Edge Probing

[thosj]

Dave,

I bravely tried edge probing WCO's today. I have a Wildhorse probe, Tool 250, tip dia .125. BP conversion with knee CNC'd. The machine works well otherwise, and the probe worked with HossMachine screens previously.

Probing simple -X, no problem. I have X/Y and Z clearance set to .100. Fast feed 15 IPM, slow feed 4 IPM. It probed, twice, backed off. X showed it was at +.1625, where it should be, .100 clearance,  plus half the probe dia.

But, doing -Y or +Y, it probed OK, but when it got done and said it was at +.1625 or -.1625, it wasn't, it was further. It SEEMED to be off by half the X/Y clearance extra. Jogging back 'til the probe light came on at the edge seemed to confirm. So what have I got wrong here?

Also, and this is the bad one. Tried probing Z, it went down to touch, stopped, then seemed to want to apply the Z clearance negatively, it moved down!! I hit the big red button before it smashed the probe so I don't know how far it WANTED to move, but I'm guessing the Z clearance plus half the probe tip, or at least the Z clearance. Now, of course, I'm afraid to try again. I tried off the part, out in the air,  just hitting the probe manually, but I don't know the routine, so I get a probe error, probably because I don't know the timing to hit the probe, it moves so little. I tried with no slow feed, so it would maybe just hit once, but I couldn't get it to work by hand.

Now I did NOT calibrate the probe, I don't have a ring gage. Is that my problem? Not sure why X works and Y does not, it that's the problem. And no explanation for Z. I set the probe tool length offset manually with a 1" gage block after setting Z and using the block on that.

I don't need a negative Z clearance, do I?

[DaveCVI]

Some days.... I just posted that we have released beta 2 to fix the minor problems that have been reported.
I even tongue in cheek said that I hope that didn't annoy the software gods.... I should probably not have written that.
just as I got done with that task,  I got the notice for your post.... 

We do have multiple users running probing operations w/o nasty things like this happening - so I guess you get to feel "lucky" - sorry.
I just finished running a complete set of probing lib tests yesterday with Beta 2 (in prep for posting it today), All tests passed. So I believe it's not totally broken.  But none of that will make your experience better -

I shipped the probing library for beta with debug tracking enabled - so there is a good chance we can figure out what happened.

Some initial questions to help me get the broad picture:
1) Are you on a Parallel port or some other interface? If not PP, what is it?

2) Simplify things before doing any more probe testing. Set the Slow prove FR DRO to 0.
That will disable the 2nd re-probe op (see manual - we don't need complicated dual probe actions while debugging).

3) how bad did the probe crash? is it functional so that you can use it for debugging or is it toast?

4) does the big LED light up on the probing page when you move the probe tip?

5) PM me so that we can get in direct contact - that will make figuring this out easier and faster.

Dave

[thosj]

I am on a SmoothStepper

I will turn off Slow Probe Feedrate

Probe did NOT crash, Big Red Button hit fast enough

The LED on the screen and the LED on the probe itself light when the probe tip is moved. The probe has worked in the past with HossMachines probing screens/macros, so the probe and interface to the BB all seem well.

I've got family company this evening, so I won't be able to work on this tonight. I don't know what time you're at it in the AM. I'm in Central Time, 2 hours later than you.

This is probably something I don't have setup right, just need to figure out what. I'll bore something to use for probe calibration if that'll help. Let me know some things, other than above, to try, or be aware of and I'll do that. I'll be here all day tomorrow so perhaps we can figure something out.

[DaveCVI]

I'm on a PP.
I do have at least one alpha tester that uses a SS (Ray L., whose orig probing routines inspired the ones in MSM) so you're not in totally uncharted territory.

Don't worry about calibration- that only adjusts from entered physical probe tip diameter to effective probe tip diameter (includes cosign error for tip shaft tilt etc). We don't need to worry about that yet.

I would like to ask that you upgrade to MSM beta 2 that was posted today. That will get you and I one the same software rev.

Let me know via Private message how to contact you directly. I'm fitting MSM stuff in between my real job tasks and I have some stuff that has to get done in the morning, but I can probably juggle some time around.
Dave

[DaveCVI]

Also, have you got anything else installed?
I prefer to keep debug environments simple - just mach and MSM if at all possible.
If there are other things installed that have to be present (say to run the hardware), then I'll need to know what they are so we can think about if matter wrt to this problem or not.

Dave

[thosj]

I snuck off from family and did a little testing.

Installed beta 2, all went well, but nothing changed with my probing issues.

For extra stuff, I have my 4 brains to run analog pots for FRO and SRO and a bunch of buttons on my panel. I have an iMach III pendant that uses a plugin, that's it.

I set slow probe FR to ZERO. X/Y/Z clearance .100. I set probe feed to 5 IPM.

Probing -X seems pretty OK, touches, sets, backs off to .1625. My probe tip is .125, set in TO. Jogging back to touch X edge it should be at X+.0625, it's close to that, maybe .058. Not close enough for real work, but close enough for this test.

Y is another story, Probe -Y, it touches, sets, backs off to .1625, but it's way further away. Jogging back to the edge with the probe, Y reads -.020, should be +.0625, so it's off by like .082.

Z, I set up pink foam and a small washer to probe to, just probing to the foam, the probe just keeps going, #6 washer gives it some stability yet pushes in if the probe goes down.  Probe cycle goes down, hits, probe LED on, stops, starts and goes FURTHER DOWN all the while the probe LED is ON. It goes down maybe .1 or something, moves up, sets Z0, but it's not .1 above like Z DRO reads, it's closer. Hard to check with the foam getting pressed by the probe tip/washer! That second down move all the while with the probe active is what's super strange. Where's it going and what's triggering it to stopo. Seems like it is moving the clearance dim. but in the wrong direction.

We'll see about hooking up tomorrow to look at logs or debug info or whatever.

[DaveCVI]

Do any of the additional things you have loaded use any DROs, LEDs etc that are in the 1700 or 1800 ranges?

Dave

[thosj]

Not that I'm aware of, of course, the plugin for the pendant I have no way of knowing. Unless you can tell me how to figure out which might be used.

[thosj]

Lee Neuman at Vista CNC, maker of the iMach III, says his plugin doesn't use any DRO/LED in the range 1700/1800.

The only other thing is the pokeys plugin, but I can't imagine that using any either.

So I guess this is pretty indicitave that I don't have anything using any DRO/LED in this range.

Would MachMAD tell me this somehow? I have that installed.

[thosj]

Ok, tried some more this morning. I ran Mach 3 043.010 with my old 1024 screenset and probing routines, everything worked fine, the probe probed, the edges it picked up were right on. Z worked fine, too.

Back to MSM Beta 2, no change from last night. X edges probe pretty close, Y off .086, Z won't work at all because it tries to go down further after touching.

I probed a 1.000 bore for center. It told me the bore was .886xx diameter, it's 1". That's sort of off by the probe tip dia. so it didn't add that to the probed diameter, I guess. After it set X/Y centers to ZERO, I put an indicator in the spindle and indicated the bore and the center was off +.086 in X, -.023 in Y. Again, probe tool is 250, .125 dia. set in tool, length set in tool.

I'm mystified. I can't make heads nor tails out of what I'm seeing.

I now have to do actual work, so I'm going back to an edge finder and indicator for now. The rest of MSM seems to be working great for me, and I like it so much I REALLY don't want to go back to 1024 sets!!!