it sounds too good to be true it probably is. This stuff is do-able. You just need to pay attention to details.

Keith