High spots between towers.

The new for 2016 RostockMAX v3!
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

High spots between towers.

Post by Gdot »

Hello

I have a Rostock Max V3 that i finally finished building. I have been struggling to get it calibrated. I have used the G29 command in conjunction with both the https://www.seemecnc.com/pages/delta-calibration-wizard and / or http://www.escher3d.com/pages/wizards/wizarddelta.php. This however never seems to end with a proper calibration. I seem to be level at the 3 tower locations as well as bed center, but all 3 spots opposite of each tower are high. I have checked numerous items on my printer by disassembling and reassembling but I have yet to find the reason I get high spots between towers but level at each tower and bed center. I've made sure towers are as square as i can make them. What makes it seem odd is the glass seems to be more flush at each tower with the melamime, but seems to sit .5mm or higher than the melamine opposite of each tower yet the nozzle rides high here. The other odd item is my z tower always needs more off set than the x and y tower, currently after G29 command, the z tower is 130 and the Y tower is 0 and the x tower is 34. I have double checked multiple times that the towers are sitting fully seated and even checked the z tower switch but I can find no issues. I manually adjusted the z tower endstop screw to bring all the offsets within a few digits but after running G29, the high spots still remain. I have checked belts, tension, servo pullys for the belts, checked for missing teeth on the belts...everything I can think of. Can someone please advise what steps I should take next?

Thanks!
User avatar
Jimustanguitar
ULTIMATE 3D JEDI
Posts: 2631
Joined: Sun Mar 31, 2013 1:35 am
Location: Notre Dame area
Contact:

Re: High spots between towers.

Post by Jimustanguitar »

Following :)
Juicy
Plasticator
Posts: 8
Joined: Sat Dec 31, 2016 3:00 am

Re: High spots between towers.

Post by Juicy »

I' ve been having leveling issues that I cannot fix either. I have probably printed over 100 of these disks http://forum.seemecnc.com/viewtopic.php?t=7361 trying to get a perfect first layer. When I print the 2mm disk, I am always lacking material in the XY half of the circle and over extruding filament on the Z half (seen here) http://i.imgur.com/6twe5XH.jpg Another file that was recommended for leveling is this one https://www.thingiverse.com/thing:50505.
My problem seems to be that I cannot get my bowden tube to seat all the way no matter what I try. There is always a "hockey puck" under the the bowden when I remove it (at 100 degrees C). Are you certain that your hot end does not have any jam? Does your extruder grind filament leaving dust all around the gear?
I've decided to ditch the HE280 all together for an e3d v6. I think it's a POS and it appears that a very large amount of people are having calibration/leveling/jamming issues with this hot end.
Here are some other threads that might help. http://forum.seemecnc.com/viewtopic.php ... ng#p104916 There is another link in that thread for calibrating but it is pretty outdated. Did not work for me but it worked for the user who linked it and he was having the same problems as you.

Hope this helps, good luck!
Attachments
75mmDisk-pt2mmtall (1).stl
Attached the specific file ive been using to test my leveling.
(24.89 KiB) Downloaded 237 times
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

Well, shockingly filament jamming was the one issue I was actually able to solve. It ended up being that my bowden tube was not seated correctly and there was a huge puck of melted filament. The bowden tube was not going to seat correctly till I removed the puck. I ended up using the timskloss method http://forum.seemecnc.com/viewtopic.php?t=11457. Once that was corrected I was able to repeatedly print with no jams as low as 200degC. Before that, I couldn't print at anything less than 215 with out jamming. I have hardly any dust or shavings built up on the extruder gear either. I have my retractions set to 2mm and speed set to about 30 or 40mm with zero issue now since getting the tube seated correctly.

