Touch-off plate for CNC router?

[acondit]

I don't know what you zipped but all of the files appear to be zero length.

Alan

[chazmtu]

Thanks! Here are the new configuration files

[Big John T]

chazmtu, I didn't get a chance to download the file this morning... I have it now and will look at it this evening when I get home unless someone else spots it first...

John

[chester88]

After a quick look I see that Classicladder will not (from the comand line) load a filename with a space in it. Your ladder program is called 'touch_off Button' rename it 'touch_off_button' and the ladder program will load and then you will see the ladder program when you bring up the GUI from AXIS. after that the pyvcp button did register in classicladder but I don't have any hardware hooked up so not sure if the probe comand was given. I will keep working on it.
P.S. using filenames with spaces is never a good idea when using Linux.

[Big John T]

YEA! chester88 to the rescue...

John

[chester88]

Ok I did a little more testing and the commands are not getting called because Classicladder doesn't hold the outputs on long enough. We need to add a MONOSTABLE object for a couple of the output lines. I set them at 500ms. I uploaded a new ladder drawing here:
http://wiki.linuxcnc.org/cgi-bin/emc...LadderExamples
We are real close now- axis shows my tool probing down looking for the probe signal.Of course I don't have one so I don't know if the rest of the program works.
I suggest that you test this by touching the ground wire on the tool instead of letting the tool touch the plate-just to make sure if it doesn't stop the tool won't be broken.
When you know it works then test with the plate as usual. I'm sure you are doing something like this I just want it recorded so other people read the warning

Cheers Chris

[chazmtu]

Awesome!! Thanks, I'll give it a try right now.

[Big John T]

chazmtu,

I took a look at your confg and noticed that you have multiple circuts in one rung. I don't think that is good form and am not sure how classicladder likes it. Most PLC's won't accept that. http://filebin.ca/gadzzw/zone.zip is the files I modified to make it do two steps so you can see how it can be done. The second step is based on a timer as I could not find anything to trigger the completion.

The thing you need to remember when doing ladder is how a PLC works...
1 Read Inputs
2 Solve Logic
3 Updates Outputs

and the golden rule of PLC's "the last output wins" which means you can only have one coil per address with the exception of set and reset.

John

[chazmtu]

I see. I didn't know you can't have more than one circuit per rung. Some classic ladder examples in the wiki have multiple circuits and coils, that was the example file there.
I can't get these files to run either though. HAL meter shows %I0 true when my button is pressed. The MDI mode switch was true also but no output on the first coil. Chester88 mentioned the values not being held high long enough. He added a timer to the example he posted here. I'll give it a try, might take a while though, my machine has trouble running the ladder editor.

[chester88]

You could run Classicladder without EMC to rewrite the ladder program- It would use less computer power.
in a terminal type 'halrun'
then type 'loadrt classicladder_rt'
then 'loadusr -w classicladder'
then use classicladder's load button to load the program to modify
and then save it after you modified it.
You can't run the program this way cause you haven't hooked up signals or added a thread but you can edit the ladder program.
Or I can send you the file.

[chazmtu]

If you don't mind sending me a copy. I'm getting tired of this fossil. I think its time to start looking for some better hardware. Other than that classic ladder file I should be all set.

[chester88]

ok try this:
http://filebin.ca/ngshze/Touch_OFF.clp
rename it to your ladder program name.

[Big John T]

I see. I didn't know you can't have more than one circuit per rung. Some classic ladder examples in the wiki have multiple circuits and coils, that was the example file there.
I can't get these files to run either though. HAL meter shows %I0 true when my button is pressed. The MDI mode switch was true also but no output on the first coil. Chester88 mentioned the values not being held high long enough. He added a timer to the example he posted here. I'll give it a try, might take a while though, my machine has trouble running the ladder editor.

I see that the examples have more than one circuit, I guess ClassicLadder does not care...

