New Home › Forum › don’t use this one, I need to sort and delete -Troubleshooting › X/Y Jitter, Middle Assmbly Problems
- October 18, 2016 at 10:49 am #19668
I have a 5’x5′ build, 48″ x 48″ cutting envelope with rail supports. I am using all the updated 525 parts, printed by me. Running with mach3 using breakout board and haoyu tb6600 drivers and 42BYGHW811 steppers 67oz-in.
I have been trying to troubleshoot a Jitter problem I have been having. It is being transferred into the work. Here’s what I have done to to try and resolve the issue.
1) Reconfirmed the frame is square and parallel. Diagonals and side rails are all within 1-2mm.
2) Tighten the belts to what I what I figure they should be and have adjusted a little beyond. Too tight?
3) Sanded all rails with 280 grit.
4) Square the rollers using a measure tape before tuning on the motors and locking it in square.
5) Played around with different speeds.
6) Played around with the stepper current with not significant change.
I have also played around with the middle assembly tension bolts and have noticed that I could not get both tension bolts a and tension bolts b to make full contact. Not sure if that matters and if so how to fix it. The jittering also seems worst when x and y are moving at he same time, like circles, curves and diagnonals. I think I have mentioned everything I can think of at the moment. P.s I do love my mpcnc, great design. Also I have used all the electronics on a different cnc with no trouble in the past.
Attachments:October 18, 2016 at 11:53 am #19671
Do you inow what your step rate is? Too low and it is noticeably twitchy. Micro stepping at 1/4 or higher is best as far as I can tell. 1/8th is really smooth.October 18, 2016 at 12:24 pm #19674
I have the exact same size machine with exact same problems. Performed all the same trouble shooting steps. My worst jitter happens on the x axis more then the Y.
My microstepping is 1/32 i think, all the jumpers on the ramps installed and DRV8825 drivers. Should we change that to 1/8 (jumpers 1 and 2 only, steps/mm = 50)?
My conclusion up to this point is the rails must be flexing on the tiny pits in the EMT conduit causing it to bounce back as it rolls. My solution very soon will be to reprint the entire thing and upgrade to 1″ stainless rails.October 18, 2016 at 12:27 pm #19676
I’m running at 100 steps per unit with 16t pulley. Running in 1/16. Should I maybe try 1/8 or half?October 18, 2016 at 12:29 pm #19677
No, the higher the step rate the better.October 18, 2016 at 12:29 pm #19678
Are you using EMT conduit as well?October 18, 2016 at 12:32 pm #19679
I don’t think that is really causing the problem as mine doesn’t happen repeatively in a certain area and to the same degree running the same code in the same location, also is seems like a little to play to be the flex only at certain times. Maybe wrong but I don’t think that really seems like my problem. Thanks.October 18, 2016 at 12:33 pm #19680
3/4 EMT 23.5mmOctober 18, 2016 at 12:33 pm #19681
I was think about adding a second set of rollers to shorten the axis’ for jobs that are smaller and require more precision. Basically cut the machine in half which would double the rigidity of the gantry rails
Does yours seem worse the slower it moves?October 18, 2016 at 12:44 pm #19683
No I would say that it seems better but it’s also moving slower so I think you just can’t notice the twitch because of the speed. But if I cut a piece of material slower it’s seems to turn out the same or maybe a little better.October 18, 2016 at 12:57 pm #19684
Do you think the the middle assembly bearing no making full contact all the time could be playing a part, tension bolts a and b. Also I am using unshielded wiring so could interference be causing missed delayed steps and the two motors on the same axis aren’t moving at the same speed for that moment? Just a thought.October 19, 2016 at 4:49 am #19704
I have tightened all the bolts on mine so all the bearings touch and it made it worse. I ran it real tight for a few hours hoping to wear the zinc coating off the rails. It did wear a bunch off but did not improve smoothness of the movement.
I am almost 100% convinced that it has to do with the texture, length and flex of the emt rails at that length. I discovered yesterday that my local hardware store has 1″ x 72 stainless closet rods. I am going to start reprinting the 1″ version. Those closet rods and quite noticeable stronger then the emt.
@vicious1 can I leave the 3/4″ legs in place? Will the 1″ corner pieces sit on top of the lock corner?October 19, 2016 at 8:33 am #19710
Yes the other size legs and locks should fit close enough. I just checked, and they are just a bit smaller so you probably won’t even notice the lock to bottom difference.October 20, 2016 at 3:18 pm #19792
So i finely tuned my stepper drivers and metered them out and it did see, to make some improve. But i decided to run a breaking code drawing circles and while examining the machine I noticed that on every rollers there is one or 2 bearing not making contact at all times. Could this be causing the Jitter and how would I fix it?
