Position measurement, encoder, repeatability error

[simasa]

Hello,
I am measuring a position using linear encoders by Heidenhain. The encoders output is 1Vpp which is then interpolated (200 times) to get TTL signal (10 KHz). National instrument counter board PCI 6602 is then used to calculate the position. Matlab is used to acquire and save the position data in real time.
A same path is measured couple of times. The measurement shows a repeatability error of 20µm.

This could be the real path deviation. But I guess the path deviation can not be as much as 20µm. thus I am thinking may be there is some thing wrong with the metrology. What could be the possibily wrong with such a metrology setup?
Thanks

 

[roydm]

What happens when you change direction, is it possible to miss a pulse? I think you need two out of phase signals.
Roy

 

[itsmoked]

TTL is not happy about 1Vpp. I'm amazed it's working at all unless I'm missing something.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[analogkid2digitalman]

What is the uninterpolated resolution of the encoder?

How are you traversing the path- motor driven, pushed by hand? If by hand are you using solid hard end stops?



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"It's the questions that drive us"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 

[simasa]

Thank for the replays.

@roydm
The measurement data shows, if a direction is changes in XY plane, there is always a 10-20µm jump in Z direction. Where as there was no movement along z axis.
actually I do not know if it is possible to miss a pulse.
Counter board calculate the position with two out of phase pulses and one reference pulse (typical with standard encoders).

@itsmoked
I use the standard Heidenhain compatible hardware to convert 1Vpp to TTL.

@analogkid2digitalman
Uninterpolated resution is 20µm. I am doing 200 times interpolation.

Traversing the path is done by a precession movement machine (very good positioning accuracy)

Regards
simars