reading Pokeys pins

[gburk]

Art

Did al little testing, its some what confusing to me, so I will try my best to explain..

The probeinvert keeps it setting now when rebooting..

Tested the probe in the MDI one line G31 Z-1 F10, second line G0 Z0 F30, started to probe with dro's set to 0 so it probed till touch plate hit and stopped, ran the second line and it retracted back to start position 0..

Ran with the probe button and got mixed results, sometimes it would only probe down a real short distance and then retract not hitting probe, one thing I did seem to find if I ran it first in the MDI then the button script it did probe down and retract ok.. but never the same results more that once or twice. then it would go back to the old way and hit plate and not retract it would dive into plate..

It still seems to me that the system is getting lost as to where it is, not sure if its still between the mach cord and work, but it seems to get thrown way off even though the DRO's are zeroed

Also I ran into a problem with the MDI at a point, it seemed to stop working the script button would still run and axis move but the G31 Z-1 in the MDI stop working no axis movement..

I think you are close with the probing..

I also ran simulator mode and had a lot of problems in that mode same as machine mode  but I took it a little further and was finding that if I loaded a gcode file and ran it, it would run maybe 1/4 or more though the file then auggie would crash, and I would get the dialog with a bunch of error lines and when I close out the dialog auggie would shut down also, I tried 2 different gcode files and both crashed auggie. tried it 4 times and crashed 4 times, I used the same gcode files in that last version with no problems. also was having major problem like before with jogging it seems to loose position when axis zeroed..

I also get a message after I run the jog script, then try to jog with the jog buttons I get the message wrong state for jog on axis.
If I hit estop and then estop on it works again..

Let me know where I have you confused, and I will try to explain that area better...

I thinking maybe we should just try to get the axis's to be in sink with the dros first then go from there. this seems to me to be messing things up somewhat.
mostly with scripts. the script gcode moves seem to be way off.. and feed rates still off..

you may want to pull this version from the web site just incase running gcode may be a problem..

Gary

[ArtF]

Gary:

  Ill be running a lot of code tomorrow on my laser so I wont pull it till
I have. Your pretty much the only one running Gcode driven scripts so
our tests shouldn't affect anyone else.

  I would stay away from simulator mode, theres too many things not
hooked into it, such as probing, all bets are off in sim mode if you
use a probe as it may well confuse it badly. So lets stay on actual
hardware till we know whats going on.

Ok, so lets take it a touch at a time so I can try to figure out whats up,
I cannot duplicate your hardware and Im simulating a probe with a button
so we need it to be simple tests.

>>Tested the probe in the MDI one line G31 Z-1 F10, second line G0 Z0 F30, started to probe with dro's set to 0 so it probed till touch plate hit and stopped, ran the second line and it retracted back to start position 0..

OK, so thats good, no problem there.

>> The probeinve rt keeps it setting now when rebooting .

  This too is good, that fix seems to work.

>>sometimes it would only probe down a real short distance and then retract not hitting probe

  This one bothers me, it shouldn't be possible unless some script code is causing the probe to be interrupted
somehow.  My probe always goes down till I hit the button, then stops. This is all controlled by the pokeys
BUT I can see if the sync if off before the probe call, it may go down a short amount as it doesnt really know
the correct starting position, so something prior to the probe would have to be screwing the sync. Perhaps
zeroing the Z axis maybe. Though I added code to try to ensure that cant happen.

This is all too scattershot though, I need you to make a script that fails that I can easily test and we know it
will fail. Something like... run this script  call , jog left once at .01 and then the probe in the script doesnt
go down. Lets get this down to one example, that I can easily test with set parameter. Its best if ,as you describe
you can show me how to make the probe not go down all the way, in that scenario I shouldnt even have
to press the button. Lets get it to display one problem so I can kill it and then move to the next ..and so
on, I find Im having trouble duplicating your issue because I dont follow your workflow.

  If I can make it fail, its usually easy to fix, the problem is Im fixing what looks like it may fail..instead of what
does fail. The probe invert I could see so it was easy to fix.

