CR10 Smart Pro ABL mesh oddities

[dkerrivan]

Recovering from a bad flu/cold (despite all my shots this fall - oh well) and decided to see if my RPi 2B could talk to the CR10 Smart Pro with OctoPrint

Yes, it could, once I located a USB A to USB C cable (I didn’t feel like disabling the Creality wifi board in the unit, just yet. Kitchen table’s tied up with another hobby (photography prints and frames )

Went through the process of getting everything up to date, and loaded the Bed Mesh Visualizer plugin because I needed to confirm how far out of whack my bed is when heated. That was an interesting exercise.

These three shots over the past 2 hours show how variable this is, and what my deviation looks like - I’m reminded of a pringle…

mkay .98 variance… (yes, I’m rounding)

hmm. .93 variance…

eep - .88?

This is not encouraging to me to a) have nearly a full 1 mm difference, and then to see it shift down, over time (as I said, about 2 hours).

I’ve mentioned in other posts, that no matter how hard I try, I have poor adhesion for anything from the center line (150 mm mark on the X axis) to the right side of the bed (300 mm mark).

Looking at these mesh results, yup, we have a problem Houston. The fade height seems to be set to 2 mm, but that’s not helping, and my Z offset is -2.25 atm.

In the past, I’ve put some orange locktite on the screws the leveling knobs connect to so that I’d eliminate vibration issues backing things off. Doesn’t seem to be helping. I know the springs get fatigued, and these may just be ready to be retired. (My springs are the original ones and are now… 2 yrs. old?)

I had thought of trying to shim the PEI plate using layers of aluminum foil, and a bit of glue stick to hold them in place. I’m not sure I want to go the route of replacing the bed, heater mat, etc. If I’m going to drop that sort of coin, I’d rather look at a multi filament extruder system.

Within this community, what are people’s experience with the solid silcone spacers vs springs? esp. the (apparently) hardened ones?

It’s clear in my case, that even with ABL enabled, the CR10 Smart Pro’s not really doing a great job of adjusting for the deviations. I’m not even sure what’s considered acceptable at this point… but it would be nice to get it down to 0.0X variance.

I sense this may lead to yet another bit of insanity on my part to get the next thing dialed in on this machine…

[Loosenut]

At two years old the springs are probably done for, if you do a lot of printing. Can’t say much about silicon spacers but some people like them.

[dkerrivan]

So was doing more digging last night… I’m running 1.0.12 for the FW. I see on Creality’s site, that they have released .13, and .14. The release notes say, in what I can only call a vague resemblance to a bug fix, it fixes a problem with the ABL. No description of what that problem is. Reading the release notes for .14, and it’s soo poorly translated I honestly can’t say what they were trying to fix.

digging on Reddit suggests it’s a 50/50 shot if I do the update. Given how the printer doesn’t seem to actually use the mesh to compensate for the ~1 mm variance I see in the mesh visualizer, it’s a faster thing to try while mulling a new, stiffer set of springs or silcone spacers…

decisions, decisions, decisons

[Loosenut]

Stock springs are not that strong. I replace mine with the Amazon aftermarket ones and they where a lot stiffer. Unfortunately knowing which ones to buy is a crap shoot.

[dkerrivan]

yah. I noticed that too…
Just realized I am already at 1.0.14 of the firmware…

So, springs or spacers it is, and more locktite, or perhaps 4 of these : Bed leveling wheel lock for Creality CR10 Smart Pro by MAXPROBE for red metal alloy aluminium leveling wheel nuts

[Loosenut]

Locktite might be problematic. Springs or spacers would be better.

[dkerrivan]

That’s why I use the orange stuff - it’s in between the red and blue for strength, and has a bit of anti vibration resistance as a result. I’ve never found a version of Locktite that can’t be undone, but it sure helps keep things in place! (My recumbent trike rotors are all now secured with locktite on the screws after one ride where they suffered rapid disassembly while I was braking going down a hill towards an intersection that sees a lot of heavy truck traffic. Bit of a sphincter moment if you will)

[Loosenut]

The red stuff only comes loose with a lot of heat which might not be good for the equipment, blue comes loose with hand tools (so they claim). I have never seen the orange ones so, I don’t know? Thanks for the edumacation.

