Mach3 just stops mid code?

[MRob]

OK! Got set up with me AXBB using Mach3 a couple of months ago, all good but in the last few days I've started having an odd problem. Suddenly all motion will stop, but Mach3 will keep on running happily through the lines of code. At first it happened just once or twice, and I would have to restart the machining process, but today it happened multiple times and just now at least 4 times on the same gcode file. It was difficult just to finish the machining. Each time it happens, I have to press stop, rewind, and start again from the beginning. I've not changed anything in my setup, no hardware or software changes. There are no differences in the lights on the AXBB when it happens, green/blue/green and the network carries on flashing like all is well. I'm running windows 10, getting updates from Microsoft but the last one was 13th Nov, and I dont "think" the issue started then. It seems to be getting worst anyway. My windows install is pretty clean, with just a few browsers and Onedrive installed. I could try shutting down the latter and see if that makes any difference, but unlikely. Finally, I was running Mach3 for the last couple of years without any such problems, although I did install the AXBB only after doing a workshop move.

So! Any ideas? Any way to get a log file from the AXBB? One thing I will try tomorrow, if/when it happens again is to let the program keep running until it completes, even if no movement, and see if it turns the spindle (Huanyang, USB->serial) off at the end as it should. I'll also double check the cat5 and make sure it is fully clear of all spindle motor wires (cable is shielded but still).

[MRob]

OK changed USB cable and made sure it is well clear of wiring that might create interference, issue still present. Also let the program continue running after movement stopped, and when it reached M5 the spindle turned off. Corrections from previous post, when movement stops the program keeps on running, but position as displayed in Mach3 remains the same. Also thinking more about this, although the problem has gotten worst in last few days, I'm pretty sure I did notice it a few times when running some much longer programs before. Finally, I said all was well until I changed over, but I forgot I was using a different PC then with Windows 7. Am now running win10 on an old laptop. So maybe the problem is there, but I would really rather get it fixed than switching over to UCCNC right now, which I have never used before and would probably take a few days to get sorted...
Sooo...... anyone? Care to comment?? Is there anything at all I can check???

[ger21]

Did it start doing it as soon as you switched PC's?
Is there any message in Mach3?

[MRob]

No message in mach3 to the best of my knowledge, pretty sure I checked that but will double check tomorrow. However I think Mach3 does recognise there is a problem, as when it has stopped it refuses to run my M code for Z probing, just beeps at me. At least, I assume that it is the same bug.
As I say above, I think it may have done this once or twice on very long programs in the past after switching PC, like a month or so ago, but in the last couple of days its become hard to even finish a program.

[cncdrive]

Did you try a different version of Mach3? E.g. R3.043.062 or R3.043.066, those are the most stabile ones which we know about.

[MRob]

Confirmed no error messages showing the history, and am running .062 - you think I should install .066?

[Robertspark]

Are you running Windows 10 could this be the problem

https://en.industryarena.com/forum/mach ... 60714.html

[MRob]

Interesting, but .062 apparently contains an update to the problem discussed, which is itself very different to the problem I am experiencing. I am running win 10 but had thought/hoped that the issues some people have with higher than vista etc were due mainly to the parallel port, which obviously I am not using with the AXBB.

[MRob]

OK, I'm pretty sure I've diagnosed the problem - transient spikes from a dodgy e-stop switch. I had connected the NO side of the switch to the AXBB, and the NC side to another device I run. After I swapped them around and the other device started getting triggered the same. So, my best guess for what was happening, is either that the AXBB was seeing the e-stop triggering for long enough to stop the axis from moving, but not for long enough to inform Mach3 that it was in e-stop. Or mach3 was not dealing with the very brief e-stops in the correct way. But then, I have no idea how the AXBB is set up, like does it know which input is e-stop? When I physically pressed the button it would stop fine, and tell me it was in e-stop. I first noticed there was an issue when running UCCNC, it was seeing the e-stops properly and telling me about them, so just Mach that was getting confused and giving me the odd behaviour.

Also, an aside, I have my probe connected with 12V to an AXBB+ terminal, and an AXBB- connected to the touch plate, which is then triggered when the grounded end mill touches. However I noticed that if I touched the touch plate just with my fingers, the probe would often trigger. Connecting a 12K resistor between the two relevant input terminals, pulling the -ve up to 12v, solved this issue - bit surprising though as I thought the terminals were supposed to be pulled up already. But hey, sorted now.