One other item I find odd, is when setting the Z height with the either the knob on the lcd screen, I have to do it multiple times till it is actually set correctly. Each iteration will seem to leave it too high after checking pinch, but it seems to take at least 2 to 3 iterations to have the perfect pinch on the paper. At one point I thought I might be losing steps, so I bumped up the current settings in firmware but this did not solve my calibration issue and only made the servos run hotter, so I set these back to standard settings. I then marked starting and stopping points on all 3 towers for 100mm travel. After running each tower down by 100mm, they were spot on. So no issue there.

One question though, what does Harblar mean by "I then checked the width of the washers under the spider plate. 4 were right at 4mm and 2 were right at 3.87. I spread the 4mm ones across from each other and put the spider plate and heatbed back on." ?
Juicy
Plasticator
Posts: 8
Joined: Sat Dec 31, 2016 3:00 am

Re: High spots between towers.

Post by Juicy »

Gdot wrote: One question though, what does Harblar mean by "I then checked the width of the washers under the spider plate. 4 were right at 4mm and 2 were right at 3.87. I spread the 4mm ones across from each other and put the spider plate and heatbed back on." ?
Good to know you are you sure you do not have the jamming issues. I think that was a V2 thread so the heated bed is a slightly different design. I am not sure but the user may be referring to the rubber parts that hold the ONYX in place and are screwed together with the blue clips? have you verified that they are all oriented the same way? One of them could be upside down? Another thing I noticed that may have been affecting my ONYX; was the wiring from the soldered joints on the bed. I taped mine down because I thought they may have been putting some upwards force on the bed. I doubt it would help but I was somewhat paranoid about it.

Is there any difference when you try printing with none (or just 3) of the blue clips securing the glass?
Xenocrates
ULTIMATE 3D JEDI
Posts: 1561
Joined: Wed Sep 23, 2015 2:55 pm

Re: High spots between towers.

Post by Xenocrates »

On a V2, there are 6 nylon spacers that hold the "snowflake" (AKA the spider plate mentioned) above the main structure of the base. If they were particularly out of tram, the snowflake and onyx could be flexed.

Most causes are a mismatch between virtual models and reality. Namely a bent bed, or a poorly chosen calibration radius can contribute, although I believe there is also a chance that it is due to tower misalignment.
Machines:
Rostock Max V2, Duet .8.5, PT100 enabled E3D V6 and volcano, Raymond style enclosure
Automation Technology 60W laser cutter/engraver
1m X-carve router

Sic Transit Gloria Mundi
01-10011-11111100001
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

I have checked and rechecked the spacers that hold the insulator bed up and I always make sure the dimples / circles are facing down.

Ok, so some interesting developments though , I always thought it was odd that the z tower offset was always 130 points higher than the x / Y towers. I kept putting paper shims on the bed mounts on either side of the z tower till G29 produced a Z tower offset within 10 digits of the x / Y towers. Then I used the escher3d calibration tool. It came up with some weird negative offsets for the x and Y but I ran with it. My first test print showed various squished areas and high spots, so I began to adjust offsets in the eeprom. I adjusted back and forth when I finally became frustrated enough that I reran G29 while the bed / nozzle were still warm. The negative offsets were gone.After a small tweak to the X tower angle and a small adjustment to one of the endstops, the entire print looked to have fairly the same squish in all areas. Then, being tired, lazy and frustrated, I adjust the paper pinch test with the Z max height in the eeprom. I ran my test print and it was the first time I ever got it to print completely with no gaps or over squishes. But...after it got done with the first layer and went to print the next layer the nozzle jammed......... :evil: . I am hoping I just have too much squish still and need to bring the nozzle up a hair. The odd thing is that the bed around the Z tower is shimmed so high now that it was hard to roll the blue clamps over the glass, yet the Z tower offset is close now to the X / Y. When I put a 6 inch level on the glass though, it is level front to back and left to right just like the top 2 melamine pieces. I cant help but think that something is warped somewhere or one of the towers are longer / shorter than the others.
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

