P003A HE280 [Solved?] Auto Leveling Issue/Bed Leveling Issue/Calibration

[miaviator]

UPDATE: The current development firmware branch: https://github.com/seemecnc/Firmware/tree/dev fixes the bed leveling on our test Max V3.



We had a redditor bring in his V3 to the depot for repairs. The printer would not level correctly and was useless for over a month.

Shameless self plug, we offer maintenance and materials contracts for printers as well as free depot repairs.

We looked into the issue before starting work and found a lot of ideas and no solutions, please point to posts if you have found others who have solved this issue: (http://forum.seemecnc.com/search.php?ke ... d+leveling) After walking through some test prints (see photos) we identified several issues with the build and found a solution for the probing issue which was repeatable on our V3 (after breaking it intentionally a bit).

Beginner level video instructions on how to correct this: https://www.youtube.com/watch?v=DxGT18rDKUM

The following 3 Items seem to have cleared up the issue:
#1 Added a Ferrite bead on both ends of accelerometer wires, at the HE280 and at the Rambo (See photos)
#2 Wiring to HE280 plug, removed, cut wire ends flush and pressed them in properly this was for the power and probe wires.
#3 Ran the V3 Probing code repeatedly until the first probe attempt doesn't crash at the X tower. This we found needs to be all of the time. If you get a crash when running the probing code, let it finish and then run it again until it runs the first tower without crashing (generally 2-3 iterations).
* We ran the probe with the bed and nozzle heated not cold. Heaters were killed just prior to probing.
**We saw no difference in probe results or first layer results with 0 3 or 6 bed clips. It didn't seem to change the bed shape. Obviously over time having pressure points on heated glass will warp the glass slightly so it be a good idea to tape the bed in place if you are concerned.

If you have questions/comments/internet hate let me know.
FLIR Photos taken to check wiring/board/motor/cooling temps: http://imgur.com/a/vArhl


If you landed here because of a similar issue, check the following items as well. It may help.

1. Fixed the black heater wire leading to the rambo: wire was (loose) (Assembly error)
2. Moved EZR feed gear: EZR extruder gear was to far in towards the motor.
3. Replaced PFTE Tube: PFTE tube had too much friction (Parts Issue)
4. Added a Ferrite bead on both ends of accelerometer wires See photos
5. Wiring to HE280 plug, removed, cut wire ends flush and pressed them in properly.
6. Advised that printing on Tape or PVA will prevent the warping/curling issues experienced.
7. Advised that PID Tuning would help prints.
8. Ran the V3 Probing code repeatedly until the first probe attempt doesn't crash at the X tower.

Results Reported by ComJak:

4 days later and the printer is still producing quality first layers. Now I'm just optimizing retracts and stuff.

But yes, in terms of the issues that I had with unlevel nozzle heights, the printer is performing 100%.

Huge thank you to miaviator for taking the time to help out. Guy clearly knows his stuff.

**** Development Tests for new firmware ****

*** proceed with caution here *** This is work in progress. You should keep in mind that you might need to return to the master if this causes your printer to not work. Remember to clear EEPROM if uploading this. AND clear EEPROM before returning to the master or installing a the new master firmware after we commit the final revisions.

This is only tested on a handful of RMAX's and Eris (Erii ?) TODAY at SeeMeCNC. But why not share the bleeding edge.

The development (dev) branch on github updated tonight regarding probing: USE ONLY G29 to calibrate the printer. G29 does it all !! You don't execute anything else. Please post some feedback while we continue to further improve the firmware. After years of wanting a dedicated embedded programmer in-house, I brought on Ryan Rittenhouse, who's very good at what he does.

seemecnc_repetier0922-DEVelopment.JPG

[mhackney]

The interesting thing, to me, in this post is the Ferrite bead. Can you describe or show (photos) of what the problem was when the owner brought this in? Many of us have been battling a persistent leveling issue that appears in the same place for all of us. I have researched this at great lengths including rotating the HE280 hot end to eliminate position of the accelerometer as the culprit, etc. I have also completely disassembled the bed, inspected, reassembled with no effect. I've also used 3 or 6 clamps with no effect (and you need to be cognizant that with the force the probe hits the bed, if you do not have these all adjusted to support the bed, it will tilt slightly as there is not a support clamp directly underneath the probe points at the base of the towers). On my printer, the issue came and went then came again. I mostly have it. When I test the end stop offsets manually after a calibration I find that they are indeed incorrect, indicating a probing problem. My hypothesis is/was flex somewhere but an electrical interference issue can not be ruled out and the power supply is mounted underneath the Z tower.

