[Rankine Cycle]

Hello all. New member here. I am trying to find out if any of the retrofit controllers (Centroid, Ajax (think there the same actually)) can use the output from typical DRO glass scales in lieu of the encoder on the back of the servo??

Reason, I ask is that I would think this opens up the possibility for purchasing a more 'loose' tired machine for the build, since the ballscrew backlash becomes a non-issue for precision. Yes, I do understand backlash has its own issues with regards to machining, like difficulty with climb milling, and perhaps when machining circles etc...

Im considering a rather tired looking kitamura mycenter 4. While I havent gone to look at the machine yet and tried to determine the actual backlash, given the overall condition, I suspect there may be some....

Kevin

[louieatienza]

Check in the linear and rotary forum, there is one guy in Greece I believe that did exactly that.

[LaserTechTim]

Be careful, more than likely you can use scales for feedback ( i don't know those contrlers) but when you have a loose machine you will have to set up the control to be loose as well; the control will look at output pulses and compare to input (feedback) if it is too tight you will have" following error" and the control will fault, requiring a larger feedback "clearance".

[minimum]

Kflop will be able to do that but you have to be aware that it's not perfect: the feedback allows controller to correct positional errors but cutting forces still will "pull" the table off it's correct position; controller compensates (with a small delay of course) but it's already "too late".

[MickGiever]

Before you go to all that expense and ag, check the play with an indicator and a pry-bar. Then do your self a great service by replacing the radial and thrust bearings that support the lead screws. I have restored the precision in many loose machines by doing only that. Think about it, how can you adjust your control for back-lash it the ball screw itself is flopping around. Those bearings usually wear out long before the ball screws do. You may not get factory-new precision, but you can turn back the clock about 85% or so. Now, wouldn't that be money better spent? You don't have to remove the ball-screw from the machine, only the end-caps and drive motors. Inspect the motor coupling for any play, as they can wear too. And while you're at it, spring another $5 for a tube of marine grade waterproof grease to repack those new bearings. Coolant and condensation migration is the major culprit causing these bearing to fail. Whenever possible, I also use a layer of grease soaked felt to block contamination and add further protection to these key components. By the way, Fadal only allows .003" mechanical backlash total for controller compensation, just to give you a goal. Any more, and the end-mill cutting edges will suffer more from ultrasonic frequency degradation and chatter, than from the actual mechanical wear.
Next, adjust the gibs for backlash with an indicator and pry-bar, just allow enough room for the lubrication to flow.
This ought to be some food for thought for a lot of you other guys also having backlash issues!
Best of luck!
Mick

[Rankine Cycle]

Excellent information here, all of it. Many thanks for the input!!!

[Al_The_Man]

Here is an instructional video that show the limitation and corrections needed when using a dual loop, (scale & motor encoder).
Tutorials - Galil: We Move The World
Al.

[louieatienza]

Here's the thread I remembered; I'd suggest checking out his other threads too. I do find it odd that there weren't many views on this thread, probably because it was in the linear motion forum. the username is rokag3

http://www.cnczone.com/forums/linear...un_ftv2-2.html

[Rankine Cycle]

Yep, that all makes good sense. I was figuring any backlash could indeed introduce hunting...

Well its all a moot point now, I just bought a solid, tight-as-can be Hermle UWF 1000 !!!! Has an ancient looking Heidenhain TNC 355 control.... dont even know if it will boot up, but I'll try anyway.