Machsupport Forum

Third party software and hardware support forums. => LazyTurn => Topic started by: Servolock on October 31, 2009, 08:54:42 PM

Title: Is this a bug in LazyTurn?
Post by: Servolock on October 31, 2009, 08:54:42 PM
Hi Rich et al.,

So, I sprung for the $175 and purchased the license to Mach3.  Solved the previous G-code garble problem.

I'm cutting a profile into wood.  I'm using three separate runs:
  1) cut the billet down from square to round,
  2) cut the desired profile into the billet,
  3) make a deep, parting cut.

Now, I used the same AutoCAD file for both run 2 and run 3.  I just made the parting cut deeper for run 3 and moved some surfaces up to the billet diameter so LazyTurn wouldn't waste time cutting the same areas again.  I didn't move any the location of the parting cut.  I just made it deeper for run 3.

I was very careful not to shift ANYTHING in AutoCAD.  Indeed, I can load the two runs alternately in AutoCAD and they perfectly overlay each other.  I also checked the positions of the vertical lines defining the shallow cut and the deep cut and their X values were identical.  Only the length of the lines, i.e. the Y ends, were different, to set the depth of the cut.

When I fire up run 3, it wants to make the parting cut about 2 mm farther out, i.e., about 2mm offset on the Z-axis, and out of position from where the shallow version of the cut was made in run 2.

Am I missing something here?  It looks like a bug in the program to me, but I'm hoping you'll have an explanation that I overlooked.

By the way, everything else is working well.  I'm very pleased with myself because the hardware I scrounged and assembled is performing just as I hoped.  Not a hint of chatter. Plenty of torque in the step motors.  I can re-run phase 2 over and over and it overlays the previous cuts perfectly.  Fun and amazing to watch.

I just can't figure out why LazyTurn is shifting the location of the parting cut.

Any ideas?

Regards,
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on October 31, 2009, 10:21:08 PM
Dave,
Been a while since i fooled with LT. Post the DXF. Can you post some screen shots?
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on October 31, 2009, 10:45:51 PM
Rich,

Here are the two .tap files and the corresponding DXFs.

If you'll tell me what screens you want and what program you like for capturing screens, I'll send them.  (I've literally never captured screenshots before.  Just never had to.)

For some reason, I thought you wrote the LazyTurn manual.  Thanks for offering to help.

Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 01, 2009, 09:21:34 AM
Dave,
I did write the manual from 400 pages of information and LT is still in development. The manual is not complete and will also change when we get into the finishing pass ( hopefully this fall / very soon ). The manual reflects were we left off in the developement.

So if you ask is this a bug? Maybe.....  during the testing i would would easily try something out against 20 files and see if i could duplicate it. Transparent to others, 2 to 4 hours were spent and who knows what other folks would put into it.

So will need to play around some.

There are free screen capture programs, if you could try a search and you should find a posting which gives you a link,
or maybe someone will post one here for you.

The screens required would be the the tool and it's parameters, and a shot of the generated pathing.

Sorry no imeadiate answer / bear with me, i'll take a look in the next few days.

RICH


 
Title: Re: Is this a bug in LazyTurn?
Post by: budman68 on November 01, 2009, 10:35:29 AM
Dave, you don't need a seperate screen capture program. When you have the screen showing on your monitor that you want to capture, just click the "print screen" button on your keyboard. That sends the image to the clipboard and then all you have to do is open up microsoft paint (comes with Windows) and click on Edit then paste and the image should appear. Then you can crop it (or not) and save it as a JPEG file but make sure you name it something odd like "Servolock1234" or it will not upload on this forum as this forum does not allow 2 files of the same name to be uploaded.

Then when you reply to a post, just click on the "Additional Options" link and it will slide open and then you can upload your screenshot using the "Browse" button.

Hope this helps.
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on November 01, 2009, 04:43:34 PM
There's another source of confusion that might relate to my problems.  I can't figure how to put my profiles in AutoCAD's upper left quadrant. 

Apparently it's not enough to simply set the set the UCS origin at the bottom right corner of the profile drawing.  I found out by experimentation that I can have the same profile in two different drawings, both with the UCS origin set to the same place on the drawing, but LT will place the profiles at different places on the billet. 

I started with a drawing that LT would accept, but I didn't like it because the profile was too far away from the chuck.  To reposition the profile closer to the chuck, I had to move the drawing over in AutoCAD.  Then I set the UCS origin to put 0,0 where it belonged again.  It's as though AutoCAD and /or LT have a hidden "machine coordinate system" that I don't know where to find.  The UCS is all relative and LT is ignoring it and using the hidden, absolute coordinates instead of the relative coordinates of the UCS.  Maybe everybody that uses AutoCAD knows about this but me.  Could easily be true; I've never taken a class.   So at this moment, I don't know how to place my drawings in the upper left quadrant because LT doesn't seem to use the UCS origin to define quadrants.

In case it matters, I'm using AutoCAD LT97. 

Thanks for any help.

Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 01, 2009, 05:13:09 PM
Select all on the drawing, then,  do a right click with the mouse and you will get the flyout menu for command selection, follow the "MOVE" commands prompts, you snap to the point on the drawing and move it to coardinate by typing in "0,0,0".
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on November 02, 2009, 12:19:50 AM
Rich,

I think what you just described is equivalent to what I normally do, which is: "Tools,"  "UCS," "Origin," then I snap-to whatever point I want to be 0,0 and left-click it.  That makes the point 0,0 in the UCS, but somehow that's not what LT wants.  Because if I move the profile a little left or right and then fix the origin like I just described, to keep the profile at 0,0 in every case, LT will still move the profile on the on the Z-axis.  As though there's an underlying coordinate system that AutoCAD and LT are using, and LT is ignoring the UCS.

