Centroid Community CNC Support Forum

tblough wrote: ↑Mon Feb 04, 2019 2:38 pm Noise is extraneous voltage induced in the wires. If your switch is NO and the controller is looking for a voltage signal, a stray 1-2V spike can set it off. If the control is using a 24V NC switch, the control is seeing 24V when the is switch not made. A stray 1-2V spike just changes that from 24V to 26V - still not made. It's pretty easy for the control to notice the absence of 24V even with a few volts of noise in the system.

Tom that is the most clear & common sense explanation of the issue I've seen. Thanks for posting! I wish Amazon would introduce a "Tom the CNC Teacher" device. Alexa's smarter brother. I'd buy one in a heart beat!

I'm wondering if there might be a misunderstanding of how the Acorn INPUTs work? Page 5 of the Acorn CNC Controller Specification Manual shows that external devices pull current through an opto-coupler that (internally inside the Acorn) activates a signal. If nothing is connected to that input, the LED will never turn on. Noise would have to be extreme. It would have to pull about 10mA through that LED and the duration of that noise would have to be long enough for the Acorn to pick up that false signal. Page 7 of the manual shows that a LOW signal is 0 to 5.9 volts and a HIGH signal is 19.1 volts to 26 volts. Page 7 also gives the operating current as 20mA. You can test the minimum current needed to produce a signal by using a resistor with one end tied to GROUND and the other end connected to an Input. A 1K resistor allows about 20mA to flow. 2.2K allows about 10mA to flow, and so on. (To verify the amount of current required, put a multimeter - set to read current - in series with the resistor.) With the CNC program running, push ATL-I to get to the screen that shows Inputs and Outputs, and watch the Input to which you've connected the resistor. At some point, the LED will not change state, indicating that not enough current is flowing through the LED to generate a signal.

It's always a good idea to use a shield cable with the shield tied (on one end only!) directly to chassis ground. So far, on my test bench, I have not had any noise problems with the Acorn board just using twisted pair wires.

mick41zxr wrote: ↑Mon Jan 28, 2019 6:48 am Check with Keith,
I think there is an issue with home/limitOk setup at moment. I had an issue where home/limit kept re-tripping on axis 1 across a variety of switch types.
Only happened in first axis. Changed setup to home all & problem went away, just used soft limits after that.
Regards,
Michael

This is the only thing I’ve changed. Just ran the exact same gcode that I had multiple limit trips in when I first started this thread. Zero issues tonight. 90+ minutes straight of cutting. I feel like this home/limitOk vs homeOk is the issue. So, that’s how I’m running.

I did notice the following tonight.. inputs 1-3 (NO homeOk for x, y, z axis) are lit green and turn red when the limit switch trips. Input 7 (probeTripped configured NC, but the switches are actually NO) is green on a fresh system startup until I actuate the probe switch, then it goes red unless the probe is tripped. It boots up latched on until it sees a state change, then it works as expected.

I’m not replacing my NO limit switches at this time..

Screen shot of the diagnostics page please.
Do you have lines over the LED on inputs 1, 2, 3?

Marty

On power up. Input 7 is green. Line over 1-3 and 5.

Limit switch inputs have been inverted, thats why the are green (NO)
Input 7 is probe tripped. Do you have a probe on the machine?

Marty

Yes.

slodat wrote: ↑Tue Feb 05, 2019 1:35 am Input 7 (probeTripped configured NC, but the switches are actually NO) is green on a fresh system startup until I actuate the probe switch, then it goes red unless the probe is tripped. It boots up latched on until it sees a state change, then it works as expected.

Slodat,
I have seen this same exact issue. I am using a TT (touchoff) tool post and when I first start up the indicator is wrong. I have to press the touch off post once before a do any work. That seems to toggle the software in CNC12. I first thought it was something wrong in my touch off indicator but I checked it with a meter and it worked correctly.

I was going to make a video of this and send it to Keith and the development team. Seems to do the same thing if you use input7 for a probe in your case or Input4 for the TT in my case.

Russ

General Reference:
viewtopic.php?f=63&t=1023

See Keith's last post

martyscncgarage wrote: ↑Tue Feb 05, 2019 12:40 pm General Reference:
viewtopic.php?f=63&t=1023

See Keith's last post

I have my NPN NO switches wired individually to inputs 1-3.