|
Register | Options | Profile | Last 1 | 3 | 7 Days | Search | Today's Posts | Mark Forums Read |
|
Thread Tools |
#31
|
|||
|
|||
Before you weld the pinions, you might try some lock-tight or similar. Or at least take off the pinion and see if there are signs on the shaft of slippage (scoring, etc).
I agree with Hennie, cutting speed helps with lost steps. The idea is to start slow, and speed up till you find some are skipping, then back off a bit for 'production' use. |
#32
|
|||
|
|||
Just FYI, I used 2 set screws in each pinion (1 jammed on top of the other) with lock-tight.
Sorry to hear you're having issues, hope you get it nailed down. Mark |
#33
|
|||
|
|||
Quote:
That's a good idea. I've got some Loctite 640 sleeve retainer. I'll add that to the shafts. I've read good things about the strength of that stuff. My rapid/cutting speed is only 200 ipm, with acceleration at 20. I can't see how that can be causing problems. I'm sure it's something loose, I just can't find it. Lee |
#34
|
|||
|
|||
I've never ran over 130 IPM, but I've got a pretty heavy gantry for a 60" Y dimension....
I believe I'd lose steps at 200IPM... |
#35
|
|||
|
|||
But wouldn't loss steps show up in both directions in my criss-cross drilling pattern?
|
#36
|
|||
|
|||
The devil is in the details. ... Start out at something ridiculously slow, then move up slowly. If you have the patience, try a 'binary search', possibly one axis at a time.
Binary search, (example only) 200 is to fast, 30 works. Go to the middle of that range ( ((200-30)/2)+30 = 115). If is works, try ( ((200-115)/2)+115 ~= 157 ), wash, rinse, repeat. This will find the sweet spot with the fewest 'tries' for your current setup. You will know any slower always works, anything faster 'has issues'. Now this is just get geek analyst in me. After you make changes, start with your 'current best speed' as the low speed, and 'desired speed' as the high speed, then as they say 'wash, rinse, repeat' to find your 'new optimum. At most if starting at 250 for the high end, you will take 8 tries. You can find an 'real close' in 4 trys on the average. (yes, the number numbers to go up one try is basically double the size of the range desired - and to be 'technically correct' you are guaranteed to find an answer in 8 tries with 256 range, a range of 512 is 8 tries, 1024 is 10 tries, 4096 is 12 tries, you get the idea. But that is my inner OCD geek talking now. If it doesn't make sense, don't worry, just use it and it will work for you.) I also suggest you make a small 'test cut' that you can use as your standard test that doesn't take long to cut so you can make another try quickly without wasting a lot of time or $$ on stock. Most of all, get it working for YOU and enjoy! |
#37
|
|||
|
|||
Fairly certain I'm losing steps.
Pulled the gantry against the stops and zeroed the machine. Ran my hole drill test at 200ipm, 20ips/s acceleration. (approx 120holes drilled) When I ran the machine back to the stops, it was .083" off on the X axis, and .034" on the Y. Still need to test at slower speeds, but preliminary testing shows no improvement at 50ipm. Very frustrated, Lee |
#38
|
|||
|
|||
I have run 250IPM cutting HDPE. 180IPM with 1/2" bit in 3/4 plywood. There is something else amiss here. Something that is being overlooked.
|
#39
|
|||
|
|||
Lee, I feel your pain, I had problems intermittently with mine. They can hide and make you want to pull your hair out. I am practically doing that here trying to think of what your problem is coming from.
|
#40
|
|||
|
|||
From what I've been reading:
1. Bad parallel cable (too much resistance. 2. Not enough memory / processing power in computer. 3. Grounding ? Does anybody know if the Gecko 203V work Active High or Active Low? Post #15 on the Artsoft (Mach) forums says that if they are reversed, the drives will lose steps at each direction change. |
#41
|
|||
|
|||
I will go down and look at mine
|
#42
|
|||
|
|||
All step low active and dir low active have a red x in them on mine
|
#43
|
|||
|
|||
Same here.
|
#44
|
|||
|
|||
are your cables shielded to the steppers?
|
#45
|
|||
|
|||
Yes. Grounded on one end only. (main Gecko mounting plate)
|
#46
|
|||
|
|||
If you have a dial indicator, you could write code to make the machine run in the X for the length of the machine and back lets say 10 times. I would do both at feedrate and then 2nd test with rapid moves. Lets first try to see if it is your rapids. Do this with the universal motor off (the router). Then do the same thing with the router on. 4 tests that will start giving a better idea of what is happening. Now if I remember correctly you have geared motors so I would expect to see 0.008" error on the return. So we are looking for something more than that.
|
#47
|
|||
|
|||
Quote:
Did you leave the shielding on the cables up to right before the point of termination inside the electrical enclosure. I have see drives act weird if the wires are next to other power wires that are excessively noisy. |
#48
|
|||
|
|||
Quote:
Going to set up the "back and forth" test tomorrow. Thanks for your help. Lee |
#49
|
|||
|
|||
SOLVED
Needed to tune the drives to the motors. They always sounded "okay", but I tuned them better then ran a X, Y, & XY rapid test at 300ipm rapid, 25ips/s acceleration. (end to end ten times with dial indicator set at home position). Came back to within .008" I then ran my crisscross hole drilling test. Again came back to within .008" and holes are now drilled concentric (no offset) Thanks for the help everybody! Lee |
#50
|
|||
|
|||
Excellent news. Thanks for sharing the solution.
|
Register | Options | Profile | Last 1 | 3 | 7 Days | Search | Today's Posts | Mark Forums Read |
Thread Tools | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Inconsistant cutting errors. | SOSx | Troubleshooting | 85 | Mon 12 November 2012 04:49 |
Stepper motor refuses to turn - Wiring & jumper errors | Gato Richy | Troubleshooting | 31 | Fri 01 May 2009 00:56 |
Stepper motors giving trouble - Motors burnt while serving as ground for elec. fault | Daya | Troubleshooting | 45 | Mon 21 April 2008 02:05 |