Hyperfine bed leveling?  

Re: Hyperfine bed leveling?

This is awesome PJR! i think you're about to save my sanity.

so of to the questions then..
I'm using prusa' version of slic3r and haven't changed the startup gcode.
...
so basically I figure i get a propper firsta layger on a 20x20square at the center and then run your stl for all the 9 points. but how should i think while doing so?
and then set the A-H points with the difference between lets say center point 100 and A point 105 which would mean that A=5?

PAX, you have to make the change to the firmware, compile and upload to the printer. Omit the G80 parameters (or set them all to zero).

At this point, I printed all 9 squares and approximated the difference values for the 8 perimeter squares based on the centre square. Then I moved the "Live Z Adjust" to what I though it should be and subtracted the Live Z difference from the parameters. Sorry if it sounds complicated.

OK, so I printed the squares, estimated that point A would need to be -100 from the centre square, and set parameter A-100. then I did the same for the other 7.

I then decided that the centre square needed an adjustment of -10, so I changed live Z by -10 and then subtracted -10 (added 10) to each of the parameters, so A became -90.

Printed all the squares again and made some fine adjustments; printed again and they were all thereabouts perfect.

Peter.

Re: Hyperfine bed leveling?

thanks Peter!
might not have been clear from my post but yeah i got that i need to flash the firmware first.
While were at the topic is there anything else that "needs" to be done to the stock firmware mainly thinking on if it is necessary to change the -50 to 50 for F/B/L/R to -100 to 100?

Re: Hyperfine bed leveling?

While were at the topic is there anything else that "needs" to be done to the stock firmware mainly thinking on if it is necessary to change the -50 to 50 for F/B/L/R to -100 to 100?

Firstly, if you try my "fix" F/B/R/L will not work. But my "fix" implements a correction of up to 200 microns at each point.

If you only want to allow larger corrections then the existing G80 F/B/L/R allows up to 101 micron correction which can easily be changed in the firmware if required. However, the stored value limits may not be changeable; I haven't checked yet if it would do any damage elsewhere.

Peter

Re: Hyperfine bed leveling?

Firstly, if you try my "fix" F/B/R/L will not work. But my "fix" implements a correction of up to 200 microns at each point.
Peter

yeah i figured that F/B/R/L would not work (kinda don't for me anyway) just figured that the correction with your method might use the same range as F/B/R/L but for each point. Glad to see that's not the case although i don't think i'll need all 200 microns but 100 might just fall short. so thanks for clarifying that!

Re: Hyperfine bed leveling?

PAX

Just a thought - the new "probe temp compensation" in 3.0.11 may halp. Probably worth trying before you go down the path of making significant changes to the firmware.

Peter

Re: Hyperfine bed leveling?

PJR
Oh really. I have not heard anything about this, I'll see if i can find anything before i get rolling with your fix

Re: Hyperfine bed leveling?

PAX

Just a thought - the new "probe temp compensation" in 3.0.11 may halp. Probably worth trying before you go down the path of making significant changes to the firmware.

Peter

i'm not very convinced that this function will deliver useful results with the actual printer setup. the thing is: in normal operation (while G80 meshbed levelling for example) you don't know the exact temperature of the probe. it would be another thing if there was an additional temperature sensor at the probe, so that you can do an active temperature compensation. things may look different inside an enclosure... i guess that's why you can switch it on and off (default) inside the menu.
i think as soon as the 3.0.11 final will be released, i kick out the sub-menu for the meshbed correction and go with your G80 amendment, which is - in my mind - a big step forward to achieve the best levelling that you can get with our printer !

Re: Hyperfine bed leveling?

I did this today and got my bed more accurate then ever before. The back left corner was way off. Maxed out to -200 and finally added another washer to the bed (total of 3). Not sure why the physical aspect is so off but now with PJR's method, my bed is nearly perfectly even when doing 9 point calibration prints!!!

Re: Hyperfine bed leveling?

Well, here's the thing. I do not need to use "temperature compensation" for the probe, so it's no something I am really interested in. But I had to make mention of that feature to ensure everyone is fully aware.

