Looking for help with TRAK-DPM cnc resolution

Cullen

Registered
Registered
Joined
Dec 21, 2012
Messages
14
Hi, I'm new to this forum and just posted a picture of my TRAK-DPM on the "Other Mill Machines/show us your mill" forum thread. I have had it just over a year and have got it cosmetically cleaned up and spit-polished. At first, It had a lot of shut-down errors when trying to run simple g-code programs. I spent a lot of time trying to figure out if it was the code or the program interpetation. I am also new to the cnc programming world, but have had a lt of experience with cad artwork generation. It machine finally just quit with a "Bad Disk" error (it runs off a floppy disk OS that gets loaded at startup). I downloaded a new copy of the OS and loaded the new floppy into the hidden floppy slot of the machine. All of the previous error shutdowns went away on the same program files. I thought that I had fixed it, but after running a few simple profiles and pockets, I found that the dimensions were just slightly off (0.030" on a 1.000" pocket length). The Z axis seemed to be the worst when trying to make thin holding tabs for cut-outs. While watching the DRO numbers it would never quite go deep enough (although the program was instructing it to go to a final depth). Then I went through a calibration routine in the OS (using gage blocks and indicator). The OS only allows for small calibration corrections (maybe 1-2%). The problem still exists with exact dimensions. I ran all the self test programs for backlash and slew rates and everything is in the nominal values. I checked tool wobble and everything that I thought would result in over-cut. The machine uses (I believe) servos and scales. Does anyone have any suggestions? I would call a factory consultant (Southwest Industries), but I have heard that they are quite expensive for service calls and I am on a limited budget after buying the machine. I'm still trying to do as much as I can before saying "uncle". Any help would be greatly appreciated.
Cheers, Cullen
 
here is another page that may help to look up your model and get info
http://www.southwesternindustries.com/index.php/model_resources
and here is a pdf that has troubleshooting by symptom ( i hope its the right machine looks the same ) it has a section about part not being sized to program dimensions and backlash checking and setting compensation
steve
 

Attachments

  • 24941_manual.pdf
    6.2 MB · Views: 5
Last edited:
here is another page that may help to look up your model and get info
http://www.southwesternindustries.com/index.php/model_resources
and here is a pdf that has troubleshooting by symptom ( i hope its the right machine looks the same ) it has a section about part not being sized to program dimensions and backlash checking and setting compensation
steve
Steve, Thanks for the interest and research. I got this manual with the machine and used it to perform the backlash and calibration diagnostics. As I mentioned before, all the numbers were right on the nominal values listed. I guess that I am going to just sit down and go through the trouble shooting list one line at a time from start to finish. I am not getting even close to the positional and repeatability numbers listed (0.0005"). I would be extremely happy to be within .001-.002" since I am not in a controlled temperature environment. One paragraph did mention writing down the calibration numbers from a file, turning it off and back on and then re-checking the values to see if they are changing. If they are, it's like throwing dice.
Thanks again...
 
here is another page that may help to look up your model and get info
http://www.southwesternindustries.com/index.php/model_resources
and here is a pdf that has troubleshooting by symptom ( i hope its the right machine looks the same ) it has a section about part not being sized to program dimensions and backlash checking and setting compensation
steve
Steve, I have a correction to make. This is not the same manual as the one that came with the machine. They are similar but definitely not the same. This one has more things to check. It is probably a later version, and definitely more useful.
I got out in the shop this afternoon (the high today was 31 degrees) and went through the position and repeatability tests on all three axis, (zero each axis on a reference using an indicator in the quill, pull away 2-3" and go back to the indicator reference). They repeat perfectly back to 0.0000 for all three axis. I did the calibration on X,Y axis with a 3" gage block and the Z with a 1" gage block. I then put a piece of clear acrylic in the vice and did a manual rectangular milling operation 0.080" deep from the surface with a 0.125" 4 flute end-mill at 1600 rpm. The coordinates for the four corners of the rectangle were (0.0, 0.0), (1.250, 0.0), (1.250. 0.250), (0.0, 0.250) and back to (0.0, 0.0). The outside edge of the rectangle measured 1.2505 in X and 0.2510 in Y. The Z, however, was -0.1060" with a -0.080" plunge of the quill. I think now I have a starting point to isolate the problem. Since the repeatability is OK (even on the Z axis), I would think that there is something amiss with the counting or the distance per count on the Z axis. I only moved the quill and not the bed in this test. Under normal operation however, the quill is stationary and the bed moves. Thanks, there is progress....
Cheers, Cullen
 
here is another page that may help to look up your model and get info
http://www.southwesternindustries.com/index.php/model_resources
and here is a pdf that has troubleshooting by symptom ( i hope its the right machine looks the same ) it has a section about part not being sized to program dimensions and backlash checking and setting compensation
steve

Steve,
I printed out the document and read through every page. It is definitely not the same as the one that came with the machine (a later revision probably). Anyway, it was snowing heavy here yesterday so I decided to spend the afternoon going through a calibration procedure (start to finish). It took quite a while, but, was time well spent as evidenced by a couple of trivial test pockets (one before the calibration and one after, using the same G-code). THANKS!!! Apparently, (in retrospect), the system OS floppy disk was going bad gradually. It was giving erroneous shut-down errors due to its own problems, not the G-code. When it finally quit with a "Bad Disk" error (actually a blessing) a new disk build did not have the "calibration data" file. Going through that cal process (with your document sure helped) and saving the file to disk, has created a working system. Now I can get serious about making chips.
Cheers, Cullen
 
thats good news
sorry i havent replied sooner. I have not been getting all my notices in my email
post some projects when you get going
steve
 
Back
Top