Hello,

I'm relatively new to this forum and Mach 3.  I have purchased a 2nd hand DSLS 3000 controled Taig 3 axis mill that was supposed to work correctly.  It is running R.1.84 version Mach 3 on a 1 Ghz computer.  This is the only software that runs on the system and the CPU loading is less than 5% when operating.  The driver test program indicates that everything looks good, i.e. there is not great excursion from the  25Khz signal trace.

My problem is that when I run a g-code program, the Reset line goes high and the log trace (lasterrors.txt) indicates that the limit switch has triggered.  To the best of my knowledge, there are no limit switches on my hardware.  I see none on the stages so I'm assuming that the limit switch error is due to some sort of following error with the encoder(?)....  

I have adjusted the stage speed down to 40 ipm and 4 ipm/pm.  

The g-code program that I'm running is using a feed rate of only 15 ipm.   If I change the feed rate down to 10ipm, it fails with the limit switch error after only a few moves.  It runs better at a feed rate of 20 but still fails at certain moves....

Does anyone have any idea of what's going on here?

Thanks in advance,

- Dave

please send XML file

Limit switch warnings are usually noise related but if you dont have switches fitted then they should not be enabled and you should not see that error.
As mamardam says attach your xml so we can see how things are set up.
Hood

If you really don't want limit switches, disable them in the inputs config page.

Thank you for the responses.  I've attached the config file.  I believe the Limit Error is actually a servo error in disguise. 

I had previously routed all control cables away from AC and spindle cables.  These tests were done without the spindle running as well.

Below is an area in the BobCad generated G-code that fails.  Specifically the XY move on line 182 generates the error.  It sounds like the stages are binding but when I jog through that area everything is clear and smooth.  It looks like a slight G-code generation error since line 181 goes to rapid movement (G00) so line 182 would be rapid as well.   In other areas of the code the Z movements (181 & 183) are always sequential, i.e. there is no XY movement between Z motion.  I've tried changing line 182 by adding a F15.0 feed rate and it still fails.  I've reduced the stage speed to 20 ipm and it still fails.  In any event, there are other places in the program where I receive the Limit Error.  This is just very repeatable.

N175 X-.6434 Y-.355 F15.0
N176 X-.6381 Y-.3505
N177 X-.6277 Y-.344
N178 X-.6172 Y-.3399
N179 X-.6023 Y-.3375
N180 X-.0012
N181 G00 Z.1
N182 X-.0086 Y.3375
N183 G01 Z-.1031 F12.2231
N184 X-.5875 F15.0
N185 X-.6089 Y.3381


Finally, the DSLS 3000 docs talk about checking the voltage across the high and low power capacitors.  Could bad caps cause the problem I'm experiencing?

Once again, thanks for the support!


- Dave

You have limits enabled so if you are certain you dont have switches on the machine you should disable them and likely your problems will go away.
Hood

Thank you Hood.  I agree that turning off those limit switches will probably keep the error from occuring.  But, don't think that will solve the real problem...

I believe the DSLS 3000 is wired to the limit switch I/O port to trigger the "Limit" when the encoder feedback is outside of some error band.  So, the system is working correctly from the stand point of triggering the limit switch when the motor/stage doesn't respond as it is programmed to.    Thus, when a stage binds or hits end of travel, the encoder feedback error increases and the "limit" is triggered.

My problem is that on the move described above, the stage doesn't move as programmed (and the encoder error band is exceeded).  I suspect that the pulse train generated by Mach 3 is getting polluted or corrupted in the DSLS 3000 box so when it gets amplified and sent to the stepper motors, it doesn't correctly move the motors.  Does this sound feasible? 

The documents on the DSLS 3000 talk about checking the voltages on a high and low voltage capacitors when this type of thing occurs.  Does anyone have any input here?  I'm going to check the voltages this weekend.   

Any other suggestions or ideas?

Thanks.


- Dave

If that is indeed the way things are connected then I would make sure there is no binding on any axis. Checking the voltage after the caps when all axis are running will tell you whether its dipping or varying too much, it should be fairly steady if the power supply is adequate.
Hood

I was correct in that the servo error is mapped to the limit switch input so when the DSLS 3000 detects a mismatch, it faults by raising the limit switch flag. 

However, the root of my problem is actually that backlash compensation was enabled in Mach 3.    If I turn BL comp off, the system runs fine.  If I turn BL compensation on, the system fails as described above.

Now the question is:  Is there a difference in BL compensation between V1.84 which I'm running and the latest version?  Also, will the license for V1.84 work for the latest version?

Thanks.

Now the question is:  Is there a difference in BL compensation between V1.84 which I'm running and the latest version?  Also, will the license for V1.84 work for the latest version?

Thanks.

I think backlash was worked on and improved since 1.84 but as I dont use it I cant say for sure.
Yes your licence will be fine for all versions of Mach to date.
Hood