rewiring to minimize EMF and heat

[BarryB]

Hey folks, the time has come to rewire Milla, my 6 axis mill.  When looking at the wiring job I did, it's surprising she's worked this well until now.  After doing my latest aluminum job, she's getting a lot of EMF.  Pict included of the wiring.  I was wondering if you wall could mention your preferred wire, control box, and wire connectors.  My initial wiring didn't include shielded cabling, so right there will be a huge step up.

Still though, I'd like to know how to turn my previous wiring job which is so messy, but it works, into something neat, manageable, and no crosstalk.

Also included is my wiring diagram, so you can see how complicated it is.





Barry

[Hood]

Shielding any 5v signals and also high voltage lines will help a lot. Also if possible keep the 5v signals to a minimum and use 24v volt for such things as lmits. This may not be practical for you as it will mean you need to convert that to 5v just before it enters your BOB but it definitely helps make things noise immune.
On my machines the majority of my wiring is unshielded as its nearly all 24v but any 5v signals and also high voltage to servos and  encoders wiring is shielded.
 I am lucky in that my servo drive  use 24v I/O and also I have a PLC so I can use 24v to it and then input to Mach via the serial modbus so that helps potential noise problems. Aslo my servo drives take differential Step/Dir signals so I convert the single ended ones from Mach to differential and this makes them much less prone to noise so I dont need to shield them.


Hood

[BarryB]

Do you recommend any brand or make of wire out there?  I want to make sure I don't screw this up again.  Also what enclosures do you find are the best and most affordable?

Barry

[Hood]

I just used whatever cable was available to me rather than a specific brand. The important thing is that it is not solid core as that can crack due to vibration, so the more strands the core is made from the better. As for shielding then best is wire braid shielding for any wires that regularily flex such as motor/encoder wires as foil shielding can deteriorate through time if constantly flexing.

For enclosures, I just used what I had, for example the Bridgeport had two cabinets so I used one, the lathes cabinet I used but the Beaver Mills cabinet was huge so I had a spare sheet of Alu that I welded up into a cabinet.
Here is a pic of the Computer and I/O part of my Beaver mills cabinet, all the drives and contactors etc are below where this board fitted in my cabinet. As can be seen I am not using any shielded wire in this pic as all of these wires are 24v except the Step/Dir which are differential and the spindles encoder inputs which again are differential.
The bottom pic is the whole cabinet, almost wired but not tidied, you can see the shielded cables that go to the drives and I have kept all of the high voltage wires well away from the few 5v signal wires I use except for the encoder wires and these are well shielded.


Hood

[BarryB]

thanks, that has given me lots of ideas.  I found the crosstalk that was causing the EMI too, and it goes directly which what you are saying.  So for planning the new control box, I will isolate the low, mid, and high voltage wires to their own areas.  I'll also shield them as well.  The control box itself I bought to replace this is also shielded too, so anything in the room won't interfere too.  I'm not taking chances this time.  I plan on making interconnects for everything outside the box too, so I won't have wires just going through a hole, or several holes, including the open side as my current scheme is.

Thanks for your help.

Barry

[JHChoppers]

Shielding any 5v signals and also high voltage lines will help a lot. Also if possible keep the 5v signals to a minimum and use 24v volt for such things as lmits. This may not be practical for you as it will mean you need to convert that to 5v just before it enters your BOB but it definitely helps make things noise immune.
On my machines the majority of my wiring is unshielded as its nearly all 24v but any 5v signals and also high voltage to servos and  encoders wiring is shielded.
 I am lucky in that my servo drive  use 24v I/O and also I have a PLC so I can use 24v to it and then input to Mach via the serial modbus so that helps potential noise problems. Aslo my servo drives take differential Step/Dir signals so I convert the single ended ones from Mach to differential and this makes them much less prone to noise so I dont need to shield them.


Hood

Second that:  Also make sure you connect the shielded part of the cable to chassis ground on both ends (sometimes only on one end for ground loops, but the rule of thumb is both ends)   

You can also RC filters to limit and home switches to keep them from false triggering.  Even if you only add additional pull ups, this can help your signal to noise ratio.  Depending on the application, how you connect the switch to NO or NC will make a world of difference to keep switches from falsely triggering Mach3. 

As Hood stated, a 24V PLC is the best.

JH

[GeorgeRace]

I have a small 3 axis home built mill using a Hobby CNC control board.  Initially I had a lot of false triggering of the homing switch's and it was difficult to home without a stop along the way on one or more of the Axis.
I ran shielded wires for all the limit switches and that really helped a lot.  But from time to time I would still get a false stop when homing.  The last thing that I did solved all my noise problems.  I put a .01 mfd and a 2 mfd capacitor on the circuit board where the wires from the home switches connected to the board.  The capacitors are between the active pin, the one with the pull up resistor, and the circuit board ground connection.  Like magic all the false triggering has been gone since that last addition. 

The capacitors are very small and can be obtained from your local Radio Shack or on line.
George