Seems like you’re getting uncommonly large deviation from one measurement to the next. Did you fully calibrate everything else before running the vision encoder calibration? For instance, for each firmware update, it’s recommended to run them all again. Vision Encoder calibration should come last. If not, you’re likely wasting your time. Vision Encoder calibration is the icing on the cake. It’s not a substitution for the other calibrations.
Also, if you moved your H2D, o did maintenance of any kind that might affect the motion system, that too could cause differences in results.
I don’t think you do though, because half my point was that the machines as they are now are SOOOOO GOOOD that there isn’t a practical need most people would have for the vision encoder plate. It’s a 1% boost when you’re 98% the way there already.
They already are and you don’t need the vision encoder plate. Bambu machines in general have really pushed the quality forward, and helped establish 3d printing as more than just a toy to tinker with. This all happened way before the encoder board existed. It’s not just a limitation of the machine though, but people’s perception that needed to change.
Part of my focus as a designer has been to push beyond the limits of what we think 3d printing is and what is possible with it.
Part of the marketing for the plate is talking about getting injection modeling accuracy. I don’t think getting that level of accuracy is just about the vision encoder board. In my opinion, the filament is going to matter too, because there’s a number of factors there that could throw it off enough to negate the benefits from the encoder board.
I’ve seen those issues before when working to make things with tight tolerances. ABS vs PETG, it’s not going to be the same exact fit even if it’s the same exact model.
So if we’re talking about getting the utmost accuracy in our print, then there’s more factors involved than just tossing on the vision encoder board. There’s other elements that need to be considered and fine tuned too, to fully take advantage of that board.
That isn’t needed for the majority of things/people though, even where accuracy is required, because the base machine is already very good. That last few % of tweaking isn’t worth it for most people unless they’re perfection chasers. Perfection is the death of getting stuff done though.
I guess I view this differently than most people who bought it. I bought it incase I need to use it a year from now. Hopefully two or three years from now. I mean I didn’t retention the belts on the X1C when it was a week old. I’m confused at people running it on a brand new printer. Wouldn’t the H2D come completely tuned from the factory?
I think the UI message isn’t all that useful. It tells the maximum and average deviation prior to running the vision encoder plate, but it doesn’t state explicitly what level of accuracy it was able to achieve. Are we to assume that since it doesn’t declare failure it was able to achieve the <0.05mm accuracy called for in the spec?
It’s clear from the numbers that people have shared after running the vision encoder calibration that even fresh out of the box, there’s notable differences between different printers in the maximum and average deviations.
It claims a resolution of 0.005mm and an accuracy of 0.05mm, so I would think it could be more clear about what level of accuracy it was actually able to achieve.
All of this reminds me what a devil’s bargain black box systems are. I seriously think this will be my last one. The perpetual wall of silence is just too frustrating to endure long-term, at least for me. It shouldn’t be on us to figure it out and then explain it to others, all the while getting zero support in doing so. I’ve had it.
I had the same firmware and the printer didn’t move. The second the vision encoder arrived i put it in and stared the calibration and did that one afer another. The printer is on a very solid table has no ams on top or anything else on the table besides the poop bucket that didn’t increase in weight or changed placement. In all honesty my h2d is a let down as I’m having z banding and layer shift issues and I’m following the forum as of now hoping someone comes up with the solution that i see a lot of other people have… even now if i use the vision encoder the numbers are all over the place and it seems like the left hotend heater/termistor went out and they are sending me a replacemnt… so yeah idk so far i have mixed opinion about what the vision encoder brings to the table
It is mostly academic for me. I don’t need tha accuracy…but I bought a machine that can look at a 2d template and calibrate itself…and the template (again, 2d image, not holographic, not microscopic, nothing preventing facile reproduction) costs $100.
So I just figure why not investigate if this calibration can be done with a DIY template. I haven’t checked your math on the DPI of a printer and how accurate this needs to be, but frankly it’s a solvable problem… (remember, something had to print what got pasted on the official vision encoder…).
You can say it can’t be done if you like. I prefer ‘it hasn’t been done yet, it seems’.
I’m curious how you got 20x on 1200 DPI requirements…are you suggesting that the encoder print is at 24,000 DPI? It may be, I’m just curious.
It seems to me DPI (for resolution of small characters) is less of an issue, and rather how precisely the printer can place them across a sheet of paper reproducibly.
If within appropriate tolerances, then it only needs to be able to reproduce the QR looking codes to a readable degree, regardless of DPI.
I’m really just spitballing and not trying to argue or anything. I’m just trying to learn here.
It can be reproduced. Of that I am certain, but the equipment needed to do so is not something that is widely available.
Getting access to that equipment eventually costs more than the Vision Encoder Plate itself.
They already are and you don’t need the vision encoder plate. Bambu machines in general have really pushed the quality forward, and helped establish 3d printing as more than just a toy to tinker with. This all happened way before the encoder board existed. It’s not just a limitation of the machine though, but people’s perception that needed to change.