[cresusillo]

I am still very wet behind the ears concerning CNC so I'm hoping this is just something simple I don't know about yet.

I recently purchased a converted Sieg Mini Mill from Syil America labeled by them as the X2. The motors they install are Nema 23 stepppers and they drive ball screws. I have set up the machine with the bundled software they provided, including Mach3 and BobCAD/CAM. When opening Mach3 there is a profile provided for the X2 which loads correctly all relevant settings. This made it very simple to get the machine up and running but I ran into something puzzling when I tried to run my first program.

When jogging the machine the motors appear to be tuned well and the steps per for all axes is 25400. I wrote a very simple pocket milling program using BobCAD to test with and when I run it the machine does not run smoothly. The end result is still accurate and well finished but the nice smooth motor motion present while jogging is gone. I was not sure what post processor to choose when writing the program so I just randomly selected the Haas post processor in BobCAD and ran with that.

The options in BobCAD for post processors are:
Anilam 3000M
Fadal CNC88
Fanuc 6M(simplified)
Haas VF Series
HPGL

Is this where my error occurs? Is it a motor tuning issue? Is there even an error? Perhaps it is normal for the machine to take audible steps while machining? I apologize if this is a ridiculous question but I didn't find a solution in the training videos, manual, or previous forum posts. Perhaps I haven't looked closely enough. Thanks for reading.

-Justin

[turmite]

Justin it sounds like you have exact stop mode set instead of cv or constant velocity. There are +- to both but I have found I can live with cv much better than I can with es. Look in the configs page, I think that is where it is located. I might also suggest you go to the mach support site and view the videos there. A world help I assure you. http://www.machsupport.com/

Mike

[gmfoster]

If you are in exact mode (proper name escapes me at the moment) then yesit will slow down to near stop at every change of direction..


CV mode(constant velocity) will smooth things out but will round over corners to some extent..

Like everything in life its a trade off...

Garry

[cresusillo]

Mike, you were correct in assuming I was in exact stop mode. I changed to constant velocity but my problem persists. Garry, you mentioned that the exact stop mode will almost halt the table at direction changes but the jumpiness of the motors occurs on straight pathways as well. The faster the traverse the more aggressive the bounciness of the motors. i.e. when the z-axis sets up for its initial height (slow traverse) there are audible, tame bounces and when the x and y-axes are moving to the starting position (faster traverse) there are audible, aggressive bounces. I will continue to investigate the problem through the videos, manuals, and trial and error. Thank you both for your responses to my post.

[gmfoster]

Mike, you were correct in assuming I was in exact stop mode. I changed to constant velocity but my problem persists. Garry, you mentioned that the exact stop mode will almost halt the table at direction changes but the jumpiness of the motors occurs on straight pathways as well. The faster the traverse the more aggressive the bounciness of the motors. i.e. when the z-axis sets up for its initial height (slow traverse) there are audible, tame bounces and when the x and y-axes are moving to the starting position (faster traverse) there are audible, aggressive bounces. I will continue to investigate the problem through the videos, manuals, and trial and error. Thank you both for your responses to my post.

Well in that case you have something running in background steeling time from Mach. Post a picture of the line driver test drew for you.

Garry

[turmite]

What Garry is talking about is another program stealing sources that Mach needs. I know one of the things that will zap it is screen savers.

Mike

[cresusillo]

here are the driver test results. is this bad? should I try and optimize my system? the computer working the machine is a dedicated tower that has no other purpose than to run the CNC so whatever is needed to get this working is ok. any advice?

[cresusillo]

Mike, I just turned off the screen saver with no apparent improvement... must be some other processing power thief. Thanks for the advice.

Justin

[Cruiser]

cresusillo, You didn't say what motor speeds your set to or what pc your running. One other possiblility that i have run into is the processor capacity of the pc ! Jogging takes up little proc, in program mode you can create a lot of math for the proc to handle and if the math exceeds the ability of the proc and pipeline etc. then you will have a very nervous system. It could well be much simpler such as the motors tuned too tight or whatever .... just a thought here. I'm waiting impatiently for my package to show up today so that i may upgrade my pc and prove my point. I knew my pc was marginal when i started this then forgot till lately, Good luck with your system and enjoy the Journey !

[cresusillo]

cruiser, my motor velocities for x, y, and z are all 60in/min, the accel. for x and y are 6in/s^2 and for z is 0.25in/s^2.

the computer is pretty beefy, Intel Core 2 Duo each one running at 1.86GHz, 2GB DDR2 RAM, and a SATA 320GB HDD. Even with a couple of things running in the background it seemed to me that Mach3 should be well taken care of. But these new computers come with about 72 processes running on them standard. I went into msconfig and disabled almost all of the startup processes but that didn't work either. I am prepared to follow the instructions of the optimization file that is posted on the ArtSoft website but it is advised in that same document not to do it unless sure of the problem. So I will bide my time a bit longer and look for other solutions.

Thank you for you looking and helping and I hope your problems are solved with your upgraded pc!

[cresusillo]

I thought that my jumpiness to the motors was due to such a horrible driver test result as seen above, but when I changed computer to 'minimal power settings' it almost completely flattened out my driver test results. Being very encouraged by this I tried to run a simple 'g0 x1 y1' command but got the same jerky motion to the point. It sounds like a helicopter. Very frustrating.... if anyone has had similar problems I would love some advice.

[Cruiser]

I see you have steppers, i have servos, i doubt you have geckos for drivers but some clone, i'd look at doc's to see ifn how to tune the motors seeing if they are tuned too tight, per say on my geckos if gain and damp are set to high side then they are tight and jerky also they hum at idle. if i calm the adj down to more neutral then at idle they are silent and when driven they aren't jerky but much smoother. Most likely your best bet is to contact syl america or whoever they are and see what they got for you, as i am truly guessing !

[Torchhead]

I'm voting for the slow processor. A 1.8 I bought and used several times on MACH3 running WIN2000 totally crapped out (jerked like a dog crapping peach seeds) on any commanded moves under XP. I worked with it for several hours and tried every trick. I switched to a 2.4G and it runs smooooooth. The 1.8 still runs fine under WIN2000.

Shared video ram can be a problem with some boxes. If it has on-board video you may have shared ram and processor cycles.

Try turning off the toolpath display in MACH and see if it calms down. If it does buy yourself an external VGA or AGP card and tuen off the on-board video.

tomCAUDLE
www.CandCNC.com
BOB's
Digital THCs
DXFTool Software
Complete Electronic Packages
Stepper & Servo Systems
Hand Controllers

[Mike Nash]

You could always try TurboCNC to rule out your drives, motors, mechanicals.

http://www.dakeng.com/turbocnc.html

You'll at least need DOS on a floppy to run it.

[cresusillo]

Thanks for the advice everyone. I will see what I can do with it today. I feel like I should also mention that the parallel port I am running through is actually on a card in a PCI slot. The computer we bought didn't have a motherboard with a parallel port already on it. I'm not sure if that will affect the performance or not.

I'll keep posting as I work today and let everyone know how much progress I make. Thanks again.

-Justin

[cresusillo]

Ok, thanks all for the time and advice. I have been playing around with the software and machine for a bit and listening to this forum and advice from others as well. The problem is solved!! Turns out that the problem with the g-code jumpiness is bug in the Mach3 software. This is probably why all my tweaking of the variables was not helping. I downloaded the Mach3 Version 2.00.076 file and after loading it my g-code jumpiness was fixed!! Again, thanks to everyone for helping.

-Justin