Well I got rid of my paper shims and ran the G29 calibration routine with the hotend and bed warmed up to operating temps, turning them off just before running the calibration of course. Then I even ran the least squares calibration with 10 points and 6 factors with the bed and hotend warmed up. This by far has made the biggest impact on getting me close to a good calibration. It seems I have a warped bed that gets worse once its warmed up to operating temps. I think I will take the bottom half of the printer apart and replace the bed insulator plate and take a good look at the 2 bottom melamine boards for warpage.
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

So, I have found by bed insulator to have some warpage to it and I have consistent high spots between my towers. I have a question though, when I run the calibration wizard, the z offset measured at the towers and center of the bed are always in the -.02 to .04 range, give or take, but at the points between the towers the z offset measured is always in the .30 +/- range. I never get below a .11 or so deviation. So today I tried randomly changing my delta rod length from 291.06 to 292, 293 ,294.06. When I got as high as 294.06 and ran G29 then the calibration wizard, low and behold, all the z offsets were pretty close to each other (-.02 to .04), at the towers and in between them. I was also able to get all the way down to .01 deviation when done. When doing the paper test at the outer perimeters, it had close to the same gap now as the center. Can anyone enlighten me on what exactly it is I have going one with this printer as far calibration?
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

Also, to add to my last post(if anyone is reading :D ). I reset my diagonal length to 291.06 and set my steps from 80 to 81 and was able to get a .02 deviation after one iteration of the delta wizard calibration using only 6 factors. So I guess I'm asking, does this info and my previous post shed any light on the actual issues is with my printer, because I know these 2 values are pretty much constant values? It's like there is some strange computational error with my printer.
User avatar
DeltaCon
Printmaster!
Posts: 616
Joined: Sat Nov 14, 2015 5:01 am
Location: Wessem, The Netherlands
Contact:

Re: High spots between towers.

Post by DeltaCon »

I don't think it is a computational error, I think you just fine-tuned your diagonal rod length.
Judging from your low deviation I guess I must do the same... ;-)
I am DeltaCon, I have a delta, my name is Con, I am definitely PRO delta! ;-)
Rostock V2 / E3D Volcano / FSR kit / Duet 0.6

PS.: Sorry for the avatar, that's my other hobby!
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

This is exactly my issue http://forum.seemecnc.com/viewtopic.php?t=4868 . 21 pages and no clear resolution that I could find....
User avatar
Jimustanguitar
ULTIMATE 3D JEDI
Posts: 2631
Joined: Sun Mar 31, 2013 1:35 am
Location: Notre Dame area
Contact:

Re: High spots between towers.

Post by Jimustanguitar »

Gdot wrote:This is exactly my issue http://forum.seemecnc.com/viewtopic.php?t=4868 . 21 pages and no clear resolution that I could find....
That's why I'm following your thread :)
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

I am still waiting on my new bed insulator to come, taking awhile since semme is moving buildings. I know its warped in a concave shape, but I dont think its the reason for my outer perimeter sine wave. Yesterday, I re-installed my existing heat bed and iterated through .1 steps/mm increments followed by a G29 till I slowly converged on the z-offsets from the probing gcode from the delta wizard. I ended up at 81.1 steps / mm. I then entered the values into the calibrator and was able to get a .04 deviation. Then after heating the bed and nozzle up for a good 10 minutes, I reran G29 followed by the delta calibrator and was able to get a .o1 deviation!!!!!! I ran my test circle, its pretty big, takes up most of the build plate, and with a .03 first layer, It looked really good as far as no gaps. There was a small area near the x tower that I think was not clean enough so the pla was not sticking to the glass. I will rerun the test this afternoon with a clean bed and retest first layer. Also, I will test a small 50mm cube to check for dimensions. If its off, I wonder if I can adjust delta rod length to adjust for the size difference? Does this effect only the x and y scaling when adjusting the delta rod length and not the Z? If so, maybe the best thing to do is to scale up by the appropriate factor in the slicer to adjust for scaling issues (if any) by me adjusting steps /mm to get rid of my outer perimeter sine wave.
Toddwagner
Prints-a-lot
Posts: 30
Joined: Wed Jun 21, 2017 8:55 pm

Re: High spots between towers.