We'd appreciate hearing more about what the symptoms were on this printer. I'm also curious on why this one is crashing so frequently probing? I assume what you mean is that the effector actually tilts - a mechanical "crash" as opposed to the RAMBo crashing?

Thanks!
Michael

[mhackney]

I'd be interested in the STL or gcode for that test print too - the large zig zag one. That's a great idea.

[miaviator]

The interesting thing, to me, in this post is the Ferrite bead.

If I have the chance to repair another V3 for this issue I will start with the Ferrite beads and then have some better observational evidence what the effect is.

Can you describe or show (photos) of what the problem was when the owner brought this in?

I can ask the owner if he has photos, lacking that there are photos attached of our initial print and the 13 point calibration readings which are written on the bed in black sharpie. This is the post he provided which "described his issue" http://forum.seemecnc.com/viewtopic.php?f=37&t=10934

Many of us have been battling a persistent leveling issue that appears in the same place for all of us. I have researched this at great lengths including rotating the HE280 hot end to eliminate position of the accelerometer as the culprit, etc. I have also completely disassembled the bed, inspected, reassembled with no effect.

I had read through a ton of your posts and the ideas of rotating the hotend etc I was not very confident that we would be able to solve the issue as you seem quite a bit more knowledgeable on these printers and willing to go to greater lengths to resolve the issue. The ideas we had came more from the think like a child camp. The wire and bowden tube being on the side can add weight and EMI sucks for sensors (We had to wrap our FSR's to get rid of EMI). Also DIY assembly by inexperienced (Apologies) users can cause weird issues. The symptoms were the same as described by you and others the tight spot between the Z & Y towers?

I've also used 3 or 6 clamps with no effect (and you need to be cognizant that with the force the probe hits the bed, if you do not have these all adjusted to support the bed, it will tilt slightly as there is not a support clamp directly underneath the probe points at the base of the towers). On my printer, the issue came and went then came again. I mostly have it. When I test the end stop offsets manually after a calibration I find that they are indeed incorrect, indicating a probing problem.

We initially did all probing with no clips holding the bed. Just masking tape. After a while we did clip the bed down with 3 and 6 stock clips. As I said we did not see much difference in the probing results. If you don't use clips you need to ensure you have the bed secured to prevent the force of the probe from offsetting the bed when it hits. I remember Steve talking about how they had to probe fast enough for the accelerometer to work and it makes sense so the tapping is not avoidable it just needs to be accounted for.

My hypothesis is/was flex somewhere but an electrical interference issue can not be ruled out and the power supply is mounted underneath the Z tower.

We'd appreciate hearing more about what the symptoms were on this printer. I'm also curious on why this one is crashing so frequently probing? I assume what you mean is that the effector actually tilts - a mechanical "crash" as opposed to the RAMBo crashing?

! To clarify I am speaking of the probe hitting and tilting on the first probe at the X tower. Mechanical. Not about the RAMBo or script crashing. Thank you for clarifying.

My V3 also has mechanical crash issues on probing as mentioned usually running the code 3 times gets a clean probe.

[miaviator]

I'd be interested in the STL or gcode for that test print too - the large zig zag one. That's a great idea.

You will laugh. It's an extruded circle. Of any size. It's all in the slicing.

We used to make the print in matter control or cura or slic3r not sure which, there was an option for concentric infill and you could make a center out spiral to test these thing as well as this line version.

https://puu.sh/s0P5w/160bca34b0.png

https://puu.sh/s0P6M/75782e8dca.png

https://puu.sh/s0P9w/985a010c44.png

EDIT: Factory FIle removed see below.

[mhackney]

Thanks for the update. Interestingly, the 12+ printers (including mine) with a leveling issue all have a thin layer near the X tower and thick near the Y-Z. The opposite of ComJak's problem. His is the only report I've seen that is opposite.

