[planekrazie]

Hey Guys,
I am looking for 8 conductor 22awg twisted pair cable for my home and limit switches so I can run the cable in my cable carrier. Where in the world can I find this. It can even be 4 conductor. I tried to use 2 conductor non twisted going through the cable carrier with the steppers. Man what a mistake. I kept getting a " Limit switch triggered" in mach3. Took the cable out of the carrier and threw them across the table and they worked like a charm. HELP PLEASE,,
Steve

[tauntdesigns]

I had got some 6-con shielded from action, but it seems they don't have any right now. All they show in shielded wire is 2-con.
But they have 3,4,5, and 10-con non-shielded wire.

http://www.action-electronics.com/wirecable.htm

[jalessi]

Hi,

If you go to your local computer store and ask for a cat5 or cat6 patch cable (length of your choice) you will have a four pair twisted cable in your possession.

The super cheap ones might be solid conductor verses stranded wire, you can tell by its flexibility.

Hope this helps,

Jeff Alessi

[barbque]

Are you using Mach? If so make sure you have your Debounce interval/Index Debounce set. Its on the general config page. If its in a cable carrier with other siginal cables like drive and AC power it can pick up noise and trip your limit switches. I set mine to somthing like 1000-1500 to keep it from tripping.

[planekrazie]

You Guys are my heros!!!! Thanks
Ill reply with the results.
Steve

[mhasting2004]

Just a quick note.

Twisted cable will not stop noise getting into the line. The idea is that both lines will have the same noise and if the inputs are differential this noise gets ignored. Cat5 cable is not shielded.
Not sure exactly how you are hooked up but the best ways around noise are either:

Run higher voltages (like 24vdc prox sensors)
Run differential line driver signals
Run current loop

BTW if you do run shielded cable only ground the shield at the power supply end.

TTL 5 Vdc thru a micro switch is going to be very prone to noise if motor cables are anywhere near it

Just my $0.02 worth

Cheers

Mark

[jeffmorris]

Mark, Do you mean to ground the shielded cables at the power supply where the ground wire that comes from the AC cord is connected?

[mhasting2004]

Hi Jeff

Yes. Basically you don't want to create "ground loops" where current can flow from your power supply end, out to the signal end (limit switch in this case) and back via the shield or vice versa.

A common chassis and/or signal ground at one location is the aim (note sometimes signal ground is not connected to chassis ground (ie its floating) but each ground should still have its own common connection point). This is usually done inside your control / power supply cabinet on a terminal strip.

Stepper motor drive signals are very noisy due to the high current transiants. Putting is a small capacitor (say 0.1 uF) across any signal lines to ground can usually kill these spikes but it will also slow the response on the signal somewhat (may or maynot be an issue).

Cheers

Mark

btw chassis ground is the one from you AC input and stops you getting fried if something makes your machine "live". Signal ground is the low referance point for signals and may or may not be at the same potential of chassis depending on ones design. Ie opto isolation seperates the ground.

[jeffmorris]

Mark, I have Probotix PBX-RF board that has opto isolation. The shielded cable has a bare wire that's supposed to be connected to ground but which ground? The board has terminals marked "GND" in a few places.

[mhasting2004]

Hi Jeff

If it were me I would say the terminal closest to where the power (and ground) are supplied to the board. Alternately to a single post off the board with its own connection to the power supply ground.

Cheers

Mark

[planekrazie]

Lots of information here. Thank guys. Maby if I get this weekend off I am gonna re-run the limit/home switches. I picked up a few .01uF caps also Gonna try them too.
Steve

[stevespo]

I know this is an old thread, but did you ever get your limit switches sorted out? I'm fighting with the same thing right now.

Thanks,

Steve

[Schweinhund227]

Yeah ! let keep this thread going !

Putting limit switches is my next big step.... and I will try running the cables on the opposite side of the table.. to avoid noise contamination ! (flame2)

what did you get Normally Open switches or Normally Close ones ???

and since I am a noob at this.... what happens when those are triggered ?
Is the machine just stops ?

and Where can I plug a PANIC Button ?

I am using a HobbyCnC pro Board ?


Thanks guys !


Norm
out !

[Al_The_Man]

The best thing is to set up a common central ground point where the Service ground, P.S. grounds and Shields are connected to, together with any ground wires out to motors or spindle etc.
Al.

[Torchhead]

It's never a good idea to run low level signal wires in parallel and close to any cables carrying high current pulsed signals (motor power, Plasma cables, VFD supply wires, etc). Even shielded wires can have a level of crosstalk.

