Incremental Probing still buggy in Ajax Plugin 1.16

[bonestock]

Setup:
Win7
MPU11+GPIO4D
Mach 3.43.22 + Ajaxplugin 1.16
Gorton 2-30 Knee Mill

I have noticed some interesting issues while working with the Ajax G31 op in the process of adapting MachStdMill probing operations to this system. First as it directly relates to the topic, G91 G31 operations are very inconsistent. About 7 out of 10 times it will ignore probe contact altogether and will complete travel to the specified Incremental position. Absolute probing via G90 G31 however will work perfectly. In reviewing the release notes it seems some G91 G31 related defects were address in release 1.16. I am not sure what specifically was addressed but I clearly have a few more for the next release.

Interestingly enough the instability of G91 G31 and even G90 G31 increases significantly when I enable a simple Brain to Pass an inverted ModBus 481 signal to the Mach Probe Switch Input. This is a very simple brain, but is required for many aftermarket probing routines that are dependent on the Mach Probe LED status.

When this Brain is enabled G91 G31 will occasionally output some very strange values to OEM DROs 2000-2005. In my debugging I have discovered that somehow the OEM DROs that should have probe contact position instead have WC Offset position. As you can imagine this is very unusable but is a fairly repeatable process (3 out of 10 times). This issue does not seem to occur if the Brain is disabled.

Additionally G90 G31 will occasionally fail to update the OEM DROs 2000-2005 completely when this simple brain is enabled.

The brain file is attached. It's nothing spectacular which is what makes this so interesting.

I am open to providing some additional debug data if needed.

-Shane

[diycncscott]

Hello,

I'll look into the incremental probing to see if I can coinfirm any issues.

Regarding your simple Brain to echo the probe input over to Mach -That is not going to work. Doing that will cause Mach to step on the 2000-2005 DRO's which will conflict with the values returned by the MPU11 to those very same DRO's.

[bonestock]

Regarding your simple Brain to echo the probe input over to Mach -That is not going to work. Doing that will cause Mach to step on the 2000-2005 DRO's which will conflict with the values returned by the MPU11 to those very same DRO's.

Hmm, That makes total sense now that I think about it. Back to the drawing board.

[bonestock]

Scott out of curiosity what would be the level of effort to toggle the Probe Switch LED from within the plugin? My thoughts are that Ajax could better control the OEM DRO values but preserve standard Mach functionality.

If this is not possible, perhaps you could elaborate a bit on the design decision to use oem dro 1999 instead.

[bonestock]

to the top...


Would it be feasible to have the Ajax plugin toggle the Probe Switch LED in Mach providing a work around for the DRO clobbering that occurs when it is triggered via a Brain?