Position measurement problem using encoders

[simasa]

Hello,
I am trying to measure the position from 2 linear encoders (XY Plance).
My hardware setup is such as:

2 Encoders (Heidenhain) provide sine wave position signal (signal period 2 µm, frequency appx. 50KHz ). Signal is then digitized and interpolated (Sine wave to TTL, 200 times) using an interpolator box (Heidenhain). Position is then calculated using a counter board.

But I do not getting correct position signal. Encoders are used to trace a circular motion. But recorded motion is an uneven circle with incorrect scale.

What could be the possible reason?
I would be thankful for any suggestion.
Simasa

 

[Skogsgurra]

If the diameter of your circle is small, it is probably a mechanical problem with coupling between movement and encoders. Often overlooked, but very common.

If the diameter is relatively large, 100 mm or so, the mechanical problems usually are easily mastered. You have to turn to the manufaturer for advice. Those encoders are not known to have any problems.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[simasa]

The radius used is 150 mm. I think too that the encoders are functioning right.
I just thought about the possibility that the input frequency of 50 KHz might be high for the interpolator (200 times interpolation). May be I should try to reduce the interpolation factor (not sure).
could it be a possible reason?
I would be thankful for some further suggestion.
simasa

 

[Skogsgurra]

2 um and 50 kHz equals 100 000 um/second or 6 m/minute. What does the data sheet say about speed? There is usually a maximum allowed speed.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[simasa]

Maximum allowed frequency for 200 times interpolation (currently set) mensioned in data sheet is 12.5 kHz. The current input frequency which I have been trying id from 10-50 KHz. I do not have possibility to test with input signal below 10 KHz.

It seems violating the allowable input frequency range.
Would be the reason?

Some extra detail: Position output signal of X and Y motion should appear to be proper sine wave form (As motion along XY-Plane is a circular motion). I am getting a kind of sine wave with sharp edges.
I would be thankful for some further suggestion.
simasa

 

[Skogsgurra]

The "Doktor" (Heidenhain, that is) opens his office tomorrow. I think that you need to dicuss a faster system. As you say, you are violating the speed limit set for the system. It does work with less interpolation, I assume?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[sreid]

Take a look at the Heidenhain IK 220.

http://www.heidenhain.com/index.php...uktDB&op=catview&search=category&CS_UID=18506

The IBV660B limitation is the output edge speed.
12,500 x 200 x 4 = 10,000,000 edges/sec which is the maximum edge rate out of the IBV660B

The encoder signals are adjusted for amplitude, etc. via Pots in the APE boxes at the end of the encoder read head cable. Heidenhain sells calibration units to use to do this.