Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: RNC on November 06, 2015, 05:21:50 PM
-
I'm assuming this is not a mach issue but this being a very active forum, figured someone would know where to look.
Yesterday, I had a random situation where the Y axis zero on my machine didn't work. In that, the machine hit the limit switch and stopped moving, but per the DRO mach was still sending a movement signal waiting for the Y to hit its home stop. The E stop in mach did not trigger on the Y, but does trigger on the other axes.
I turned the machine off/back on and referenced all to their home positions again and it worked fine, didn't think much of it, maybe a glitch.
Today, the same story but the y limit never trips in mach. I manually observed the signal inputs while hitting the switches with my fingers, and sure enough, the Y axis limit switches don't show a signal in mach, but the switches do work, because the machine stops when I jog to the limit manually.
This is on a Sieg KX3.
I'm guessing perhaps bad pin on the board or internal cable? That's my best guess...
There's an ethernet smoothstepper so the connection from computer -> machine is an ethernet cord. I can ping the smoothstepper with no loss so there's no issue with that.
-
suggest you check the ports and pins settings for inputs that they are still set.
then have a look on the diagnostics page and see if they are tripping
(presume you have physically tested the switch with a meter...?)
... I've heard of this but never experienced it... you may have a bad / corrupt profile.... suggestion on that one would be to add / install a new profile (don't over-write / delete your existing one though! >>> Go to start >> Programmes >>> Mach3 >>> Mach3 Loader and create a new profile.... then install the smooth stepper driver off their website (given it was working before and not now its not an issue with the plugin more likely) http://warp9td.com/index.php/sw#PluginEssMachThree ... but download and install the latest version... then setup just your input pins ..... I'd suggest closing down that profile, opening your existing one an copying the pin mapping .... then open up the new one and copy across the mapping (just for the inputs as all you want to do is toggle them).
Once you've done that (note you will need to restart M3 so that the ESS can reload the settings to the FPGA) ... toggle your switches and see if they work.... if they do... then its probably a profile thing (and you need to recreate all of your settings pins ports and motor tuning etc just by copying them across) .... if not .... a hardware thing.... switch wiring .... breakout board (optoisolator if your bob has them) ... of the ESS ...
Method of elimination ... may want to flip your wiring between the troubled axis and another pin on the ESS (physically) and see if that clears your fault.
Rob
-
Yeah, the profile hadn't changed, like I said, worked one day and not the next. That's why I'm assuming this has to be a hardware failure.
Corrupt profile? Can that happen randomly? I mean, it's just a text file.
-
Like I said I've read of it but never experienced it.
Can you swap the wiring at the ESS for the switch if so and the problem goes away it may be the physical pin or profile, if not its cabeling or faulty switch (oxidisation of contacts)
Yes profile is just a text file
Hopefully you've got a few options to troubleshoot
Rob
-
If it's any help I have had a corrupt .xml / profile on two occasions in the past whilst using the ESS. Probably my fault and caused by making setting changes.
Tweakie.
-
TWEEAKIE, Unless YOU went into teh XML and made changes there is no way that mach3 SHOULD currupt teh XML. BUT there is a bug there that causes it. It is NOT your fault.
This happens with the LPT as well.
It is a LONG STANDING BUG.
(;-) TP
-
TWEEAKIE, Unless YOU went into teh XML and made changes there is no way that mach3 SHOULD currupt teh XML. BUT there is a bug there that causes it. It is NOT your fault.
This happens with the LPT as well.
It is a LONG STANDING BUG.
(;-) TP
what evidence should I look for of this?
I have a backup of the working config, here's the diff. I don't see anything here that seems grossly out of place (the input differences are me turning off the Y home so that the machine can be used, by manually zero'ing Y).
The OEMDROs are I assume what is included from Sieg, I built a config from the one they distribute.
