Servo Motors..Gearbox or not??? 1

[Servo54]

Hello,
I have a JVL AC Servo Motor mounted directly to a Tolomatic Slide.This slide has a Kevlar belt.For the past 2 years we have been adjusting the Load Factor which affects the filter parameters every month or so.Recently I updated to the newer Firmware to get better control over the Filters.But I am not sold on this.
Every other servo in our plant is run using a gear box.Is it a general rule of thumb to have a gearbox on servo motors??I am also worried about the kevlar belt.I heard this could be stretchy??

 

[Skogsgurra]

It depends a lot of what your load is. A general rule is to try and get the inertia of load (as seen by motor) as equal to motor (and coupling) inertia as possible. If, for instance, you have a motor with 1 kgm2 (it is a lot, yes, but makes numbers easy) and a load with 16 kgm2, you will then select a 4:1 gear to make motor and load inertia about equal.

But that's not all there is to it. You also have to select top speed, tourque and available power to suit the application. And don't forget that the gear adds some inertia, too.

Gunnar Englund

http://www.gke.org

 

[Barry1961]

Can you give some more information about the application?

I would guess that this is either a very high speed application or that the kevlar belt is turning a screw to get the linear motion.

Barry1961

 

[Servo54]

The Tolomatic Slide is run by a kevlar belt which pulls through a pulley system.no ball screw.The speed on the JVL motor is 9000 RPM 3000 Vel RPM.Torque is @ 300%.The Load Factor is set @ 2.1.
This application requires accuracy and speed.I find the the motor oscillates the most at the end positions(Pick and Place)I have never had to touch any other closed loop controls on any other servo drive in the plant.This one is the only one that does not have a gearbox on it and the only one that has to be retuned very frequently.
I am purchasing a CGI 3:1 gearbox to try to smooth things out but still maintain my speed.I hope I am going in the right direction.
Thanks

 

[Barry1961]

I have not used the Tolo on high speed applications but have used belting with Kevlar tensile members. The Kevlar does not stretch like steel or Polymide but it will still have some spring.

The spans I was working with would be as long as 40 feet and the use of belt guides/supports was a must. With spans under 6 feet we never needed a belt support.

The teeth could also be deforming and producing a spring into system. This is common with the small round tooth “power grip” type belts.

I have never found a self tuning feature that would work well when there is a mechanical spring in the system. I found that when doing point to point that the best way to set the gains was to keep the integral at zero and increase proportional and derivative until there was a slight hum then back off some. The proportional gain would be roughly ten times the derivative.

I don’t think I would run a servo motor over 4500 rpm and most of the time I kept them under 3000 rpm. Of course I was only running 5 feet a second. There are probably many people who run motors at 6000 rpm who will disagree with me though.

The reflected inertia through a gear reduction is a square function, 2=.25, 3=.11, 4=.063, so if you are able to run now without the system becoming unstable a 2:1 reduction should be enough for inertia problems.

Barry1961