The reason your computer is having trouble running the ladder editor is your base period is too small. The computer spends all it's time cycling the base period and has no time left for other things. In the example you sent me I had to change the base period so I could run it on my computer. Create a new directory and put all the files I sent you in there plus any I didn't send from your original directory and up the base period until your computer quits choking and run it again...

If you want to do it through the stepconf wizard then up the latency test number then save it and test. Try a number that will give you 30000 to start.

John

[chazmtu]

Thanks for the files and for mentioning my base period, I forgot all about it. I knew it need some serious tweaking but It got pushed aside when we started this. Hopefully that will get me a little performance boost. I'll let you know pretty soon what I come up with. I think I'm starting to get a feel for how the config files work now.

[chazmtu]

OK, here's where we stand. The first two mdi commands run flawlessly every time. The third ran once out of 10-15 tries. Since its the only output without a monostable output I'm gonna add one after the timer and I think we should be all set.

As for my base period, I took your advice and set it up to 30000 and It does run better but classic ladder still too slow to use with axis running but the manual method is tolerable. I'll have to do some testing for that later.

I also bought a laser crosshair module off e-bay for zeroing the x and y axis. Luckily that button will be easy to set up. After that I think I owe you guys a youtube video of our work in action for all your excellent help. Looks like I might be done before the day is through.

[Big John T]

chazmtu, as they say down in Arkansas... SWEET!

John

[chester88]

Yes of course the last output needs a monostable - LOL - sorry I should have caught that! Glad to hear this works. I am wondering how accurate/repeatable the touch off is- I imagine for routing +- .005 would be fine.
John Thanks for mentioning the one circuit one rung deal. I have come across this problem on another project with out knowing what the trouble was!
Chazmtu congrats you are the first person to have this feature on a machine!
I am glad you kept at it and got it to work - I will add it to the wiki and I would bet John will add it to the manual.
Will love to see the video.
Chris

[chazmtu]

IT WORKS!!! I attached a copy of my config files. USE AT YOUR OWN RISK! IT HAS A FEW BUGS. Some of the statements below are for anyone who wanders in later on. Any idea's on how to get rid of the bugs? They're a little strange and potentially destructive. I was thinking the in_pos pin changing states from jogging might start the center of the program on the third circuit down, though it shouldn't with %B0 being open.

Known Bugs:

1. If I jog the Z axis up and switch to MDI mode, it automatically runs the program. This was repeatable in that session but when I restarted EMC , it stopped. I haven't been able to repeat it since.
2. If I jog the Z axis up after pressing the button in MDI mode, it either runs one or several very short probe moves and returns the error 'G38.2 Probe move finished without tripping probe' (probe move is mdi command 1) for each move, or it will run the last two commands which set my Z height to 0.25in then rapid up 0.75in. The choice appears to be random.

Tips:

• The button won't work if your not in MDI mode.
• Don't use large rapid values for the third MDI command because of the bugs mentioned above.
• Use small distances for the probe move. If the probe extends beyond your machines soft limits the button won't work.
• NEVER jog the Z axis after pressing the button. You don't want to damage your machine.
• See the HALUI section of my-mill.ini for the code ran by the button.

[chazmtu]

Thanks very much for the help! I learned way more from this than I would have thought. I didn't expect it to take so long. I don't suppose you would have any problem with me linking my site the wiki? I might keep working on this.

I don't really have a way to test the repeatability, though +/-.005 is probably better that my machine can get right now. Make sure to add a warning to the wiki, someone could probably do some real damage it they're not careful.

Chuck

[chester88]

Hmm the program should never run unless you are in MDI and you have pushed the pyvcp button. Can you put a Halmeter on the pyvcp signal and see if it is ever on when it shouldn't be on? Feel free to link the sites
Also when you say you jog z after pressing the button do you mean that after pressing the button and letting it finish probing if then you jog (by switching to manual or by using a MDI command) it starts to probe again?