Inches = original
production = the one I'm using
diffs = diffs
suggest you check the ports and pins settings for inputs that they are still set.
then have a look on the diagnostics page and see if they are tripping
(presume you have physically tested the switch with a meter...?)
Like I said, one day works, next day doesn't. Nothing had changed in the config. I did not check the switch with a meter because it still works, as far as stopping the motors when it trips. There is no 'grind' as the motors try to run past their thread limits, they stop when the limit switch is triggered as they should, but the input in mach isn't showing up. Yes I tried tripping the switches with my fingers while watching the input signals, notta on that axis, the other axes do show a signal and do work. That's why I'm thinking a bad pin on the board or the smoothstepper must be the culprit, up to there everything works, the signal just isn't making it out from there to the PC.
Switching the pins on the ESS is a good idea and could eliminate the dead pin theory, didn't think about that, thanks!
-
Update: opened the machine up, re-seated the internal ribbon cable, now the switch works again.
So as I suspected, bad connection at the pin itself.
Time for a new cable.
-
TWEEAKIE, Unless YOU went into teh XML and made changes there is no way that mach3 SHOULD currupt teh XML. BUT there is a bug there that causes it. It is NOT your fault.
This happens with the LPT as well.
It is a LONG STANDING BUG.
(;-) TP
Hi Terry,
The ESS also writes to the .xml and I think that is when the corruption occurred.
(I have never had a corrupt .xml when not using the ESS).
Tweakie.
-
Update: opened the machine up, re-seated the internal ribbon cable, now the switch works again.
So as I suspected, bad connection at the pin itself.
Time for a new cable.
Oxidisation ?
Can also occur on the switch contacts given there is no high current flow when the switches make and break they are reliant upon the mechanical abrasion of the two sliding contacts to clear any oxidisation.
Rob
-
Can I make a suggestion on your production XML file...
Your Axis are running as follows:
X-Axis = 12700 steps / inch, with a maximum velocity of 59.064" / min, so that is ~12.5kHz, [59.064 / 60 x 12700]
Y-Axis = 12700 steps / inch, with a maximum velocity of 59.064" / min, so that is ~12.5kHz, [59.064 / 60 x 12700]
Z-Axis = 12700 steps / inch, with a maximum velocity of 38.378" / min, so that is ~8.1kHz, [38.378 / 60 x 12700]
Now when you look at the ESS Config, you have
X Axis = 32kHz (ok, that's the minimum setting)
Y Axis = 32kHz (ok, that's the minimum setting)
Z Axis = 4MHz .... suggest this is changed to 32kHz
In fact I'd suggest changing them all to 32kHz (even the axis not used, just in case the ESS sets up a loop at those frequencies regardless of if the axis is enabled or not, as it will do no harm).
There is now a good explanation here on the Warp9 website (suggest reading the rest of the whole page too as it's more informative than the manual ever was):
http://www.warp9td.com/index.php/faq/f-a-q-mach3#MachThreeStepFreq
Your controller frequency is set to 250hz... you may wish to improve that to 1khz (unless your PC is not up to it), again there is an explanation on the Warp9 website here (actually above the post I listed above... hence suggest you read the whole page
http://www.warp9td.com/index.php/faq/f-a-q-mach3#MachThreeControllerFreq
You also have debounce set on the Mach3 Config page.... but if you want debounce with the ESS... you need to set it within the ESS config ... explanation here:
http://www.warp9td.com/index.php/faq/f-a-q-mach3#Debounce
And finally you have the lookahead set at 20 lines on the mach3 general config page (but as I noted above you have the ESS controller frequency set at 250hz (~4 second buffer) ... so it's looking for a buffer, but mach3 is only looking at 20 lines for the motion planner ... suggest increasing this to 200 lines... bit of a guide here too for you to read (some of the other settings may be worthwhile to know about even if you don't use them:
http://www.machsupport.com/wp-content/uploads/2013/02/Mach3_CVSettings_v2.pdf
Hope it helps
Rob