R2 v2 Bed S3D First Layer Gap Issues

My R2 v1 bed pogo pins started to fail, and Robo quickly sent me a new v2 bed. I'm running my machine stock with their buildtak surface on the v2 bed. I should note, I am able to complete prints on my R2 - but the gaps in my first layer are something I didn't have prior to the bed swap. Here's what I've done so far:

• Auto-leveling has been disabled in Octoprint (replaced with a G28 vs the 2.0 firmware G36).
• v2 bed has manually leveled/fine-tune leveled.
• z-offset currrently set to -6.52

Originally, I tweaked a R2 S3D profile from @Ed Ferguson dated 2017-08-27 (linked here)
I'm using eSun PLA+ filament. I used to run 215/60 on the previous v1 PEI bed with Ed's profile - but those settings v2 Buildtak bed made parts were near impossible to remove!

Lowering my hot end to 205 and bed to 45 has made print removal much easier, and hasn't had any effect on first layer gaps, and has reduced stringing too (retraction distance 1.50mm, speed 2400mm/min).

I've included a photo gallery with details/settings in the comments - please feel free to leave comments and refer to them (I'll update the gallery as needed): https://photos.app.goo.gl/9y4mpd4eDEgXTGky5

Using these guides (visual Ultimaker and Simplify3D Print Quality) and previous forum discussions here, I believe I likely have the following issue(s):

• Under extruding
• Extrusion Multiplier tweak needed (never touched that setting)
• first layer width/height
• print temp
• z-offset closer (hesitant to touch this - when lowered to -6.50 or lower, prints SUPER hard to remove)

I know to want a "perfect" first layer might be akin to madness, but I know the gaps I have can be sealed up a bit more. I feel like my Simplify3D profile isn't translating variables correctly, or I'm missing something else that should be obvious. Been toying with this issue for about a week, and figured I should seek some guidance from you all I've also included my current S3D profile.

 

While I certainly could be wrong (and I likely am) - I thought the value saved in the EEPROM was actually kept/honored. Making any adjustments and manually saving the value in the EEPROM definitely has an impact in the squish factor. I thought that was the point of having the manual level and fine-tune leveling wizards - to help you calculate a value used when you print even if the auto-level is off

I started out with -6.54 before lowering to -6.52.

I will try -6.50 again with the profile settings I uploaded, and send the image to the gallery.

 

While we wait for the others to chime in - I manually adjusted the eeprom from -6.52 down to -6.50 to -6.48. I saw NO visible changes in the print (photo gallery updated), unless the adjustment of .04 was taking you a bit too literal on the tiny bit too high

It would appear you are indeed correct, so I'll have to go back through the manual leveling/fine tune wizard to help adjust my z-offset. While I could use the thumbscrews - it took me over an hours to get it leveled out nicely as it is, so I'll run those wizards for now (open to other suggestions).

 

Mark, my apologies, but I am unsure how to accomplish that without using the manual and fine-tuning wizards. I can and have used the thumbscrews on my R2 to help level out the bed as best as possibly while using the fine-tune wizard with both circles and lines, and I know I can use the motor controls to raise and lower the bed. Am I missing something obvious that can help save an offset without these wizards?

I encountered severe issues with the auto-leveling change after the 2.0 firmware upgrade. Jerry from Robo had said once I had the bed dialed in (I mentioned using the manual bed calibration and fine-tine wizards) to replace the G36 and G29 with just the G28 in Octoprint. I did exactly that - and my Octoprint still only has the G28 in the area where the previous G36 and G29 commands were.

I've included an updated R2 circle bed level image in the gallery after adjusting my z-offset using fine tune wizard with both circles/lines, and now and my z-offset is now "calibrated" at -6.34 (why the "huge" difference from -6.52 I have no idea).

 

@mark tomlinson you sir, are a fine genius and gentleman. I think I might have overdone my squish with the -6.34, but if you take a look at the latest image in the gallery - I think we've got a winner if I increase the distance just a little bit between my nozzle and bed.

