Any ideas on what to do about lost steps?

[Brian Harvey]

I've had my xylotex 4 axis kit with the 425 oz motors for over a year now; however, I've not had the money to purchase all of the software until now.  Up to this point I've been able to simply jog the machine around which seems to work just fine.  When I create a simple design to cut out, the machine will start out the first pass just fine and then will lose steps.  It's happened along both the x and the y axis.  I had one simple pattern to cut out 3 times along the length of a board and thinking perhaps the depth of cut was too much (4 passes with .25 end mill through .75 thick cedar) so I increased the passes to 10.  First pattern it went all the way around fine the first time and the second time around it lost steps and began cutting in the wrong location.  It seemed like it worked ok for the rest of the passes, so I let it finish with the first pattern hoping the second one would be a good one.  It went through 8 passes without missing a step and then on the 9th pass it missed again.  

My next thought was if I could only get a template cut out of a thinner material, then I could use that to make the pieces using a band saw and a pattern bit.  I looked online as lost step issues and on the xylotex site and I looked at multiple factors.  I was using an extra long 25 pin cable, so I switched that out for the cable that came with the unit.  This did not help.  I thought perhaps the machine I was using was too old/slow, so I replaced it with a newer faster machine.  Still the same issue.  I saw on a thread a where a number of people talked about using something called a UC 100 to correct for their issues, so I purchased one of those.  That too didn't fix the issue.  

I'm really hoping that someone out there might have other ideas for what is causing the loss of steps.  Again, there are no issues while jogging the machine; however, during program execution through Mach 3 there are lost steps seen with both the x and y axis.

Regards,

Brian

[Jeff Pollard]

Hi,

  Generally, loosing steps in because of too high a top speed, or too fast an acceleration.  Try running the program with lower rates and see what happens.

  Also mechanical issues can cause loss of steps.  Make sure that there is not any binding in the ways in all areas of motion/travel.  Lubricate any screws that might have too much friction as well.

  check for mechanical backlash.  This can cause parts to not be cut as expected.  Use the software backlash compensation as required.

Jeff

[Brian Harvey]

Well, it's taken me some time to get back out to the shop and play around with the settings, but I am finally able to report back...

I tried slowing down the machine and had some success.   If I drop the feed rate in mach 3 to 10 (down from 100), I can sometimes get a part cut out.  The longer the machine needs to run the worse my chances seem to be.  I watched some videos on troubleshooting this and then experimented using their tips.  The one that seems to be the most revealing is where I marked a line on the motor housing and the shaft, then I went into ports and pins to disable the other motors, and finally ran the program.  Again even at low speed it lost steps.  I unhooked the screw from the motor, reset everything and ran the program again.  Same results as before, lost steps.  This seems to say there must be something either with motor, controller or computer.  I'm only losing steps in the x axis.  I'm using both the A and Y axis to drive the gantry.  I'm not sure if I should switch out the motor for one of my others and then try to run the program again or what my next step should be.  Any other advice or troubleshooting tips would be greatly appreciated.

[Brian Harvey]

I need to also provide additional info based on my original posting.  Initially it seemed like it was both x and y and also jogging seemed to work ok.  The more I troubleshoot the lost steps and have gone through the diagnostic steps of isolating each axis and running the program... I am now seeing that the lost steps are only in the x axis.  As for jogging... similar to my earlier reply, jogging does seem to work fine... up to a point.  The longer the controller and motors are on and running the worse it gets.  By that I mean, when I first turn on the machine and jog it left, right, forward, back, up and down... everything seems to move just fine.  If I run a program to the point where it starts to lose steps, after stopping the program and clicking reset I am seeing and hearing motor hesitation and jogging the x axis does not seem to work.  I have a jerky motion and sometimes mach thinks I'm moving but the machine is doing nothing.

[Jeff Pollard]

Hi,

  What Mach shows on its screen doesn't have anything to do with reality.  It is only what it "thinks" it has done.

  I'm wondering if you are getting overheat problems.  Are both fans running on the drive box?

Jeff

[Brian Harvey]

I moved the drive box from underneath the table of the CNC in order to check the the fans.  Both seemed to be running and the air never seemed to get very hot.  I ordered brand new motors so that I could switch out the x axis to see if that might have been the issue.  Even with a brand new motor, the x axis is still missing steps.  I'm starting to think it's something in the drive box, but I'm not sure how else to determine that.  