Proper conditioning can have a lot to do with keeping noise out of the logic (PC) side. In area that is often misunderstood is that Opto isolation is virtually worthless if you don't run the opto input side (with the switches) on a separate and floating ground system. In simpler terms the ground for the opto inputs is not shared with the same ground as the motor power. Most noise is "conducted" noise across a common connection (i.e. a shared ground). Common Mode noise (noise between a source and return) can be cancelled out only if it's a differential circuit. Twisted pair helps but not enough to fix things in a non-differential (on side referenced to the logic ) setup.

The inputs can have some bypass caps (we use .1MFD Mono) before it gets to the opto.
It would appear the the Probotix Breakout does have split supplies but the input is referenced to the Motor Neg side (VS the the Logic Ground side) so if you ground your motor side supply to the same side as the logic ground you basically have no galvanic isolation so noise can travel across one side to the other. It's all about the impedance of the grounds if they are tied together. The motor side has lots of nasty noise from driving all of the motors.

Software debounce is nice but solving the problem at the source is more effective.


TOM Caudle
www.CandCNC.com
Totally Modular CNC Electronics

[Al_The_Man]

Also remember that unless your PC is a laptop the 5v PC logic is at ground potential through the mother board screws, so the PC logic is going to be be at the same common or ground potential as your 120vac universal router spindle.
For this reason I usually bond any 5v common leaving the PC to the central ground point, so far this has eliminated any spurious noise for my applications.
I disagree with the shared ground theory, I use this method extensively and have not experienced any problems of noise, The only reason I use opto's is to transition from one one system to another, e.g. 24vdc to 120vac.
I also bond to ground my servo supply DC common.
Al.

[stevespo]

I think I've solved my limit/noise problems by grounding all my limit switches to the parallel port (pin 25) and installing the .1uF caps as a filter to the input and pins 18-21 (also parallel port ground).

I grounded my stepper motor shielding to the AC (earth) ground, but the limit shielding worked best when attached back to the parallel port ground. I know they both ultimately end up at the same place, but I think I was encountering what Tom called "conducted noise across a common ground". The single ground terminal did NOT work for me. Two terminals with the filter has been the right approach.

BTW - I am using a very basic BoB with no opto isolation or filtering. My noise problem has been present for years, but greatly aggravated when I recently ran all my wiring through Igus chain. I also installed a VFD/spindle and that generates a lot of noise as well.

Shielding, "proper" grounding and filtering with a low (100) debounce has seemed to fix my problems. I hope.

Steve

[CNCMAN172]

Steve,

Great to hear you ironed out your noise issues. They can drive you insane and I found the best way was to attempt to keep all my low TTL cables away from my AC and Servo power lines. I was constantly killing my system when I attempted to run the AC power line from my router down the IGES track guides. I had to remove the AC line and run it from overhead to avoid my servo drives from constantly erroring out. I have optical sensors for home switches and they are all connected using shielded wire attaced at the controller end. That is great but the AC power cord to my Bosch router was just way too noisy. Actually when I opened the router I discovered the normal A/C cords on wood working routers are never shielded. They basically have three wires, Neutral, Hot, and Ground. I was going to put in a new shielded cable to help cut down on the noise but think I am going to add a VFD spindle as well. I have a large 7.5HP Colombo spindle that I will be using on my big machine not yet under construction, however for my small 24" machine I needed a smaller router. I started with the Bosch 1HP router and that died in less than a month because I attempted to cut Aluminum. I swapped it out for the 2-1/2HP Bosch 1617 and it seems to work great but is way too noisy to run for hours on end. Which VFD spindle did you get? I have been looking at the Chinese 2.2KW water cooled units as several people have said they seem to work very good and the noise level is significantly lower than standard wood routers. On my large Colombo I purchased a special Indrant double shielded cable and it seems to be pretty quiet, noise wise. Actually the spindle makes less noise than my 2-1/2HP router amazing. Anyway would be interested in which way you went. Thanks

Russ
CNCMAN
:rainfro:

[stevespo]

Russ, yes I was on the verge of going nuts from troubleshooting the noise issues. Chased down a lot of dead ends, but all it really took was $3 worth of capacitors. I was planning on running shielded cable everywhere, so this was a good opportunity to get that taken care of.

I have the 2.2kW Chinese spindle and matching VFD. There is another thread with a fair amount of info in it ("Cheap 1500W Chinese spindle", or something like that). So far it's been great, but I haven't run much work through it. It is quiet and powerful, definitely better than my PC 890s. The water cooling is a bit of a nuisance to setup, but once it's going it's no big deal.

I'm temporarily running 16-3 unshielded "tool cord" to the spindle, and there is a whole lot of noise. Luckily my cap filter has taken care of it until my shielded wire arrives. I bet it would help out your situation as well. I used a .1uF for each input.

Steve