Post by Toddwagner »

Gdot wrote:I am still waiting on my new bed insulator to come, taking awhile since semme is moving buildings. I know its warped in a concave shape, but I dont think its the reason for my outer perimeter sine wave. Yesterday, I re-installed my existing heat bed and iterated through .1 steps/mm increments followed by a G29 till I slowly converged on the z-offsets from the probing gcode from the delta wizard. I ended up at 81.1 steps / mm. I then entered the values into the calibrator and was able to get a .04 deviation. Then after heating the bed and nozzle up for a good 10 minutes, I reran G29 followed by the delta calibrator and was able to get a .o1 deviation!!!!!! I ran my test circle, its pretty big, takes up most of the build plate, and with a .03 first layer, It looked really good as far as no gaps. There was a small area near the x tower that I think was not clean enough so the pla was not sticking to the glass. I will rerun the test this afternoon with a clean bed and retest first layer. Also, I will test a small 50mm cube to check for dimensions. If its off, I wonder if I can adjust delta rod length to adjust for the size difference? Does this effect only the x and y scaling when adjusting the delta rod length and not the Z? If so, maybe the best thing to do is to scale up by the appropriate factor in the slicer to adjust for scaling issues (if any) by me adjusting steps /mm to get rid of my outer perimeter sine wave.
Any news on the new test?!?!
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

So, I ran some 10 mm cube tests and found that the x, y dimensions were in the 9.8 to 10'ish range....but the Z axis height was growing. I measured around 10.2mm +/- in height. This did not seem right so I am thinking this is attributed to the increase in steps I was trying. So I am resetting to 80 steps and re-calibrating to measure cube dimensions for comparison.
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

So, at 80 steps / mm my dimensions are still the same with a 10.2 mm height of a 10 mm test cube. So I retested with a .2 mm first layer as opposed to a .3 mm first layer and I got a height of around 10.15mm. To be continued...
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

This weekend I plan on running more test cubes with the 81.1 vs 80 mm steps to compare dimensions. I found it very odd that my measurements were fairly similar between the 2 settings and that my height measurement was more than the X / Y dimensions . I would think that there should be some difference between the 2 step settings . Something I found interesting as well was when I set my steps to 81.1 and my outer perimeter z probe offsets began to converge, the horizontal radius would decrease until it got to 143.20 something. The theoretical horizontal radius is 143.28, but when I set all settings back to standard and go though the calibration routine, the horizontal radius goes all way up to 144.1 to 144.2. I understand that the higher the horizontal radius is, the higher your outside perimeter points become compared to the center of the bed. Its almost like the points between the towers are obeying the higher horizontal radius setting and the points at the tower are not. It makes me think the problem is not the high points between the tower, but the low points at the tower. Anyway, something to ponder I guess.
Toddwagner
Prints-a-lot
Posts: 30
Joined: Wed Jun 21, 2017 8:55 pm

Re: High spots between towers.

Post by Toddwagner »

I look forward to seeing your dimension results. I was having roughly the same issue with my prints, so I took your advice and adjusted my setups/mm and when I arrived to 82.25 with a zprobe offset of -.2 my print was finally looking normal. I immediately printed a single layered 250mm disk and got these results... The print certainly looks level but between the x and z towers the lines didn't fuse together much. Did this happen to you ever?
Attachments
image.jpg
image.jpg
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

I can see the rough spot, it looks to be an area on your glass that maybe wasn't clean enough. I've noticed when that happens the plastic tends to bubble up in that area due to bad adhesion to the glass. Not sure though. All in all, your test print looks really good from what I can see from the picture. A question though, I think I remember you from another post, weren't you just having an issue where you were too low just between the Y and Z tower at the outer perimeter but fine between the other towers? Seems like that would just be fixed with the proper tower rotation. either way I'm curious to compare dimensions between us for standard steps and poor calibration and modified steps and low deviation calibration. ***EDIT*** I see you updated your photos. That looks like your nozzle was too high in those areas.
Toddwagner
Prints-a-lot
Posts: 30
Joined: Wed Jun 21, 2017 8:55 pm

Re: High spots between towers.