  So if Im following you, I can zero my DRO's, then use the probe button and it doesnt always go down
all the way? Is this while running code and after an M6 hit, or just the pressing the probe
button in standby?

Im thinking if th eprobe doesnt go down all the commanded distance the only possability
is the engine isnt synced, we just need to figure out whats unsyncing it..

Art
 

 

So can you give me a scenario and a script that fail?




 

[ArtF]

Gary:

  On reflection on this, Id like for you to try an experiment for me,
can you reproduce the problem without triggering the probe at all.
So put the device at a good height. Make a simple GCode program..
like

G0X0Y0
M6
G0X10Y10
G0x0y0
m30

  Then, in the m6 dialog run, test your buttons. Can you make the probe not go down far
enough..if so I can reproduce it by following your instructions. I should
be able to fix whatever we find, and then we can see if the probe hit
works after we make sure no-hit scenarios work. Im just trying to cut
down on the possible things that can be biting us.

Art

[gburk]

thanks art

Will test it on the machine later today.

As for the probe testing its always been tested with the probe on screen button and scripts, not with the M6 dialog so lets say the M6 call isn't in play endless I am running a gcode file..

I did reinstall the last version of auggie last night and had no problem running the same gcode file in simulator mode. but as you request I will try all testing now in machine mode..

Will get back to you later, I had a feeling you would want to throw in the towel after my last message.. :o

Gary

[ArtF]

Gary:
I think I found the major source of all this. Try the version I just uploaded.
I found if I zeroed the axis the next probe wouldnt work, this has been corrected so
now after a jog and zero the probe works fine.
  Let me know of any change.. Good luck

Art

[gburk]

Art

Will give the new version a try,

Think I have a few issues to work out first, I had a few minutes and went out to test, I am having homing problem now its will home but not backoff, I will check the pokeys config to see if any thing changed, but I do get the message axis homed, but I can't jog down, once it homed and I try to jog I get the message wrong state for jog on axis so i'm stuck and can't move the axis endless I hit estop and re hit estop..

I'm hoping it just because it not backing off.. I have auggie reverse set to true, or is this option just for the direction of homing and not back off?.
I did run mach4 a few days ago maybe it changed the pokeys config.. but I have homing set to backoff in mach4

Gary

[ArtF]

  Gary:

  Hmm, you may want to update just in case its that bad version doing it. I took out some changes and fixed
a couple bugs I could prove. I dont think it woudl affect your homing but you never know.

Art

[gburk]

Art

I installed the new version, and fixed homing it was a pokeys config that got changed and the logic was wrong and there were no backoff pulse's.. so now backs off, and that also fixed the jogging not working when homed looks like it has to back off and the home switch needs to go off, or jogging does not work.

here is the code I tested it with its in inch

G0X0Y0
M6
G0X2Y2
G0x0y0
m30

Ok first m6 dialog I just hit exit and continue button worked fine the x and y moved good and returned to 0.. no problems
Second run m6 dialog I hit the probe Z touch off button
Its setup to read the probe.panel dros.
It reads the Z Max Travel distance its set to 1, in the engine.gcode  I have the Z set to Z- it should it go down works for me
Also reads Slow and fast feed rates 10 is slow 30 is fast at least what I have them set at
and the Retract if probe not hit will be set to the Probe start position usably = 0  endless you start from a different position

That worked fine did, not hit probe but went to -1 and retracted to 0 and then x and y moved to 2 and back to 0..

I did not do any manual jogging for this test...

Is this what you were looking for..?

Gary

[ArtF]

Gary:

Yup, thats  what I needed. So all that worked well. Now try it with jogging after the probe, or before
and see if it fails or works..

Art

[gburk]

Art

Tested
1 - ran gcode then jogged while M6 dialog displayed, exited without probe and all when good returned to 0 then x2 y2 and back to 0

2 -ran gcode  then jogged with M6 dialog and then hit Z touch off button let it run in air and go to max distance, all went well then went to x0 y0 then x2 y2 and back to 0..

