Jack -
I'm sensing as electrical noise problem ....
Please see below
Dave,
Got the probe hooked up but it wont probe. It moves slightly but then probe error is Z not found. Switched back to touch plate and it works fine.
So a simple probing operation start movement, then movement stops and MSM says Z not found.
OK, first, let's switch from Z to X for test movements - much less chance of crunching a probe that way.
Keep the probe tip depth just below the objects top surface that you test with - that way if the probe does not stop it will just scrape lightly across the top. This is why I say test in X not Z.... X will scrape across the top, while a Z failure will smash the probe tip into the test object.
Is the error behavior consistent and independent of axis?
A touch plate and a probe are opposite re electrical interface. The TP is active low (I.e. the input signal is +5V when inactive and goes to 0V when the tool touches the plate).
A probe is low when inactive (it's a closed switch in the inactive state) and when triggered, it goes high (switch opens).
Thus noise issues will show up on the probe more than when using a touch plate. A small, quick noise pulse at the input will look like a probe trigger to mach (for the probe whihc is active high)- and mach stops motion when it see a probe trigger input. For a TP, the same noise is just more "I'm not active" to the input and so it does not cause trouble.
I disconnected the probe and just used the wires. I got the same result so it isnt the probe itself. The amount it moves varies a lot from very slight to an inch and I stop it manually.
This is often not a reliable test - the problem is that touching the wires tends to create lots of short input pulses - and any one of them can look like a trigger signal. See more comments below.
It seems to be when I switch from low to high, I think, the probe distance changes. why would it say Z not found if it got a probe hit?
Are you set up for single or double actions?
If you are set up for double, please change back to signal while debugging (set the slow probe FR DRO to 0 in the probing parameter panel).
For why it would say this, please see section 8.11.1.1 of the MSM manual.
Noise on the input can make the mach probing sequence fail - and what you describe is the condition MSM sees when that happens.
Sometimes the probe active light flashes when nothing is happening.
This is the most telling part of your post for me. This tells me you have a noise issue and that mach is seeing noise as probe triggers. There should never be any flashes of the probe LED when the system is idle.
I'd bet that if you looked at the input line with a scope you'd be seeing noise on the line.
I check the debounce and it was set to 2500 along with debounce index. I changed them to 500 but it didnt make any difference.
That is a LOT of debounce. Each number represents 40 microseconds (with mach running at 25KHz kernal speed), so 2500*40us = 0.1 seconds.
with that setting, mach should only be looking at input signals that are longer that a 1/10 of a second - that's an eternity for a logic signal.
Given that you see the LED light flash when nothing is happening and the debounce is set that high, I suspect you have a serious noise problem.
BTW, a lower debounce number means mach will pay attention to shorter input signals - so making the number lower would not improve the situation.
For test purposes, you can try an even higher debounce value in mach - at what debouce level do the random LED flashes stop?
NOTE:
DO NOT try to run a probe operation with really high debouce values! The higher the debounce, the longer the input signal has to be present for mach to see it as a probe trigger signal. That time also means that the probe will overrun the trigger point by more - as the spindle keeps traveling during the debounce time. That can result in more overrun than the probe can handle.... the amount of overrun is a function of probe velocity and deceleration for the system. This is another reason I say test with X, and keep the probe depth shallow - it lowers the chances of crunching the probe.
Do I need the probe led fix like in the manual?
Probably not, but I don't know what the input circuit of the control you have looks like electrically.
While I strongly suspect at this point a noise problem, it could be an interface problem between the probe and the control which in turn makes the input sensitive to noise. I have no way to know without knowing what the interface circuit in the control is.
Does the control vendor provide schematics for the electronics input for the probe signal connection?
If so, I could take a quick look to see if I spot anything obvious.
However, I'd guess that you may need to have a discussion with the control vendor about this - if they are the only ones with schematics, you pretty much have to talk to them to get info.
Any suggestions.
Thanks
Jack
The above may not be what you wanted to hear... but it's my current educated guess given what you've told me.
Dave