Hi everyone, is been a long time between visits to this forum, but I've finally realised a need for probing on my cnc router.
I am using an Ethernet smooth stepper with the latest plugin for motion control.
To keep things simple and avoid complications I am initially using the standard Mach3 router screen set.
I have the tool on the router connected to signal gnd and pin 16 on parallel port2 connected to a piece of circuit board material. The input for the probe is configured in Mach3 to match these physical connections.
When I jog the Z axis tool down, upon touching the for surface of the board, it always lights the probe led in the mach3 diagnostics screen.
My probing results are inconsistent however, using G31 Z, for example, sometimes Z motion will stop when the probe touches the surface of the board (as expected) other times, despite the led in diagnostics coming on to show the input for probing as active, the Z axis to will continue to bury itself in the circuit board until I hit stop, or the Z axis motor stalls.
Other than this new issue, all of the inputs and outputs work well on what has been a workhorse router.

Try to set your denounce - But DO NOT set debounce TOO HIGH
ref
https://warp9td.com/index.php/faq/faq-mach3#Debounce

Try to set your denounce - But DO NOT set debounce TOO HIGH
ref
https://warp9td.com/index.php/faq/faq-mach3#Debounce

A long time ago, I tried a new type of controller called a USB smoothstepper. It had endless problems with electrical interference from the VFD that drove the spindle on my lathe, and I tried all sorts of filters etc to try to remove some of the noise. Greg eventually added the adjustable "debounce interval" to the driver and that was certainly a big help with that type of interference. When the Ethernet version of the smoothstepper came along, electrical noise was not such an issue because of the way that Ethernet works, but I have noticed that the debounce interval is still there in the ESS driver.
You are suggesting that I increase the debounce level, and to be frank I hadn't even thought of that as a possible cause - Ill try it now. I will set the probe debounce to 20 micro seconds. Thanks for the reply.
Dave

I was thinking debounce for the probe should be as low as feasible for better sensitivity.


BUT for future:-
I would recommend you make a probe using Quality NC (normally closed) metal switch for better reliability and absolute foolproof operation.

I am in process of making one (almost completed) even though I do not have no problems with my STD probe. But I once broke a bit because I FORGOT to clip the ground wire to the bit (an operator error).
With a NC switch, the top could be replaceable hard (cabinet door bumper) plastic to protect the bit tip and the probe could be used for non-conducting ceramic and diamond bits too. Switch opening to trigger is theoretically much most reliable.

See attached for a NORMALLY CLOSED probe

I was thinking debounce for the probe should be as low as feasible for better sensitivity.


BUT for future:-
I would recommend you make a probe using Quality NC (normally closed) metal switch for better reliability and absolute foolproof operation.

I am in process of making one (almost completed) even though I do not have no problems with my STD probe. But I once broke a bit because I FORGOT to clip the ground wire to the bit (an operator error).
With a NC switch, the top could be replaceable hard (cabinet door bumper) plastic to protect the bit tip and the probe could be used for non-conducting ceramic and diamond bits too. Switch opening to trigger is theoretically much most reliable.

I agree and when I get around to working on my mill I probably will make my own probe. I didn't mention my application in this instance, but it is to route circuit boards. Given that the surface of the board is clean conductive copper, the router is all metal and the sharp bit (my finger will attest to its sharpness) is steel, I think my setup should work.

See attached for a NORMALLY CLOSED probe

I think I've got a probe like that somewhere....

Quote from: reuelt on May 01, 2019, 01:04:37 AM

See attached for a NORMALLY CLOSED probe

I think I've got a probe like that somewhere....

I have just wired up my new probe switched the probe input to active high and tested that works very well. Of course I had to make sure that the retraction distance is higher than the new probe thickness in the code. Now I have a reliable and fool-proof probe. NO MORE non stop downward Z movement during auto tool zeroing for me again.