How to connecta an incremental encoder

[goargamo]

Hi everybody, i was wondering if someone could help me with a doubt i have concerning to the conections of an incremental encoder. I have this encoder with 6 different cables:

1) Vcc

2) GND

3) channel A

4) channel B

5) channel Z

6) "shield" fg

I connected Vcc to 24volts dc, gnd to ground, and channels A and B to the high speed inputs HSI.0 and HSI.1 respectively... MY DOUBT IS ¿¿¿where shoud i connect "Z" and "shield", or should i leave them unconnected???

Thanks a lot for your answers....

[Joe Tauser]

There's no place to connect the Z channel on a Unitronics PLC. Hopefully you can get by without it.

Connect the shield to a ground terminal.

Joe T.

[goargamo]

Thanks for the info Joe... I connected the shield (FG) to ground but I´m still having kind of a problem.

Im using an incremental encoder of 2,000 pulses, and in my program i preseted that every 8,000 pulses i should count a turn (8,000 because i'm usign a "high speed shaft encoder x4"). The thing is that i'm not counting the number of turns i should...

¿do you know why this could be happening?, my PLC is a M90.2.T1

Thanks again

[Joe Tauser]

Look at the register where your pulses are going and see if you are getting 8,000 counts for exactly one turn of the encoder. Also remember that a signed 16-bit MI register can only hold 32,767 counts - are you resetting the counts after you increment your turn counter?

What is the frequency of the pulses coming into the PLC?

Joe T.

[goargamo]

Hi Joe. I will check that. Yes, i'm resetting my count every 8,000 pulses (supposedly). The frecuency coming into the PLC is about 300rpm, (5 revolutions per second x 2,000 pulses per revolution = 10,000 pulses/second)... is this frequency to high =s???

Thanks again

[Joe Tauser]

If you've got the X4 mode set you're actually getting 40,000 counts/second, so you're reaching your 8,000 counts in 0.2 seconds. This shouldn't be a problem, but I am curious why you have such a high resolution encoder on this. If all you're doing is counting turns, a 10 pulse/rev encoder would work fine. Are you doing some sort of measurement as well?

Joe T.

[goargamo]

I´m using this encoder because it was the only one i was able to get (the supplier of the company only had this one in existence) =s. Yes the only thing i'm doing is counting turns i don´t really know why this is happening.

To try to correct the error in the count of turns i have modified the 8,000 pulses corresponding to a turn to a different number (6,810) and i'm getting a smaller error but i thing this shouldn't be this way, i'll show you the ladder i'm doing so hopefully you find an error jajajaja, thanks a lot Joe

[Simon]

Getting back to the question of counter speed, you say it is 10,000 pulses per second, which is equal to the maximum speed of the M91-2-T1 ( note the max input speed usually relates to the raw pulse frequency not the decoded quadrature count frequency)

It would be helpful to check that you get 8000 counts per rev just by a simple test, without turns counting logic. Then we can isolate whether the problem is the encoder interface or the turns counting logic.

You can test this by turning the encoder more slowly and see if the count is more accurate.

You may not get to the maximum speed possible if the pulses from the encoder are not a clean squarewave (for example they can become slightly triangular (trapezoidal) or suffer other distortions).

I hope this helps

[goargamo]

Thanks I will do that test with a low frecuency to see if i´m really getting 8.000 pulses per revolution. Thanks a lot