So far I've 

• replaced the 15 ft 25 pin cable with the 3 ft cable that came with the drive box
  
• replaced the pc with a newer faster pc running windows 7
  
• switched from 25 pin printer port to using a UC100 USB connection to the drive box
• isolated each axis within ports and pins and then ran a sample program to check for lost steps
  
• Removed the lead screw from the motor and allowed the motor to run the program by itself
• marked a line from the rear shaft to the motor housing in order to determine any lost steps
• replaced motor with brand new motor

I'm at a loss for where to go from here...  Is there something that can be done with the drive box?  

[Jeff Pollard]

Hi,

  Try this test:  Set the software X axis as pins 4 and 5 (step and dir respectively), and set the Y axis as pins 2 and 3.  Now the red cable will be the Y axis and the yellow will be the X axis.  Then plug the yellow cable in to the X axis motor, and the red cable in to the Y axis motor.  This will test the Y  drive chip in the drive box on the X axis (the one you were having problems with). If the drive box X axis is the problem, then the problem you were seeing on the X axis should be gone, and should now appear on the Y axis.  If the problem stays on the machine X axis then something else is going on.

Jeff
 

[Brian Harvey]

I'm excited to be able to report the issue followed to the y axis. Now that it feels like we've found the issue (potentially), what can be done to repair it?

[Jeff Pollard]

Hi,

   By Y axis, do you mean the problem is still on the RED cable (which was the X before)?

Jeff

[Brian Harvey]

Correct... the issue moved from the x to the y.   No issues on the x axis after switching the pins like you said.  

[Jeff Pollard]

Hi,

  Perhaps I should look  the box.   But first.....

  The reason I was stressing RED, as opposed to X or Y, is because X and Y mean nothing to the drive box.  They are just human notations (and software notations).  The box (and thus my interpretation of what is physically going on) only knows about the cables coming out of the box (Red, Yellow, Green and Blue).  While you answered affirmatively, you used x's and y's.  What I'm interested in is the color of the cable that has the problem.  Is it, and has it always remained with the RED cable, even after changing things around in software? (or did the problem move to the yellow cable)

jeff

[Brian Harvey]

I can go out tomorrow before I leave for work and snap a pic of the cables.  I believe the cable colors you mentioned are correct.  I have 2 of the cables coming out of the box going to the 2 motors driving the Y axis, I have 1 plugged into the motor driving the x axis and 1 plugged into the motor driving the z axis.  I printed off the instructions from your earlier post and when I went out I was pleased to find that the Y axis motor closest to the X axis was the yellow cable and the X axis was the red cable.  Since I use the A axis slaved with the Y to drive the gantry, my fear was the one I needed to swap with the X would be at the opposite end of the machine.  

[Jeff Pollard]

Hi,

  When you take a look at the machine, remember to only be concerned with the color (red, yellow, green, blue) of the cable, as your logical X, Y Z, and A are only really significant to your software use.  The way the drive box "thinks" about it is:

Pins 2 & 3 - RED
Pins 4 & 5 - YELLOW
Pins 6 & 7 - GREEN
Pins 8 & 9 - BLUE

  You could set up the software "X" axis to use pins 1 and 14, in which case the drive box would not respond at all, and the software would think everything is just fine, and still show numbers changing on the screen and drawing pictures on the graphics screen like everything was OK.

  Unless there is an overheat condition, the axes usually either work, or they don't work (assuming the motor cables and parallel port cables are all firmly seated and good).  Thus, if you were able to cut the first pass fine, then the box would appear to be running fine as far as driving the motors is concerned.  Things like backlash compensation, or trying to go too fast a speed, or accelerating too fast can only be fixed by changing the software parameters or changing the machine to have less weight or friction, or taking lighter passes with a sharper tool, etc..

  If you need to send the box in, please contact me through e-mail.  Go to the "Contact Us" link on the left hand side of the Xylotex web page for the e-mail address.

Jeff

[Brian Harvey]

Jeff,  

I tried send you emails the last 2 days and both have bounced back with the following error, "LMTP error after end of data: 552 5.2.2 <*****@x*****x.***> Quota exceeded (mailbox for user is full)"

Note: I changed the email address in order to obscure it on purpose.

Here are the photos I snapped after moving the red cable from the X axis to the Y and the Yellow from the Y to the X and changing the pins as you mentioned within Mach 3.  The final position after executing the program was off in both x and y; however, this is due to the fact that my y is driven by 2 motors and the lost steps caused the gantry to skew slightly.

[Jeff Pollard]

Hi,

  I actually did receive the e-mail, but with no pictures.
  You should have received my reply.

Jeff