Zero mark in encoder

Most of the incremental encoders have a zero track that gives a pulse every rotation. In most controllers if you configure the encoder and you tell the system it has a zero track, the controller will check f.i. for a 1024 pulse encoder that 1024 pulses after the last zero track there is a new zero track. I believe if there would be one pulse missing the controller issues an alarm.
I have a machine with an encoder for length measurement, and I'm not sure if it is counting ok. I know that in the controller the zero track is not activated. I hope that by activating the zero track that I will be confirmed if there are miscounts by the encoder. Especially about EMC disturbance I am not sure.
Is this true that if the zero track is activated and I get no encoder errors I may conclude the counting is ok? What are your experiences with the 'zero track'?
In specific my system is Siemens S120, encoders connected to SMC 30, but it could be any drive system.

Is the encoder part of a motor or is it mounted separately ?

L D[AR2 said:

Is the encoder part of a motor or is it mounted separately ?

Click to expand...

It is on a SMC30 encodermodule that is connected to Simotion D410. Then encoder itself is not used for the drive, it's a length measurement for the Simotion D410 CPU. But in the configuration of the SMC30 you can select that there is a zero track.

Plc_User said:

I believe if there would be one pulse missing the controller issues an alarm.

Click to expand...

I am nor familiar with Siemens S120 and Simotion, but have never seen such diagnostic feature.

How are you checking the encoder counts against length to be not sure of the count?

L D[AR2 said:

How are you checking the encoder counts against length to be not sure of the count?

Click to expand...

It concerns a flying shear application and when I trace the cut lengths in the controller the cutlengths seem correct, when we measure the cut pieces, they have variable tolerances of several milimeter. I have checked already so many things and from the traces the machine seems to cut at the right postions, so
incorrect counting is one of possibles sources to check.

How many counts per millimetre ?

I have used encoders with marker pulses (MasterDrive and S120) but I have never seen a fault reported due to incorrect pulses in 1 rev - it may be detected, I've just never seen it reported.

Enable the marker pulse and then change the encoder configuration to be double the actual pulse count. The controller will receive half of the expected pulses in one rev - you will soon find out if there is any missing pulse fault reporting.

If this is the same application as before you shouldn't be missing counts, the frequency isn't that high. You haven't said whether the lengths are longer or shorter than they should be. If the cuts are long that is because the master encoder is slipping. It is a little of both then then control is simply not that good.

The encoder decoder usually counts 4 phases per line and the phases or states must be in sequence. If the phases or states occur out of sequence and error is generated immediately. There should be no need to hook up the Z input but it doesn't hurt.

Check the power supply for drop outs and the encoder lines for noise but noise will usually give you extra counts or a fault.
Usually the problem is with the mechanics not the electronics.

Do you have the ability to graph the motion profile at the millisecond level or faster? Then one can see the counts change and look for abnormal encoder changes.

Haw to resolve the alarm no pluse from encoder