[mhackney]

Thanks! Ha, that is a funny and makes sense.

[miaviator]

Thanks for the update. Interestingly, the 12+ printers (including mine) with a leveling issue all have a thin layer near the X tower and thick near the Y-Z. The opposite of ComJak's problem. His is the only report I've seen that is opposite.

That is what we saw after our initial testing. The extruder was off the bed at X/XZ and Squishing at Y/YZ.

[mhackney]

Careful, sharing a factory file is not permitted by the S3D license agreement.

[DerStig]

Thanks for the update. Interestingly, the 12+ printers (including mine) with a leveling issue all have a thin layer near the X tower and thick near the Y-Z. The opposite of ComJak's problem. His is the only report I've seen that is opposite.

That is what we saw after our initial testing. The extruder was off the bed at X/XZ and Squishing at Y/YZ.

That's my EXACT problem

[mhackney]

Yup, you are on my list!

I am going to pick up some Ferrite beads tomorrow and see what happens. I've spent a LOT of time investigating mechanical reasons for this but have come up empty. Perhaps miaviator is on to something.

[miaviator]

Careful, sharing a factory file is not permitted by the S3D license agreement.

#Sarcasm or #TrueStory?

[mhackney]

#TrueStory

[bvandiepenbos]

@Michael,
have you tried moving the probe points in 15 mm or so ( I used 127 MM radius vs. OEM 142.68), to see if probing is more reliable?
I am wondering if probing at the extreme travel very close to tower is any of the issue.
Also try probing opposite of towers and see if reliability changes.
You can change probe locations in eeprom by altering the values for...
Z-probe X1, Z-probe Y1
Z-probe X2, Z-probe Y2
Z-probe X3, Z-probe Y3
(flip the sign of the Y values)
See my other post for more detail
http://forum.seemecnc.com/viewtopic.php?f=36&t=11052

I should note, mine is a V2 with HE280. That means bed is not the same as your V3.
I have not had any failed probing unless a booger of hard plastic on tip of nozzle causes platform to tilt when it hits.

[mhackney]

Hi Brian, yes I have and I published the points so others could try them. No effect and I still got the occasional mechanical crash although not as frequently. I've reordered the probe sequence, I've rotated the probe sequence, and they all have no effect. I've swapped the X and Y stepper and the X and Z stepper (and endstops).

[mhackney]

You can also use David Crocker's online wizard to calculate the probe points, I showed how to do that in my V3 build thread a couple of weeks ago.

[morgandc]

Two questions -

1. If this was an interference issue, why would the level be so consistently off in the same direction?
2. Does a proper manual leveling work? I tried (first time ever) and it didn't seem to be better.

Thanks.
-Dwight

[mhackney]

1. if it has something to do with EMF from the power supply mounted in the base, then it could be consistency off in the same direction
2. yes manual calibration should work. That's what I'm doing in fact, I know precisely how far off the end stop offsets are after auto calibration and I go in and manually edit those (making Z and Y bigger and Y is a but bigger than Z) and can print a nice consistently thin layer on the entire bed.

[Xenocrates]

Careful, sharing a factory file is not permitted by the S3D license agreement.

#Sarcasm or #TrueStory?

S3D's license agreement forbids sharing any and all output from the program, as noted in the two paragraphs:

All copies of the Software made by You as well as any output generated by the Software shall bear the proprietary markings (be they patent, copyright, trademark or trade secret) of Simplify3D borne by the copies transmitted or provided by Simplify3D or third party to You and in the same form and location as the original. All rights not expressly granted herein are reserved.

and

Except for the use expressly authorized, You agree not to sell, transfer, license, modify, copy, reproduce, distribute, transmit, display, perform, publish, or create derivative works from such materials.

You can see even more of the weasely language that I take issue to here:

In short, S3D owns your settings and factory files, and also the crash logs and G-code. You aren't allowed to share or distribute them. Their CS reps refuse to clarify that it doesn't include Gcode or factory files, even when asked point blank. They reserve the right to change the EULA at any time, without need to give notice, and the use of the software after the effective date of the new EULA counts as acceptance (This is explicitly spelled out too), even if you haven't looked at the new EULA.

