[carbidecraters]

Is it possible to use Mitutoyo's linear tape for feedback on positional accuracy instead of motor feedback? This may take care of any positional errors that may occur or add up. Anyone heard of this?

Here is their patent #

http://www.freepatentsonline.com/5894678.html

[LMSC]

Unless your accuracy is =/- 1/8th of a inch I don't think that this will work. I looked on the US office patent website and the description starts as follows:

What is claimed is:
1. An electronic tape measure, comprising:

There are inexpensive scales sold by myself and others companies that will work for what you need.
Regards,
Tim Barnard
LMSC
877-376-4373

[carbidecraters]

The patents may not be the correct idea I was under the impression that the linear tape was new to this year and is to be shown at IMTS in Chicago. I hear it will revolutionize the scale industry and will replace most of the glass scales.

[Torchhead]

Taking positional feedback from the table presents several challenges.

1. Resolution is low compared to geared down motor shaft position with moderate to high encoder count. On a servo that has a 500 line encoder (2000 PPR) your raw resolution is .18 deg of rotation. Depending on either the leadscrew pitch or gear/belt reduction between the motor and load that can translate easily to a resolution of better than .00005 inch. Why does it matter? Because the servo loop uses the finer resolution to obtain smoother operation and accuracy.

2. ANY backlash in the system will cause instability and usually lead to oscillation using the conventional drive based loop control. The machine will shake and rattle. While entertaining it also renders cutting impossible. What happens is the loop circuit reads the position of the linear encoder and tells the motor to move to correct it. If backlash is present the motor will take out the backlash and then move the load (usually too far) causing overshoot. The system sees it and moves the load back the other way with even more overshoot, etc, etc. Most conventional servo loops do not just move to an encoder line then stop. They actually have a tiny oscillation between two lines (delicate balance) so the back and forth happens several times per second even when at rest. Thus the "hum" of your servo and it's being locked on power up. Link that up to a "loose" loop with lots of inertia and you have to dramatically dampen the loop (lower the response time) and that will also lower it during actual moves.

3. The concept of encoder on tape is not new. You can buy encoder tape from USDigital. Perhaps reading the tape from a single surface accurately my be the breakthrough.

If you go with a more sophisticated control with dual loop circuits you can use motor encoders for the dynamic torque feedback and the tape for true positional feedback. Have a look at the CNC "Brain" advertised on this site.

TOM CAUDLE
www.CandCNC.com

[carbidecraters]

Thanks for chiming in Tom. I dont know the specifics but the big mfgs are chatting about this as a replacement for current open and closed loop system. I wish I could make IMTS this year but I will be getting married.