[kitedemon]

That is a lot. When I went though this with a sidewinder, after lot of back and forth I found out most ABL systems fail at a bit past nozzle diameters. That is from Antlabs people. No wonder you are having issues.

If the bed itself is warped springs will not really help. New bed might be the only help.

I got mine slightly better, not fixed mind you, by loosing all the screws/springs and then heating the bed to the max for 30 mins then letting it cool down to room temperature. it helped ‘relax’ the bed and sorta helped flatten it. Mine changed over time, so it got worse and worse. I eventually just replaced the whole bed.

[dkerrivan]

I’ve used the red stuff lots, and I’ve undone it with a basic picnic screw driver with no issues. It does keep things from vibrating loose, like nuts on a snowblower, but I never need heat to remove them or screws that I’ve set with it, but if I do have it happen, I’ll definitely try applying some heat - so thanks for that!

[dkerrivan]

Yah, that may be in the cards, but ouchie on the pocket book. I may, after tonight’s first layer test with the spacers, need a new PEI sheet. Manual 5 point (with heat) had me at -3.80 Z offset. Ran the 25 point. Then sent the print job to the printer via OctoPrint. I couldn’t get to the Zoffset adjustment fast enough (OctoPrint seems to make the touch screen less responsive)… it sounded like I ground the print head all around the board… eventually got it to -2.55 and it seems to be behaving. I’ll have to run a second test to make sure the first two squares being printed are good. I also noticed under a brighter flashlight some odd discoloration in the PEI sheet, which has only had water, dawn dish soap, and Elmer’s glue stick applied to it. So it may be subtly delaminating too. More $$.

Keeps me off the streets right?

(I do have more nozzles so that may get changed out too, which willl mean more leveling work…)

[dkerrivan]

Well, who’d a thunk it.

Cleaned up the PEI sheet - yup. small scoring happened… but it is still usable. Second 1st layer test run. See below, not terrible.

Bed Mesh (using bedmesh.com) since I’ve removed the G29 code from my profile to try and see if with the silicone spacers if we’ll stay dialed in or not (will be running several 1st layer tests today, and yes, I’m using M420 S1 Z3 to restore the mesh and fade out in 3mm)

I think I need need to move z-offset to -2.50, or perhaps -2.45 or so. tests will tell. The variance is down to .78 now, and I may still try my idea of shimming with some tinfoil. May. As I’ve said before, I know how much time you can waste chasing that last 5%…

[Loosenut]

@dkerrivan

I’ve undone it with a basic picnic screw driver with no issues

With that kind of luck you need to buy loto tickets. Loktite says it need 500 deg to loosen the red stuff.

[dkerrivan]

Well, I did manage to wrap the handle of a T-style spark plug wrench handle around my hand once - so maybe I just have more torque than they expect? (The manager at CT was a bit stunned when I returned it for a refund, and no, I don’t exercise my hands;/arms save for typing )

As for the lotto tickets, all I seem to do is win another free ticket.

[dkerrivan]

Second run of the 1st layer test at Z offset -2.50 went well. Test 3 with -2.45 under way now. Hoping to eliminate that bit of drag through on the left back corner of each square. I suppose I could dig around in Octoprint and find the z-hop settings… or drop it in my start g-code…This week’s efforts are to A) get OctoPrint dialed in, and B) resolve the ABL issues. B seems to be going well.

Was very puzzled with having to convert 500 mm/s to 30000 mm/min for Octoprint. Not sure why they have that unit when all the other slicers out there seem to like mm/s or mm/s^2.

[dkerrivan]

Update: z-offset -2.45 looks about right!
Need to setup my PrusaSlicer profile to use zhop I guess.
Now to look at some calibration checks after all this fun.
I did find one of my POM wheels has a nick in it - no clue how that happened. So, time to order some new ones.