HSC Encoder processing high frequency vibration

[plc_user]

After talking to support serveral times.  I am still very much concerned with the HSC Inputs on the MD series boards!  Everything checks out.  Wiring is fine, noise isn't an issue even though it is a very difficult to prove.  I don't understand what conditions are occurring other than vibration that would cause the HSC to count in one direction, than decrement the other on a random basis.  That's what is occurring.  The count never remains stable even when there isn't movement in the encoder.  Vibration is the only thing that I can see that is creating such an irritating situation.  What I feel is occurring is the frequency of vibration in the machine that is connected to the the Linear Cable Encoder is making the HSC not count the transistions correctly.  At low frequency it may be fine but is it possible that frequencies above a certain range might make the HSC inconsistent.  Is that a internal software issue!  Everything works fine until I connect the encoder to the machine.  For example, if I remove the encoder from the machine and pull the cable of the encoder slowly it works fine I attach it to the machine that has vibration it becomes inconsistent.  I have tried all possible angles to resolve the problem.  The only thing I am coming up with is the firmware in the PLC cannot handle the situation I am having.  I have used the SETSYSTEM 4 command to help alleviate the problem.  Still no success.  

Please if anyone has had a similar problem let me know what I doing wrong.

Thanks in advance

[plc_user]

UPDATE, Still no success.  It can't be electrical noise.  I switch to a linear cable encoder with differential outputs.  I used a converter to convert back to NPN as to remain compatible with the inputs on the MD1616.  The counter still contains an inconsistant count.  Sometimes it increments or decrements, randomly.  

Has anyone used an encoder with this board and had the same occurrance?  Let me know.  I really don't understand what is going on here.

Thanks

[AlbertMartinez]

What is the frequency of Encoder pulses coming into PLC HSC inputs?

Are you using a quadrature Encoder with pulse phase A and pulse phase B?

If you are using quadrature Encoder pulses, phase A must be wired to PLC Input #3 and phase B must be wired to PLC Input #4 that is if PLC HSC #1 is being used. If PLC HSC #2 is being used, then, wired phase A to PLC Input #5 and phase B to PLC Input #6.

If you are not using both quadrature Encoder pulses, for example if you are only using phase A let us say connected to PLC Input #3, then, Input #4 must be permanently disconnected (open) in order for leading-edge of pulse arriving at Input #3 to increment count. If Input #4 is permanently connected to ground,  the leading-edge of pulse arriving at Input #3 decrements count.

I am going to assume that you are using single phase (not quadrature) and the other HSC complement input is changing states and thus causing a reversal in direction of count .

Sincerely;

Albert Martinez
martinezalbert@yahoo.com

[plc_user]

Problem solved, I THINK.  After connecting inputs 3 & 4 to an oscilliscope, I feel I have found the problem.  I have my MD1616 an enclosure sharing space with several mechanical relays.  Two of which the PLC controls and two that are manipulated by the operator and/or a photoelectric sensor.    When the photoelectric sensor is activated(sensing an obstruction) it activates a 120VAC relay.  The relay inturn activates an air solenoid valve.  This photoelectric sensor, relay and solenoid are all 120VAC.  This circuit is in no way connected to the PLC.  The only thing in common is AC supply that DC Power supply uses.  When this sensor is obstructed and then not obstructed, I notice a quick swing in voltage on the inputs.  Going mostly have a logic high to almost a logic low.  Maybe enough to consider it a low.  This tells me that it is electrical noise induced by switching the AC parts of the circuits.

I removed the problem by placing varistors accross all of the solenoids and relays AC and DC.  I am still not sure why this is occurring though, the PLC is not even connected to that part of the circuit.  Please SUPPORT can you give a reason why that is?  

Thanks
PLC_USER

[support]

That is reason why it is almost mandatory that you must add a varistor across the contact of a relay that controls an AC solenoid coil, motor etc. Without a bypass mechanism such as varistor for AC or diode in the case of DC control, a collapsing magnetic field can generate more than 1000V across the relay contact, which can produce an electric arc that jump across the contact. This not only severely shortening the life of the relay contacts, but more importantly it presents a short but high energy electromagnetic field on its neigbouring circuit which is enough to be picked up by the wire between the encoder and the PLC.

Our PLCs normally use software debouncing techinque to filter out such short burst of noise. But when used as encoder input, the filtering is disabled in order for the encoder to operate at high speed (up to 10KHz).  That may be the reason why the noise seems to only affect the encoder input but does not affect other digital inputs.

[plc_user]

Thanks again for all your support.  I understand that all relays, need some sort of snubber(maybe with the exception solidstate relays).  I wouldn't have guessed that the relay's contacts needed it also.  Guess I should make a note of that in my user's manual.  Darn it, a simple thing like this causing so much trouble.  

Thanks again, you have a great product.