By trial and error I can figure out where LT thinks its origin is.  I just wish I knew what's going on.

Thanks,
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: budman68 on November 02, 2009, 05:03:41 AM
Dave,

I'm at work this morning so I can't really send you any screen shots to help visually but there's yellow a ruler/scale for each the Z and the X in L- Turn.

Are you adjusting those to your liking?

Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 02, 2009, 08:50:17 AM
Open the OPTIONS MENU>SET PATH OPTIONS and check that there is no Z offset.
(section 5.5 / page 8 in the manual).
Also as Bud posted, see section 6.2.1 page 23 of the manual. But the Z start point should default to 0 idf that is what you have
in the OPTIONS.
Also make sure you installed  the latest version of LT. I believeit can be found in reply #1067 page 107 of he LazyTrun thread.
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on November 02, 2009, 11:45:02 AM
Rich / Bud,

I double checked.  Yes, I'm using the May 28 LT.

The Z-offset is at 0.1.  I don't remember if I put that in, but if I did it was a long time ago.  I haven't been doing anything to it for the last two weeks at least.  I haven't touched either the Z-offset pointer or stock location pointer on the ruler.  That's why it's a mystery to me, that the profile moves on the stock even though AutoCAD always shows the profile at the UCS origin.  This is not a showstopper.  I can work around it, now that I'm aware of it

My showstopper is:  I apparently have some hardware problems after all.  As I am cutting the profile, Z-axis stepper acts like it is briefly sticking or something, i.e., getting slightly offset.  I don't think the problem is mechanical.  I cut the Z-axis speed and acceleration in half and I'm still seeing the same problem.  I am suspicious of my Dell laptop.  I'm going to switch to a different platform, running a fresh install of Win2K, my favorite Windows version.

I'll be back after I get a handle on my hardware problems.

I must say, it is gratifying and reassuring to get the level of attention you guys are giving.  I saw this forum and this process before I chose Mach3, and this was a major factor in my decision to go with Mach3 instead of the alternatives.

Thanks!
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: budman68 on November 02, 2009, 11:49:35 AM
Let us know how it goes, Dave, as the Dell laptops can definitely be "iffy" on whether or nor they work.

There's a thread floating around on this forum about Dells and I have a D600 and it works fine but I think the next model up (D610?) has serious voltage variances and creates havoc.

Good luck-
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 02, 2009, 03:15:44 PM
My Dell latitude can only be used with a SmoothStepper and the same goes for the Dell Optiplex.
I think Dell's are just a poor choice with on board graphics. But That's just my opinion
The best thing i ever did was to have a PC put together.
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: budman68 on November 02, 2009, 03:48:07 PM
Hi Rich, with the Dell OptiPlex, it really seems to depend on which form factor as well as some of those samller ones are basically laptops in a tiny size tower form.

I have a Dell OptiPlex GX260 which is a full size tower and workhorse  8)

Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 02, 2009, 04:03:50 PM
Hey Dave,
The Optiplex is great for everything and they were used at work for doing high end 3d design.
But for running my steppers via PP, it just plain sucks and never was able to get a good signal out of it no matter what was done or tried. Steppers sounded just like they were being hit with a mallet and limited to 8 IMP.  The SS was the only cure. What can i say!
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on November 04, 2009, 03:04:18 PM
Rich / Dave

I found my hardware problem.  It isn't the Dell laptop.  It's worse.  Apparently, I'm fighting a resonance in my Z-axis actuator.

I slowed the Z-axis *waaay* down and the problem went away.  Now I have to figure out how to identify the resonance and how to deal with it.  This is an area I have no experience in.  Can you point me to a recipe or tutorial on how to measure and compensate for resonance?  I'm hoping I don't have to do this by trial and error because that could take forever. 

Now, there's a Lovejoy-style coupler between the step motor and leadscrew.  I'm suspicious of the rubber spider between the two halves.  Seems to me that this could invite resonance problems.  Yes? No?  The leadscrew is almost 30" long and at least 1/2" dia., pretty massive.  And the stepper shaft is only 1/4" and is probably twisting under the torque and contributing also.  Would a larger stepper with a more robust shaft help?

Should I post this on a different part of the forum?

regards,
Dave   
Title: Re: Is this a bug in LazyTurn?
Post by: budman68 on November 04, 2009, 03:23:30 PM
Hi Dave,

yes, post this in the "General" area of the forum and you'll get more views that way. Somebody else would be better at helping you with this as this is not my area of expertise. As a matter of fact, I have NO area of expertise,  :D

Just a thought, have you tried switching motors and or couplings with another axis to try and narrow the problem down a bit?

Good luck,
Dave
Title: Re: Is this a bug in LazyTurn?
Post by: RICH on November 04, 2009, 05:21:07 PM
Dave,
Do re-post the problem as Budman suggests. Povide a description of the problem as best you can.
Give info on the ipm / velocity / accel your running. A picture of the axis which shows support,  how the motor is mounted,and a general shot of the machine  is worth a thousand words.
There is a difference between a motor resonance and mechanical resonance of a shaft and what occurs.
That should provide folks with basics to assist you.
RICH
Title: Re: Is this a bug in LazyTurn?
Post by: Servolock on November 04, 2009, 05:32:00 PM
Dave / Rich,

Wilco. I'll post in "General."

I do have two more identical motors.  I just have to put on my connectors in order to use them.  I don't expect it will make a difference though, because this is a small step motor driving a large leadscrew and the original design didn't require high performance.  If I want performance I'll probably have to spend money.

Dave