hello,
I use a autonics rotary encoder ( E40S6-3600-6-L-5 ) in my project. the maximum speed of the rotary is about 900 rpm.
but at speeds above 100 rpm, I have a lot of noise in the rotary signal, especially on the Z signal. I use a Schmitt trigger (74LS14) to reduce noise.
it is better now but still, noise on the Z signal yet.
is there any effective circuit or any methods for removing noise from the encoder signal?
thanks.

 

How long is the cable to the encoder?

Does the encoder have a differential output?

 

How long is the cable to the encoder?

Does the encoder have a differential output?

cable is about 3 meters.
yes, +A,+B,+Z and -A,-B,-Z (Line Driver Output)

 

Use differential line receivers.

3 meters using single-ended signaling, especially in a noisy environment is going to give trouble.

 

Can I assume that:
- all the encoder signals are ran together inside the same cable?
- the cable is shielded?
- the shield of the cable is bonded to ground?
- the cable is routed away from large EMI sources?

If all those are YES then I suspect you have a bad connection or bad conductor in your cable which corresponds to your noisy Z signal. You should have the same amount of noise on all your signals, if noise is the real problem.

Below is a recent finding of mine. The encoder had a 30cm whip which was spliced onto shielded CAT6 ethernet cable to extend it about 20m, routed alongside high power cables. Noise was not present on the other signals so I did not investigate noise as the cause of the problem. I found a loose connection at the splice, and after the splice was properly reconnected there was no more noise on the signal.

 

Per @Sensacell , I second using one of the common differential driver I.C.'s, I prefer to use the encoders with this built in, but where faced with no choice, I use an external type.
Also on long runs, I use Belden 9891 cable, very pricey, but avoids headaches later on!

PS:the spec sheet shows differential output, are you using it in this fashion, or just single ended?

 

I think the 1 uF electrolytic capacitors are too large to be used for noise suppression as they will degrade fast edges. Also they will not be as effective as film or ceramic capacitors at filtering high frequency noise.

Les.

 

Shouldn't need any capacitors if the right internal line drivers are used, also is the OP using the full differential capability or using them in single ended mode?

 

Post #1 schematic seems to suggest that he is using singled output and I was assuming it was just the A and B channels.
But in post #3 he says he is using differential mode. If that was the case he would use a differential receiver chip. (Not a Schmitt inverter.) I think we need clarification.

Les.

 

Agree.
Beginning to sound like a totally wrong application!.