motion problems
27 Mar 2016 00:53 #72201
by jim_dewit
motion problems was created by jim_dewit
Hi, completely new to CNC but was given a CNC mill recently, so needless to say the last few weeks have been a crash course in cnc hardware & electronics, linuxcnc, gcode, and CAM (fusion 360)...
The machine is a ZAY7025FG has travel of about 9x8x14 (xyz) 370 Watt motor (kinda wimpy if you ask me)
the electronics are:
drivers KL9082
BOB C10
motors KL34H2120-60-4B (1200oz-in) on all axes
power supply KL7220 (72v 20A)
leadscrews rolled C7 with 5 turns per inch
The mill itself is in almost new condition and the conversion was very well done in my estimation, with one failing however. The thrust bearings are just simple thrust bearings (no preload) which are trapped against the shoulder of the leadscrew and the stepper mount. But with this system I was able to get the backlash down to 0.0017 on X and 0.0027 on Y. I adjusted the .ini file accordingly for backlash compensation and tested with dial indicator and this is working well with backlash on the order of 0.0001
So I figured the mill was working well and so I tried to make a part with a bearing pocket when I started noticing problems.
First I noticed the diameter of the circle is not coming out to size. Was supposed to be 0.906 but coming in about 0.012 under. I chalk this off to cutter deflection or some other unknown and simply compensated by commanding a bigger hole. Doing this I was able to get a really good fit for the bearing - the circle is within 0.001 on the diameter no matter where I measure it, and the bearing has no play in the pocket at all... So far so good.
Then I measure the location of the pocket on the part and its off by about 0.05 in. both in the X and Y axis... So I tried to replicate this and found that when I command a 2 inch move, it goes 1.954. Both X and Y do this. So this is probably the reason for the undersized as well, not cutter deflection. The odd thing is, repeatability is perfect. I can move the axis all over the place, hundreds of moves at 60IPM and it will come back to the exact same position as it started.
My only lead right now is something to do with the drivers? I could not find the timings for my drivers anywhere, but I know they are actually made by Leadshine (not Keling) and found a driver with the same specs and so used those timings.
steptime 2500
step space 2500
direction hold 10000
direction stop 5000
I also tried 10000 for all values to see if it made a difference - it didnt. The other lead is I do get realtime errors, although my settings are above the latency I get during latency tests... And again, I never have a repeatability issue, I can jog around for miles and always can return to the same spot on the part and DRO.
Help?
The machine is a ZAY7025FG has travel of about 9x8x14 (xyz) 370 Watt motor (kinda wimpy if you ask me)
the electronics are:
drivers KL9082
BOB C10
motors KL34H2120-60-4B (1200oz-in) on all axes
power supply KL7220 (72v 20A)
leadscrews rolled C7 with 5 turns per inch
The mill itself is in almost new condition and the conversion was very well done in my estimation, with one failing however. The thrust bearings are just simple thrust bearings (no preload) which are trapped against the shoulder of the leadscrew and the stepper mount. But with this system I was able to get the backlash down to 0.0017 on X and 0.0027 on Y. I adjusted the .ini file accordingly for backlash compensation and tested with dial indicator and this is working well with backlash on the order of 0.0001
So I figured the mill was working well and so I tried to make a part with a bearing pocket when I started noticing problems.
First I noticed the diameter of the circle is not coming out to size. Was supposed to be 0.906 but coming in about 0.012 under. I chalk this off to cutter deflection or some other unknown and simply compensated by commanding a bigger hole. Doing this I was able to get a really good fit for the bearing - the circle is within 0.001 on the diameter no matter where I measure it, and the bearing has no play in the pocket at all... So far so good.
Then I measure the location of the pocket on the part and its off by about 0.05 in. both in the X and Y axis... So I tried to replicate this and found that when I command a 2 inch move, it goes 1.954. Both X and Y do this. So this is probably the reason for the undersized as well, not cutter deflection. The odd thing is, repeatability is perfect. I can move the axis all over the place, hundreds of moves at 60IPM and it will come back to the exact same position as it started.
My only lead right now is something to do with the drivers? I could not find the timings for my drivers anywhere, but I know they are actually made by Leadshine (not Keling) and found a driver with the same specs and so used those timings.
steptime 2500
step space 2500
direction hold 10000
direction stop 5000
I also tried 10000 for all values to see if it made a difference - it didnt. The other lead is I do get realtime errors, although my settings are above the latency I get during latency tests... And again, I never have a repeatability issue, I can jog around for miles and always can return to the same spot on the part and DRO.
Help?
Please Log in or Create an account to join the conversation.
- Todd Zuercher
- Offline
- Platinum Member
Less
More
- Posts: 5007
- Thank you received: 1441
27 Mar 2016 01:34 #72202
by Todd Zuercher
Replied by Todd Zuercher on topic motion problems
Sounds more likely that your kit came with 5mm pitch screws rather than 5 TPI. (the math works out almost perfect.)
Please Log in or Create an account to join the conversation.
27 Mar 2016 03:20 #72203
by jim_dewit
Replied by jim_dewit on topic motion problems
man.... its too late to go out to the shop to test this out, but this seems very likely / embarrassing. none the less it probably would have stumped me for a long time - thanks and ill confirm tomorrow morning!
Please Log in or Create an account to join the conversation.
27 Mar 2016 17:49 #72219
by jim_dewit
Replied by jim_dewit on topic motion problems
changed my leadscrew settings and preliminary looks like that fixed it. Im new to this board but I love it already. Thanks!
Please Log in or Create an account to join the conversation.
Moderators: piasdom
Time to create page: 0.234 seconds