Post by Toddwagner »

*Sorry to piggyback off your post here, but I think this would be helpful to other people who may be having similar issues as us*
I agree that the nozzle is too high on that side, but now my issue is how to fix that without lowering the nozzle on the opposite side as well (causing my initial problem of the nozzle being to close to the bed near the y tower) :? Are there any tips anyone can give about this problem?
Toddwagner
Prints-a-lot
Posts: 30
Joined: Wed Jun 21, 2017 8:55 pm

Re: High spots between towers.

Post by Toddwagner »

ALSO!!! This should help in the journey to dimensional accuracy! http://forum.seemecnc.com/viewtopic.php ... l+accuracy
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

I am actually printing that right now!!! :lol:

I took a step backwards sort of. I re-flashed my firmware to reset everything. Then I began to increase my steps/mm by .1 increments. After each iteration I would G28 followed by 3 G29 commands. I would then run the Rostockwizardprobe gcode and plug those numbers into https://www.easycalculation.com/statist ... iation.php to calculate deviation. I actually ended up at 80.95 steps /mm before my deviation began to rise again. The deviation got to around .04423. After that I then plugged those numbers into the http://www.escher3d.com/pages/wizards/wizarddelta.php calculator. This got me all the way down to .02718 deviation. I am now going to run the test prints from http://forum.seemecnc.com/viewtopic.php ... l+accuracy . Then make whatever adjustements are required from 626Pilot's instructions.
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

So, my measured 50mm test piece was 50.2mm give or take. So doing the calculations I ended up with a new 292.2 DR length. However this through off my deviation. So I slowly turned down my steps/mm till I ended up at 80.5 steps/mm. This gave me the lowest deviation(.02946) with my new DR length. I printed another test piece and I was getting around 50.08mm to 50.02mm. So I decided to try my full bed test print circle and it was a huge failure..... (See attachments). I had always thought I could see a little nozzle life going away from the X tower as opposed to approaching the x tower. I modified the Rostock probing code to test from bed center to incremental spots in a straight line to the X tower and back to the center through those same spots. This is what I was getting: -.01,-.03,-.02,-.02,.01,-.01,-.02,(first step heading away from X tower).20,.12,.06,.04,0,-.01. Sure enough, every time I test, the nozzle is lifting when it returns back and then gradually drops back to normal offset as it moves. I reset everything back to stock setting in firmware and after 3 iterations of G29, its still the same, so it has nothing to do with modified DR or steps. I will test all 3 towers, but I seem to have some backlash or something else wrong. Not sure what though. Also, I tested by cutting the distance to halfway to the x tower and returning. The lift seems to be greatly reduced but somewhat still there (-.01,-.01,.01,.01,.04,.06,.03,.01,0).
Attachments
Wierd_Pattern.jpg
Test_Circle.jpg
Gdot
Printmaster!
Posts: 41
Joined: Mon Jun 05, 2017 1:21 pm

Re: High spots between towers.

Post by Gdot »

Well, checked belt tension and adjusted, checked pulley alignment and adjusted, tried higher current settings in firmware, ...no luck. I even swapped out the x tower servo since I had a spare from another project....no luck. Every time the z offset is fairly close from bed center to X tower ,then going from x tower back to bed center, it jumps up by as much as .12 and then settles back down by the time it gets to bed center. -.01,-.02,-.03,-.03,-.01,-.01,-.03,-.07,.15,.12,.08,.04,.02,0,0.

At a loss.....
Post Reply

Return to “RostockMAX v3”