Machsupport Forum
Mach Discussion => Mach4 General Discussion => Topic started by: Tronn on June 25, 2018, 01:08:52 PM
-
So I have been trying to setup probing with Mach 4 using a contact probe, but I cannot even get the calibration routines to complete without getting the error message: "Can't transition from MDI running to stop." I am using a probe from drewtronics (https://drewtronics.org/s5000led) and a PMDX-424 board on a retrofitted Bridgeport M-Head. I have confirmed that Mach 4 is seeing the probe, as the indicator on the touch module does trigger when I contact something with the probe. The probe itself also has an LED indicator that turns on when contacting something. I am seeing this behavior in both the touch module and the probing screen set. Anyone else run into this problem?
-
Try changing your plunge rate when driving down to probe. Make it real slow to start with and if its already really slow make it a tiny bit faster. I had this problem ages ago. Pretty sure this fixed it. Let us know how you go.
-
I had some minor success. I tried setting the fast feed to 2 ipm and the slow feed to 1 ipm, as suggested, but still got the "Can't transition from MDI running to stop." message. I then tried the other extreme, fast feed at 50 ipm and slow feed at 45 ipm. With the faster values I was able to get the xy calibration cycle and the probe radius cycle to complete successfully and repeatably. I then tried playing around with the other probing screens and had success with bore probing, but back to the same "Can't transition from MDI running to stop." problem on corner finding. Does anyone have insight to why probing is so sensitive to feed rate during probing?
-
There are issues with how Mach4 and PMDX SmartBOBs interact when using the Mach4 built in probing routines.
Could you try running a G31 probe cycle by itself using MDI?
Please contact PMDX directly by email. We have a BETA plugin that is not in general release and would like you to test with it.
-
Thanks for your reply Steve. Email sent to PMDX support. I will try a standard G31 probe cycle when I am at the machine tonight.
-
Im seeing similar behavior with a G31 sent via MDI. When running these from MDI it will do one of three things:
1) Probe successfully and halt motion when the probe contacts the part. Motors remain enabled.
2) Fail to probe and issue "cant transition from MDI running to stop" message.
3) seemingly probe successfully, but motors are disabled after probing cycle. Mach 4 reports that a limit switch is triggered and disables the outputs. The limit that it is reporting is one that I have my motor faults hooked up to. This fault output is triggered whenever motors are disabled though, so I am unsure if the motors are causing the disabling, or if the disabling of the motors is causing the limit.
-
I am having this issue with the PoKEYS57CNC board. Sometimes it works correctly, but I am unable to duplicate it 2 times in a row. Sometimes I get the MDI RUNNING to STOP error, and sometimes it blows right through and breaks the bit before it stops with the error message. I am using metric measurements with the folowing settings, 100 probe rate, Retract 2.54, 500 prep feed rate, 2.5 prep distance, Tool Diameter 2.4, Touch Plate Height 10. Build 3804
-
Ok, I solved the problem with the MDI RUNNING.... error. I changed the wxMilliSleep command on Line 90 in mcTouchOff.lua to
wx.wxMilliSleep(70)
I think its trying to check too fast at 20ms. 70ms seems to work well. This is probably CPU performance based. As long as the probe speed isn't too high the slower test speed should do well. At 80mm/min, that is < .1mm @ 70ms.
-
I am having similar issues with my setup. It is a PMDX-424 and Mach4 with latest build 3804.
I think that I corrected the "Can't transition from MDI running to stop" issue by adjusting the wx.milliSleep value to 70. However, now I'm getting a touch, retract, set tool1 offset and then a LUA error. (see below)
Any thoughts?
Thanks
Mach4 History readout:
Z axis is performing a touch move.
Z axis is moving to a requested machine position.
Tool 1 Height Offset Set: -19.7255
Lua: Error while running chunk C:\Mach4Hobby/Modules/mcTouchOff.lua:1178: attempt to compare nil with number stack traceback: C:\Mach4Hobby/Modules/mcTouchOff.lua:1178: in function 'frameMainOnUpdateUI'
C:\Mach4Hobby/Modules/mcTouchOff.lua:912: in function <C:\Mach4Hobby/Modules/mcTouchOff.lua:911>
-
Hello RGUERI1, that issue was found in fixed in build 3805, please update and let me know if this addresses your issue, best regards Jim.
-
JimDingus,
Thanks, I'll try it out.
-
Can someone direct me to the 3805 version? The website's latest download is still 3804.
-
Can someone direct me to the 3805 version? The website's latest download is still 3804.
ftp://ftp.machsupport.com/Mach4/DevlopmentVersions/ (ftp://ftp.machsupport.com/Mach4/DevlopmentVersions/)
-
Thanks
-
Been away from my machine for a while. I've updated to the 3805 build, as well as the latest dev build, and now I'm getting an "Error(7) Motion Underrun." I have tried to reinstall profiles, adjust millisleep setting, manually perform G31 through the MDI, changed computers and adjusted buffer settings in the PMDX-424 performance config tab, all to no avail.
Currently I get the error consistently on the 3rd-4th probe. I am also using the modified M40 and M41 scripts for parameter logging.
This is getting very frustrating and I seem to have hit a dead end.
Has anyone had success in creating a probe level map file? I have also tried the auto generated probe file from "G-code ripper", "Autoleveller AE" and "Autoleveller AL." I have even manually typed the G31 commands in, through the MDI.
I am currently at a loss. Any help would be appreciated.
-
Hi,
I use Autoleveller AE. I had a problem initially that sometimes there would be recorded more data points in the probe log file than
there were G31 calls. I tried all sorts of ways to find out why without success. In the end I wrote a Lua script that could deduce the correct
number of probe points and correct, if necessary, the probe log file.
The full discussion of it is on the Autoleveller forum.
May I suggest that you double check that Mach has not inadvertently put more data points in the file than you expected.
Craig
-
JoeAverage,
I've eliminated the M40 and M41 calls, from the G-Code, until I can get it probing more than 3-points without getting the Underrun Error. So at this current point, I'm just probing and not writing to any file. As I'm typing this, I'm wondering if that is part of the problem. Last week when I was working on it, the M40/M41 calls were still in the code and I was getting the error at random probe points. Sometimes it would successfully probe 5-10 points and then hang up. However, now it consistently errors out at point #1 or #3. Not sure why it's chosen those two spots but I haven't successfully probed more than three points in the last 30+ attempts.
I'll put them back in and see if anything changes.
Thanks