Wire Drawn Linear Quadrature Encoder

plc_user1973

Member
Join Date
Aug 2006
Location
Pennslyvania
Posts
21
Greetings, this my first post to this forum so here it goes. I have a PLC from Triangle Research (T100MD888.) I have a wire drawn linear quadrature encoder connected to the PLC. The encoder has differential outputs and is place 50+ feet from the PLC. The PLC doesn't have differential inputs but is connected through a differential to open collecter unit.

The problem is: As with most people that use quad encoders my position data is not consistent with what is measured with a "tape" measure. At the end of a work day. The position data is not consistent to what it should be.

I am using this equipment to control the position of a sawmill carriage. The carriage holds the log at a position in relation to the sawblade. This is making me look very bad. I have exhausted what I feel is all means to correct the problem and the only thing I can think is the problem is the wire drawn encoder.

I believe the encoder itself is fine, but the spool and wire apparatus is what is in question.

The sawmill carriage moves at 6 in/sec and the encoder has a resolution of 500 pulses/in. Therefore 3kHz is being generated at the outputs of the encoder. This pulse rate is well within the 10 kHz response of the PLC.

I don't believe its a noise problem due to the differential out's and I also have tested the machine at rest and there is not problems with position data. To further see if it was noise, I connected the encoder outs to an interrupt channel with the machinery running but at rest and there was no accumation of interrupts being generated.

I feel that the acceleration of a "crashing" log against the fense of the machine which is monitored by the encoder is greater than the accerlation rate of the retraction spring of the encoder and it causes the wire to wrap around the spool in an inexact means, which could lead to positioning errors.

Can anyone give me some insite on this.

Thanks in advance.

Roger
 
Potential problems with the system as you describe:
1) Slippage between the wire and the encoder wheel.
2) Slack forming between wire and encoder wheel during abrupt motion.
3) Stretching of the wire.
4) Inertia of the encoder, keeping it in motion even after the driving element is stopped.

1, 2, and 4 are all parts of the same thing, and result in cumulative errors. 3 is generally a long term, slow drift, but may be affected by temperature as well.

Solutions? None are perfect, but...here are a few ideas.
1) Have a 'Calibrate' switch, somwhere that is guaranteed to be hit every pass. Use this to zero the system to a known home position.
2) Use a 'Capstan' method of attaching wire to encoder - That is, make one (or more) wraps of the wire around the encoder sheave.
3) Use a chain and pulley, instead of a wire and sheave.
4) Use an alternate sensor, such as an absolute magneto-restrictive type, instead of an encoder. These can be purchased in fairly long lengths (Turck, MTS, etc).
6) Use a different cable. I've used similar systems, and all have failed miserably when using a lubricated aircraft cable. Switching to a non-lubricated cable fixed all slippage problems. (along with a multi-wrap capstan).

Not a complete list, but at least a few ideas.
 
Thanks for everyones, replies. I can only hope to hear some more. This problem is making me look very incompentent. I have been through more engineers and techies at the relavent companies. I would like to keep the PLC the same because lets face it, it is easy to program. I would entertain a SSI encoder. Would a timing belt and appropriate cog be better in this type of monitoring application connected to the SSI?

Please if you guys/gals are willing keep the advice coming. I am sorry though, a "Thank you" can be all that I give as payment. :)

Later
Roger
 
plc_user1973 said:
T Would a timing belt and appropriate cog be better in this type of monitoring application connected to the SSI?
Yes, but it seems like we are telling you what you already know. The problem is getting that data back to the PLC.

I am a big fan of Stegmann SSI multi-turn encoders. I have many saw mill customers that use SSI encoders for head rig and end dogger carriages. Other SSI encoders brands may work but I avoid brands that extend the SSI standard in non standard ways.

So how long is the travel?
 

Similar Topics

Hi there, Trying to get some ascii serial communications working via RS485 (COMMREQ functions). I have attached our wiring for the COM2...
Replies
1
Views
948
What's everyone use for I/O drawings on new installs. I am trying AutoCad Electrical, but can't download it due to a MS Office application running...
Replies
11
Views
1,266
I've been tasked with adding a 3-wire pressure differential transmitter to a system and drawing up the wiring to an existing PLC board. The...
Replies
7
Views
740
Hi all, I wonder what's the best type of cable to use for RS-485 2 wire + GND That is, Data+, Data- & GND I know that Belden 9841 (not...
Replies
10
Views
1,014
This is something I figured out that electrically doesn't make sense to me. An electric blanket with a single heating wire looped through it and...
Replies
12
Views
983
Back
Top Bottom