All roller bearing were making contact at time of installOctober 20, 2016 at 3:23 pm #19793
Take the middle assembly out, by removing the roller motor mounts only, and inspect everything again, something obviously moved or seated in a bit more. while apart Check to make sure the frame is square, all the roller bearings touch again, and the middle assembly is still square with the z axis in place.October 20, 2016 at 4:07 pm #19798
Pulled the motor mounts, checked the frame everything in the outer rails are square and parallel.
The rollers have all worked loose, I have attached a video below. At time of install they were all tensioned the same and were snug on the rails. What would be the best way to snug these up again?
Also it do look like the middle is out of square. What would be the best tension bolts to adjust this?
Attachments:October 20, 2016 at 4:11 pm #19802
The rollers snug up with the long bolts, and the middle is best to try and just mess with the 6 long bolts but if it doesn’t square up take it apart and re assemble it. It doesn’t seem like a bad thing, your parts just relaxed a little I would guess. I know it’s a bummer to have to redo some of it but it will probably be much more solid this time around.
Could be the plastic, could be the weather who knows.October 20, 2016 at 4:12 pm #19803October 21, 2016 at 9:42 am #19827
Ok took the middle assembly apart and redid it and made sure it was square. So cut out some circles and they are from circles. What could cause this. It seems like it’s out mostly in one axis. I checked the belt tension and all seems even as before. Any thoughts?
Also.. I done some test squares and they were perfect width and diagonals.
Attachments:October 21, 2016 at 9:46 am #19830
Cut larger circles, much easier to calibrate and diagnose issues. Or if they are large put in a tape measure or something to show scale.
Make sure you are not using arcs. It looks like you have arcs on and they do funny things to small circles. I have not calibrated them in the firmware as I haven’t had time to fully dig into the formulas yet.October 21, 2016 at 9:53 am #19831
They are 50 mm and I’m not running your firmware. They were produced through cambam and ran through mach3. Also ran some from estlcam just to make sure.October 21, 2016 at 9:56 am #19832
Again. Make sure you had arcs turned off when you generated the gcode . Can’t really help with other firmware other than the one I recommend.October 21, 2016 at 10:01 am #19833
How would I turn off arcs in estlcam.
Edit. Thanks found it and it was turned on. So I will retry.October 21, 2016 at 10:28 am #19835
Recut with arcs turned off. But same results.
Attachments:October 21, 2016 at 10:35 am #19838
Okay. A lot more information is needed.
What board, what firmware, What speeds are you cutting, what depth of cut, what rpm, what bit, what material.October 21, 2016 at 10:44 am #19839
Using mach3 with a parallel breakout board and haoyu tb6600 stepper drivers. Cutting at 1200 mm/min. 0.6 mm per pass. Makita rt0701c running at 30,000. 3/16 endmill 4 flute. Cutting in hard plastic.October 21, 2016 at 10:51 am #19841
It could easily be a firmware/acceleration thing on how your board handles tiny changes in direction. Marlin has look ahead, not all boards do and need the software to do it for them, I don’t know what yours needs.
A 4 flute bit is not a good choice for that high of rpm.
How long is your zaxis? more than 3″? that could be your problem. How large is your machine? too large could also be the problem.
At this point I can’t really help you it could be many thing. I would say list as much detail as you can, including answering the questions in the troubleshooting sticky so maybe someone else using similar hardware can help you.
Pictures go a long way. You don’t really give much information to go on so full pictures of your machine, pictures of your bit, screen shots of your settings. You need to post it all. Most people here are using the recommended hardware, making it very easy for me to diagnose things with little information. You need to flood the info here so someone might spot the issue.October 21, 2016 at 11:04 am #19842
I can upload and give as much information as possible able my software and driver and the machine. I answered all the troubleshooting quesions at the beginning of this thread. But I do not believe that this is related to the electronics. I have cut these same circles on this same machine without this issue. I had Jitter problem so you recommended resquaring the middle and tighten the rollers. That is all that has changed for these results to change. My thoughts were that maybe there is still a problem with the middle assembly or too much belt tension on one axis causing missed steps. Thanks for all the help.October 21, 2016 at 11:20 am #19843
I have a 5’x5′ build, 48″ x 48″ cutting envelope with rail supports. I am using all the updated 525 parts, printed by me. Running with mach3 r1.83.027 using breakout board and haoyu tb6600 drivers and 42BYGHW811 steppers 67oz-in (2.5a per phase at 3.1v). I’m running them at 1.14a in parallel with 13.8v supply. Some pictures attached to describe other information.
Attachments:October 21, 2016 at 2:24 pm #19855
Your machine is extremely tall for such a heavy router and large build volume you benefit greatly from reducing it by at least half. More accurate and faster cuts.
Your acceleration looks really high ours is set at 400mm/s/s yours is almost 700.
You are running your steppers at 1/4th power? Either wire them in series or turn them up. (we run ours at 50% power put have larger steppers. I also have series harnesses coming so running them at 90% long term is possible).
You must be logged in to reply to this topic.