Messages

311

General Mach Discussion / Re: Abnormal Condition warning

« on: October 17, 2008, 11:58:54 PM »

Thanks for all that insight. Now that I understand it I guess it makes some (very limited) sense why it's activated in my particular program when nothing else has ever done it.(in my limited experience).
So now I'll ignore it.

I am now beginning to be suspicious about my installation of Mach on this test laptop. The code presented sometimes loads and comes up with an error about "the start of an arc is not consistent with the end point" (something like that). Then the program sits on the bad line of code and nothing works. The only way to get rid of the error seems to be to clear the code and shut down Mach and restart it and then the code usually loads and runs fine.
 This and the other "abnormal condition errors" seem to come and go. Some other strange things as well. Only on this program.
I noticed tonight that the code loads and runs fine on my actual CNC setup. No flashing lights or hangups.
Has anyone tried my program to see if it triggers any of these problems?


Sage

312

General Mach Discussion / Re: Height Following

« on: October 17, 2008, 02:47:45 PM »

I'd be sceptical that an ultrasonic system would work, especially in the vicinity of the weldng/cutting process. Welding, cutting etc apparently produces a lot of sonic (and probably ultrasonic harmonics) "noise". Evidenced by the fact that hearing protection is recommended / required for welders. Never would have thought so, but apparently true.

I guess it really doesn't matter what you decide to use as long as you standardize on the voltage levels the software expects to receive. Output level shifting would be simple enough to implement for whatever sensor you choose.

Sage

313

General Mach Discussion / Re: Abnormal Condition warning

« on: October 17, 2008, 01:31:14 PM »

Thanks guys. That's good insight as to what Mach is "thinking".

After trouble shooting step by step (inserting m1's here and there) I found a couple of places causing different "abnormal condition" warnings. (I'm a rookey at Gcode so go easy on me).
 The one about CV seems to come and go. The last time I ran the program it was being generated while the code was drilling. Maybe the look ahead nature of the program execution is confusing the real line that is causing the problem but there are no arcs being cut until  quite a while down in the code from where the light comes on.

The other one seems to be caused by a bit of code something like this. (I think it's acceptable - no?)
G91
G0 Z-.020
G90

The light turns on for the G91 line and goes out on the G90 line. Under normal feed rates it just flashes briefly going un-noticed mostly.
The error reported in this case is "G90 / G91 State".

I'll attached the code. I'd appreciate your comments.

You'll probably want to change all the feed rates to 10 instead of 1 unless you like watching paint dry.
The CV error is visible near the beginning of the program during BOTH of the drilling sequences.
I left some M1's in the code to stop during the "G91 / G90" problem in the outline milling subroutine at the bottom.


Sage

314

General Mach Discussion / Abnormal Condition warning

« on: October 17, 2008, 09:13:24 AM »

I generated some G-code which causes the "Abnormal Condition" warning light to flash on the Mach control panel. If I clck on the warning light it lists the problem as "Exact Stop vs CV mode".
The code runs ok.
I  created an offset in lazy cam on something that had square corners. It insisted on rounding the corners so instead of fighting with it I accepted it's code any way and then fixed it up myelf by hand to remove the arcs it generated and replaced then with straight moves. Not sure if that led up to this problem.

Is this warning something I need to be concerned with.

Sage

315

LazyCam (Beta) / Another small bug to look at

« on: October 12, 2008, 11:51:51 AM »

If anyone ever does get around to fixing up Lazy Cam. In the Text generating tab. The pull down font selector works strangely. Using the down arrow works normally (list moves up and alphebetically increasing at the bottom) but clicking in the lower part of the scroll window the list moves down while alphebetically the list gets higher at the top. Very strange. I guess the list is a virtual thing because in effect the same font name is at the top and the bottom of the list. You have to try it to see what I'm getting at.

