Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Encoder Noise Guidelines

Status
Not open for further replies.

kiltach

Mechanical
May 21, 2009
3
US
Hi,

i'm a pretty much a mechanical engineer but I'm putting together a control package for a precision mechanical stage.

The short of it is i'm trying to figure out good guidelines to acceptable wiring lengths in encoder lines.

The long of it.
I'm using linear encoders that output 1VPP waveform.
That will go into a quadrature interpolater about 1m away.
5-12+ meters away will be my control cabinet. (still figuring out how far away it will be). I will run the encoder lines into the control cabinet in twisted pair shielded cable with differential signals. The encoder lines will run onto a breakout board which will split my line into my controller (inside the cabinet) and to a U.S. Digital encoder display ~3m away from the cabinet. (Currently just a shielded 37pin dsub cable for my 3 encoders with differential signaling).
The only real guidelines I can find are that anything over 6ft should be differential lines. I will have 3 low power stepper lines (Nema 17) in my system, but in their own shielded cables. I'm not sure if the line length will be too long (~15m + interconnects) for noise reasons

Bonus Points. Should the encoder be able to supply enough power to signal both the controller and encoder display?

Manufacturer (Heidenhain) was fairly unhelpful (try it and see, which honestly if I have to buy new cables/interconnects/breakouts/filters at the end of the project will make me lose money, before that and I can charge my customer)
 
Replies continue below

Recommended for you

Cable length is one side of the story. PWM in switchers and drives is the other side.

If there are servos or drives utilizing PWM, then you have to make sure that all structures are very well bonded. Not just using cable screens and 360 degree conducting gable glands, but using heavy copper straps between machinery, cabinets and panels.

For real good results, avoid PWM altogether.

Gunnar Englund
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
 
Well, in my case I really only have some motor to stepper driver lines (5 phase, 90W). They use pulse train not PWM.
IF 20m isn't a big deal for a shielded differential signal then i'm really not too worried about noise sources.
 
I have successfully run quadrator (sp?) encoder signals 100's feet. The key is high quality shielded, twisted cable. We used a Belden cable that had multiple TSP (twisted shielded pairs) within a single shield and sheath.

Generally speaking an encoder should drive just one load, and that load is designed to send the signals to where ever they need to go.
 
Quadrature signals are much less noise sensitive than analogue (sine) signals that shall be interpolated for higher resolution.

The interpolator is 1 m away. So that part is probably OK. The resulting signal, which could be square wave or RS485, or anyting, is much less sensitive to noise.

Adding a display will not cause any problems. It is usual to use separate signals (using a signal splitter) to avoid disturbing the drive's actual speed value when working with - for instance - Measurex or Accuray units that also need speed information. Sometimes two encoders are used to avoid such problems. But that hardly seems to be the case here.


Gunnar Englund
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
 
djs,
thanks that's really what I was looking for that people successfully run cable that long, I was planning on using that exact Belden cable as well.

Skogsgurra,
I felt that was the case, thank you.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top