spindle electrical noise

[balsaman]

Guys,

The electrical noise from my router/spindle on my machine causes my limit switch circuit to false trigger. I put a ferrite donut on the limit switch circuit cable with no luck. The limits are NC and the input goes low when there is a problem. All the cabling on my machine is shielded EXCEPT for the limit switch circuit. Is there a better solution of do I need to rip out the cables and replace them with shielded?

If I turn down the spindle RPM the problem goes away. (80% or below is fine).

Is there something I can put in the sindle circuit that may help?

Eric

[bunalmis]

You can use noise canceller at the button inputs. Also this is good for debounce problems.

Yes you must use shielted cable for all input pins of controller.

[Mr.Ed]

Eric,

Do you think it's purely an electronic problem, like strong (electro) magnetical induction in cables, or is it a vibration in the construction that might not be detected by touch or hearing. That could allso trigger a switch. Only one occurance would halt the software as well.

With an oscilloscope you could try to find irregularities in the signals. The ferrite ring was the first most obvious thing to try. Would it be a good idea to ground your machine to a clean earth? Ring shaped constructions could run large induction currents, like your gantry perhaps? You could go for magnetically operated switches like reed relais?

I dont know, it could be a million things, just trying to think my way out of this one

Keep us posted, this could affect more builders.

Good luck...

Ed.

[Mr.Ed]

Come to think of it,

You tried to find the solution near the switches. Wouldn't it be more obvious to try to solve it near the source of the trouble?

Maybe putting more ferrite cores close to your spindle? Is the spindle earthed, like maybe the rest of the machine? If it has a steel / aluminium case it probably should be?

I know my spindle has a plastic casing, so no shield there!

Ed.

[ToyMaker]

If your spindle motor is dc, mount small capacitors (0.01 uf or so) on the motor terminals. Use 3: one from each terminal to the case of the motor, and one between the terminals. This will reduce a lot of the brush noise.
Since your limit switch is n.c. and goes low to activate an error, there must be a pull down resistor, you could connect the n.o. contacts on the switch to ground and use a pull up resistor (5-10k) on the ckt. This would solve a vibration problem.

robotic regards,

Tom
= = = = =
A clinician learns less & less about more & more until he knows nothing about everything. A researcher learns more & more about less & less until he knows everything about nothing.
- - anonymous

[HuFlungDung]

I had a similar problem, and it was due to vibration. I replaced the "El cheapo" limit switches installed by the original retrofitter, with some heavier duty units and solved the problem.

[balsaman]

It's an AC Porter cable spindle. I dont *think* it's vibration. I jumpered the z axis limits and still had the problem. I beat on the machine with my fist and had no faults.

I suppose I need to run shielded cable.

The ferrite donut is installed as close to the breakout board as possible.

The machine and spindle are grounded.

How about a small value cap in parallel with the limits?

Eric

[creative_mind]

I'd go for bunalmis' circuit. I think you're dealing with transient noise, actually it's a pain in the you know what.

Keep it simple, use a 4.7k pull up resistor, that is one end of the resistor tied to +5 volts and the other end to the input signal, and a .01 microfarad capacitor tied to the input and ground.

Use it near the input and of course, by all means use shielded cables. Sometime it's a good idea to use ferrite core in series of the input signal too.

I've had a similar problem testing high speed actuator for transient analysis for aircraft applications where the transient spikes would mess up the computer.

[balsaman]

There is a pullup in there. I will let you all know how I make out. Thanks for the help.

Eric

[balsaman]

Problem solved with a cap between the limits and ground at the terminal strip. Thanks everyone!

Eric