Hello Guest it is December 22, 2024, 01:19:32 AM

Author Topic: glass scales and servo's  (Read 5337 times)

0 Members and 1 Guest are viewing this topic.

Offline comet

*
  •  338 338
glass scales and servo's
« on: October 13, 2007, 12:02:58 PM »
hi,
  I would like to fit linear encoders onto my smalll ballscrewed servoed
miller. I have already got motor encoders and gecko 340's.
I was wondering could a piece of electronics be designed that would instantly sence
a movement pulse streem and switch from the rotary encoders to linear for positional accuracy?
then once the pulse stream stops switches (maybe 1/4 sec pause)back to the rotary motor mounted encoders
to stop them from hunting.Even though I only have .002" backlash I think the machine
will burn its servo's out sharpish without this type of system. I would think it might be possible to do this
in hardware rather than delving into Mach.
  Any thoughts?
   Tony
Re: glass scales and servo's
« Reply #1 on: October 13, 2007, 08:49:34 PM »
That piece of electronics will cost over $10,000+ for us hobby guys a encoder on the back of the motor does just fine:) 

Offline comet

*
  •  338 338
Re: glass scales and servo's
« Reply #2 on: October 14, 2007, 10:25:30 PM »
dennis,
 I was hoping for a more constructive reply,
 perhaps you could quantify yours as your obviously an expert on the subject.
   Tony

Offline Chip

*
  • *
  •  2,055 2,055
  • Gainesville Florida USA
Re: glass scales and servo's
« Reply #3 on: October 14, 2007, 11:24:31 PM »
Hi, Tony

If your ball screws have slop, Liner slides will cause your servo to hunt and seek like mad.

Trying to come up with an interface to combine liner slides and motor encoder would be expensive.

I have interfased liner slides on one mill, It had pretty tight ball screws and is somwwhat more nervious compared to motor encoder mills.

Spend your money re-building your ball screws or buy new ones.

Not saying it can't be done.

Just My Opinion, Chip

 
« Last Edit: October 14, 2007, 11:27:47 PM by afn09556 »
Re: glass scales and servo's
« Reply #4 on: October 15, 2007, 08:50:40 AM »
Nice Idea, like it. The reality is most of us hobbiests have machines with some sort of backlash and as linear scales are now becoming cheaper it would be nice to use them as they provide an accurate position measurement.

If one was to develop a circuit to do the task of switching between encoders, a CD4052BC chip could be implemented. It's a 4 channel analouge switch which has 2 select pins to switch between the 4 channels.

So that would be one way to take care of that task.

Question?
Would the switching action produce a voltage spike that would be read as an encoder count.

I was also thinking of the problem of the time taken between switching from one encoder to the other and the motor not having any control but then if there are no encoder pulse changes then the motor should just sit there.

Is this true?

How can Mach3 control this setup?

Would it be possible for instance, to use the 'Whilst Is Moving' command to implement a macro pump that choses the correct encoder?


I don't see it costing $10,000 but then would something like this work?


I am currently converting a mill to servo CNC using the standard method of motor and encoder but as the machine also has linear scales with DRO it would be a shame not to use them.

Offline DAlgie

*
  •  314 314
    • Algie Composite Aircraft
Re: glass scales and servo's
« Reply #5 on: October 16, 2007, 01:58:40 AM »
Don't think you would like both encoders and linear scales. The Accurite mill setup has these. While I feel the Accurite has some of the easiest to learn and use interfaces around, it can be a finicky machine as well. Have one that, unless you use a new mill, it will fault out from a small amount of chatter, and nobody seems to be able to fix it. Most of these are problem free, but it's something to do with the linear scales and encoders not being happy with each other. Like others here said, just stick with encoders, millions of machines around the world work well like that, without a problem.