3- ran this test 5 times, ran the gcode at M6 dialog jogged z x y then hit the Z touch off button, and let the probe hit the touch plate, it stopped good and retracted to Z1 correctly it did not seem to mess up at all never seemed to get confused and dive into the part just hit and retract..

I did not use an probeinvert with g31 to backoff the probe I have a g1 or G0 I forget Z-RetractHight, Also I haven't tried the probing yet with the on screen buttons.

It looks good so far, any other test you want to try before I get ahead of us.. it may be a good day for testing cooled off a little at least for 24hours

Gary

[ArtF]

Gary:

No, Id say test at will now. I think the error I found was instrumental in all your problems.
Thats not to say other problems dont exist, but I like the way its looking so go ahead and play.
If you see a problem, try to isolate it down to a procedure just like you did with these ones,
if I can repeat it, I can find it generally, though Im amazed how complex it is in there..what
was I thinking..lol..

  Thats the problem with a program like Auggie, you soon start to forget exactly how you
did what you did. It has a whole lot of things going on at once and its enough to boggle
my head if I get too scattershot.. :)

  Good work on the testing, Ill have to add you to the credits for probe development.

Art

[gburk]

Art

Thanks that sounds good..

I did test it with the Z probe screen button and that worked fine also..

I did find a couple times that the after the probe was hit and retracted, I could not jog with the jog buttons got the message again wrong state for jog on axis
but its had to pin point it only does randomly, I assuming its like the homing if I didn't back off the switch I would not be able to jog.. so my best guess auggie for some reason misses the probe state when it releases but not to often though.

Also not sure if the Feed Rates are changing correctly, when it start the probe G31 F5 it seems to chance to 5 every time but when I do the retract to G01 Z1 F30
the feed rate seems to change to 15.75 not 30 and does this every time.. doesn't effect the probing just letting you know..

One question will auggie loose its home position on estoping , am I still good with the last homed mach position?. when I hit estop again..

Also I though you had said I needed a G0 before the probeinvert,1 is that still correct of can I replace the G0 retract with a G31 to check for a release on the probe..?.

Thanks gary

[ArtF]

Gary:

>> I assuming its like the homing if I didn't back off the switch I would not be able to jog..

  Shouldnt affect it really. Unless the home switch is also set as a limit, it will be ignored except in homing.
If it does share as a limit, then yeah, bad things can happen if it doesnt retract. There is a state DRO on the screen
on the right, you may want to take notice and see what it is when it refused to jog, let me know and Ill
see why its sticking in that state.

>>One question will auggie loose its home position on estoping

  No. Homing changes the machine coordinate and is the only thing that does. So Estops dont matter.

>>Also I though you had said I needed a G0 before the probeinve rt,1 is that still correct of can I replac

  No, all thats needed is a block on motionstill after the probe, you can then invert and probe off.

>> when it start the probe G31 F5 it seems to chance to 5 every time but when I do the retract to G01 Z1 F30
the feed rate seems to change to 15.75 not 30

Ill do some checks and see if I can repeat this, so far I cant get it to happen. Its a weird one..I cant imagine how
that feedrate number is changing.. is there a 15.75 anywhere in later code?


Sounds like its getting better, we're close. :)

Art

[gburk]

Art

Will let you know what I see, The feed rate may only be in the G01 from the script and not the G31..
and as far as I know there is no other feed rate changes going on.

May take a couple days having computer problems that's connected to the mills, haven't seems this yet, when I power the computer on all the drive lights power the power light comes on and sometimes I here the bios beep, then all power go off, tried booting 20 times or so and shuts off a second or 2 after power on.
if anyone has run into this let me know..

Thanks gary

[ArtF]

Gary:

Did over 10,000 holes drilled today with auggie so the base seems ok..

Sorry to hear about the cpu.. it sounds like maybe a device is screwed, Id unplug all the drives and
devices and see what it does..

Art