I'm still curious to know about other ways to level the bed, and why there appears to be no more value adjustments with any value saved to the eeprom. I'm guessing whatever value gets saved during those wizards gets buried somewhere in Octoprint that I cannot see, but the R2/Octoprint honor that value even if the G28 has removed the bed leveling routine/wizard.

 

Thanks Mark, that will make for some good reading. After a little more digging, I believe this is where the Marlin Plugin values are being stored/recorded: https://help.robo3d.com/hc/en-us/ar...tting-your-z-offset-through-the-web-dashboard

My R2 is at the office, so I won't be able to verify until Thursday, but perhaps I can test that theory on my C2. As far as I can tell from skimming your resources and my Octoprint startup code, the use of the G28 instead of a G35/36/29 combination should have auto-leveling disabled. Presumably it might retrieve the z-offset from this stored area, since no leveling occurs when I submit a print.

I'd still be curious to hear from Geof or Waldo about the z-offset/eeprom storage and it being used or not. My slight tweaking of the epprom z-offset value from -6.54 to -6.48 clearly was too small. I need to lower the bed a bit more, maybe and lower the left half my bed just a smidge as that side is a bit more smooshed.

Thanks again for helping me determine my z-offset was the culprit - just goes to show how often that can be the crux of getting everything else setup correctly

 

@Geof My R2 is at the office, but I need to calibrate my C2 for proper first level as well. Would I be correct that the steps would be similar?

I can run the M502 and M500 on my C2 in a little while. I can definitely confirm that if you use the fine tune wizard on either R2/C2, the autolevel does run - even if the G36/G29 has been removed and only the G28 remains.

I was trying to explain yesterday (albeit poorly) - without those various leveling wizards - I honestly have no idea how to help level my bed and run tests that I can use to make small adjustments and set my offset. (I did read some of the various Marlin bed leveling types Mark linked me to)

My apologies for what I'm sure seems like a silly questions, but how does one go about setting and testing a z-offset without those wizards?

 

Thank you all for the responses. Let's see if I got any of this right

• Remove auto-leveling code from Octoprint and other slicers
• Mechanically/manually level the bed at the 4-corners as best one can (manual bed calibration with paper, or perhaps use a small torpedo level at each corner)
• Home the Bed (G28)
• Run Z-offset wizard (fairly sure this issues a G28 /runs an auto-level)
• Print single-layer item that covers extreme edges of bed.
• Manually adjust z-offset (adjust values in eeprom? adjust in Octoprint?)
• Test again with single-layer print.
• When happy with results, send M500 from Octoprint

@WheresWaldo although I understood your statement about only being able to make 0.20mm adjustments - that feels like a bit of a big jump. Why do the wizards offer to make adjustment as small as 0.01mm if they aren't actually able to make that adjustment? I believe even all of the various z-offset wizards run an auto-level once, but with the other leveling code stripped out of Octoprint (and S3D) - the auto-level never runs again.

 

My mistake on the torpedo level - totally make sense now that you've explained it the way you did And thanks for clarifying that 0.02mm adjustment. Do you make that adjustment in the eeprom setting, or manually go into Octoprint and set the z-offset by adjusting by 0.02 there?

 

I feel like a broken record - but thank you all @mark tomlinson @Geof and @WheresWaldo I ran through the steps we discussed, and after applying some small tape squares in the corners where the bed was lower in the rear - and I have a beautiful spiral bed level test (see gallery for updated picture). I issued the M500 command via Octoprint, and received the following output:

Send: M500
Recv: echo:Settings Stored (498 bytes; crc 36253)
Recv: ok

The machine survived a reboot, and printed out the circle level again just fine, so I'd rate this a huge success.

I'll run a few more test prints (found a few from thingiverse that will reach the edges with a single layer square and circle) just to be sure. Leveling the C2 bed with tape shims manually wasn't nearly as painful as I thought it would be

I'll report back on my R2 once I'm back in the office tomorrow. Thank you all again for your assistance - I'm convinced there's not much the three of you (and some other super helpful forum members) can't solve