Sage

316

LazyCam (Beta) / Re: Help with reversed order entities

« on: October 10, 2008, 09:37:34 AM »

Rich:

I see you patiently waiting.

You had a question


Sage

317

LazyCam (Beta) / Re: Help with reversed order entities

« on: October 10, 2008, 09:36:21 AM »

It's interesting that the disconnects show up in lazyCam because they are not showing in AutoCad. I am aware of this problem and corrected some disconnects using the extend line and trim line functions in autocad and they look clean. At least for all reasonable zooms in Autocad. So perhaps there is something amis in the mathematics of the DXF conversion that LazyCam has a difference of opinion on. Apparently this is an issue else LazyCam wouldn't have the conection resolution variable.

 Presumably clean and optimize are supposed to remove these problems from the file and it appears to do so, evidenced by the fact that after cleaning it is translated properly as 4 arcs for the whole cam. It's only when you offset it with lazyCam that the extra entities (presumably due to disconnects) show up again. As mentioned - if I do the offset in AutoCad before sending it to  LazyCam it imports into LazyCam properly as 4 arcs.
BTW after offsetting and moving the rapid point, I've seen extra entities added right in the middle of what was a perfectly good arc. So something isn't right.

Still. I agree with you. It is a powerful tool that I can't do without. It's all a matter of understanding the limitations and all of the work arounds and this thread will help in that regard. At least LazyCam generates the geometry - something my math limited brain appreciates. The rest of the code can be cleaned up and optimized by hand.

Sage

318

LazyCam (Beta) / Re: Help with reversed order entities

« on: October 09, 2008, 09:07:13 PM »

No 8 isn't a problem. It's just a curosity why it needs to create them. If I use Autocad to create an offset in the origial dxf file it still imports into LazyCam as only 4 arcs.
 The extra entities are very small but I've had other files with them in and they create a very small pause and leave a small mark on the part. Luckily in this case they don't (not sure why) but it's a good thing because I don't really want bumps in the cam.

Sage

319

LazyCam (Beta) / Re: Help with reversed order entities

« on: October 09, 2008, 07:18:44 PM »

Ok. It looks like V3.00.2 from the Mach website corrects this "out of order entities" problem. Still not sure if this is the latest since I thought the one posted by Ron was supposed to be newer but I trust that the website should have the latest and greatest. I'll see what happens. I can't recall what the posted one was supposed to fix.

SO another question:

How do you get rid of all the tiny entities created when you do an offset.

Docltf:
You will notice your file has a bunch of extities not in the original dxf (originally only 4 simple arcs in the file).


Sage

320

LazyCam (Beta) / Re: Help with reversed order entities

« on: October 09, 2008, 06:23:02 PM »

Docltf:
Yes, Your file seems to be ok. Why is the question? What did you do different from the process I described in the first post. MY connection tolerance is fine (also .001). This is evidenced by the original import of the DXF file which is clean and has only 4 simple arcs the way it was drawn in Acad. It's only when you move the rapid point that it screws up.
Did you try moving the rapid line and start point??

Budman68: You are correct there is something fishy connected with zooming in. When I zoomed in on the point at the end of the rapid as I moved it the entities did remain in the correct order A COUPLE OF TIMES. Then I guess I didn't zoom in the magic "enough" and moved the point and the entities re-ordered themselves and were then screwed.
THIS IS CRAP !!

Docltf:
I'm using what is reported in help\about as version 2.61 but I'm pretty sure I downloaded the version posted by Ron Gingery in a message here recently (January??). I also understand the version numbers are all screwed up so who know what I'm running.

Maybe you can point me to the version you have. As I understand it the latest version on the MACH website is not the newest one. The one posted here is apparently newer.

It's all very confusing an pretty hoakey. If it wasn't for the fact that it sort-of-works and I have no immediate alternative I'd ask for my license fee back.

Having said that. I'll will stick with it because I have faith in the author. When we finally get to the bottom of this it will be another chapter in the manual.

Sage