I feel it's wrong to give people money and then have them take rights from me below the table, and so I've personally decided to not use S3D (Even if I don't have to pay for it). For the same sort of configurable in slicer supports I could use Craftware, or Meshmixer if I feel like working them out ahead of time. I can use KISSlicer for excellent pathing (arguably better than S3D, and with more useful settings exposed, plus a lower rate for the pro version and an education and bulk discount). Best of all, I don't have to give money to a bunch of rats who would happily use a shrinkwrap license to take ownership (or a least patent rights) on your newborn child if they offered a physical product.

And don't take this as annoyance at not being given something for free. I have nothing against paying for software, especially good specialist software. I own probably twenty licenses for bits of useful software, usually purchased at the top levels available, such as a lifetime rather than subscription license, usually with multiple seats. But I buy software that is useful, effective, well made, and from honest developers (Or a reasonable facsimile thereof).

S3D is a dangerous thing with it's license as it stands to the community, especially as they attempt to push it with kickstarters, with CAD packages, with printers, and on so many blogs and youtube channels. It seems they want to promote their product (Which is fine), and are willing to push a near monoculture to do it, with a poisonous EULA that no one will read (Which isn't).

Take it with a grain of salt, of course. It may be they have no bad intentions and merely want to (Very vigorously) defend any IP rights they have in what they consider a superior product. But that's more benefit than I'm willing to give them from doubt.

[ComJak]

4 days later and the printer is still producing quality first layers. Now I'm just optimizing retracts and stuff.

But yes, in terms of the issues that I had with unlevel nozzle heights, the printer is performing 100%.

Huge thank you to miaviator for taking the time to help out. Guy clearly knows his stuff.

[PhantamaroK]

I'm still really struggling with this bed leveling/calibration issue. I have tried adjusting temps, several calibrations, decreasing the probe radius, using only 3 clips, using washers on the bed clips, etc. and it still seems like it's printing on a slant.

[img]http://i.imgur.com/vAtRU98.jpg[/img]

See how it's super smushed on the right (Z/Y) and lifts off the bed on the left (X/Z)?

Can anyone offer me any more insight on this? I've already scoured the forum. At the very least, can someone give me a step-by-step of how to manually calibrate the v3? I apologize for my newbness; this is my first 3D printer.

[DerStig]

I'm still really struggling with this bed leveling/calibration issue. I have tried adjusting temps, several calibrations, decreasing the probe radius, using only 3 clips, using washers on the bed clips, etc. and it still seems like it's printing on a slant.


See how it's super smushed on the right (Z/Y) and lifts off the bed on the left (X/Z)?

Can anyone offer me any more insight on this? I've already scoured the forum. At the very least, can someone give me a step-by-step of how to manually calibrate the v3? I apologize for my newbness; this is my first 3D printer.

This is the same problem I have X/Z tower lift Y/Z good to smashed

[crocky]

Yup, you are on my list!

I am going to pick up some Ferrite beads tomorrow and see what happens. I've spent a LOT of time investigating mechanical reasons for this but have come up empty. Perhaps miaviator is on to something.

Did you get the ferrite beads yet and if you have did they make a difference?

[Qdeathstar]

simplify3D's EULA is standard for any software company that is not open source. Not a big deal. Also, it is unlikely that a court would uphold all of the EULAs terms, especially the one about agreeing to automatically agree to any new terms in the EULA without expressly agreeing to them.

[miaviator]

I'm still really struggling with this bed leveling/calibration issue. I have tried adjusting temps, several calibrations, decreasing the probe radius, using only 3 clips, using washers on the bed clips, etc. and it still seems like it's printing on a slant.

See how it's super smushed on the right (Z/Y) and lifts off the bed on the left (X/Z)?

Can anyone offer me any more insight on this? I've already scoured the forum. At the very least, can someone give me a step-by-step of how to manually calibrate the v3? I apologize for my newbness; this is my first 3D printer.

I'm working on a step by step video guide for this. I'll include referral free links to ferrite beads which can be purchased. Should be done this weekend. Lacking that if you read the first post it goes over what we believe the issue is when this happens. Until then printing smaller footprint items or using a raft should be fine?