@Cory - I have never been convinced about the benefits of adding washers under the bed. If your printer passes XYZ calibration then the probe can "see" all the calibration points. If that calibration does not pass then add washers as the chassis may be bent/twisted.

When the probe can "see" all 9 points the only thing to do is to adjust for probing sensitivity - level correction. Washers will not change this.

Peter

Re: Hyperfine bed leveling?

So my live Z is around 350. If the back left corner needs a high level of correction, should I move the pinda closer or farther from the bed to adjust sensitive. After I remove the washers of course.

Also, have you tried this with the new temp calibration? Does it make everything off? I have the newer firmware and may play with it when I have more time between prints.

Re: Hyperfine bed leveling?

So my live Z is around 350. If the back left corner needs a high level of correction

Cory, I am guessing here, but I think you mean that Live Z is currently at -350? Also what do you mean by "high level of correction" (more or less)?

If I assume correctly, then your Live Z is quite "average". If your back left corner is not sticking/printing correctly, then it needs lowering, or making more negative, so you can use down to -201 microns with the amended code.

Also, have you tried this with the new temp calibration?

No I haven't and there is a very good reason. I am a "creature of habit". When I find something that works, I try very hard to be habitual with it. When starting a print, I have always ensured that the probe is at least 100mm above the bed in the same position. So when I print the calibration squares, I will get the same probing results as with any other print I do.

The only time this changed was last summer when it was 15 degrees (C) hotter in my office than usual. At that time I did have to change my Live Z values a little, but that was for the whole bed. I didn't have to change the side correction.

Peter

Re: Hyperfine bed leveling?

Awww yes sorry, I meant -350. As of now my H point is -200 after I increased the max. Still not perfect so may need to re flash he firmware with a higher limit to adjust. Also noticed my zip tie holding cords from extrude had the lock piece too low and it was hitting the bed connection cover.

As of now I have my starting Gcode in simplify3d to add my calibration point adjustments, and have extruded at 100mm above home point, warm up bed, then do calibration, and finally warm up extruded.

Re: Hyperfine bed leveling?

Cory

To save a minute or two, you can heat the extruder to 170 while the bed is heating. I also then set the bed to print temp during the 9-point calibration; don't get any oozing during cal.

Peter

Re: Hyperfine bed leveling?

I believe it's been said before that the bed should be at print temp prior to calibration since the heat changes the measurement? Or is it simply the probe being warmed up prior to calibration?

Re: Hyperfine bed leveling?

Cory

M140 S<BED> ; Heat the bed to print temp - don't wait
M104 S170 ; Head extruder to 170 - don't wait
M190 S<BED> ; Heat the bed and wait
M109 S170 ; Head extruder to 170 and wait
G28 W ; Home
M104 S<TEMP> ; Heat extruder to print temp - don't wait
G80 A-90 B50 C125 D75 E100 F40 G0 H-70 ; 9-point cal with correction
M109 S<TEMP> ; Heat extruder to print temp and wait

Peter

Re: Hyperfine bed leveling?

So I've managed to get all points nearly perfect except one with PJR's update.

G spot is still significantly too high. I have attempted multiple time to adjust settings and set max bed correction to 300. Even with a G-300 its still needs at least -50 more. Considering it is so large over the others, I assume its another issue.

Could the calibration spot on the bed be malfunctioning? My regular Live Z is -305 I believe which is in normal range so not sure what adjusting PINDA would do. Plus I don't want to recalibrate every spot again.

The wires for the bed temp run near that spot, could that be the cause?

Re: Hyperfine bed leveling?

So I've managed to get all points nearly perfect except one with PJR's update.

G spot is still significantly too high. I have attempted multiple time to adjust settings and set max bed correction to 300. Even with a G-300 its still needs at least -50 more. Considering it is so large over the others, I assume its another issue.

Could the calibration spot on the bed be malfunctioning? My regular Live Z is -305 I believe which is in normal range so not sure what adjusting PINDA would do. Plus I don't want to recalibrate every spot again.

The wires for the bed temp run near that spot, could that be the cause?

Interesting. I have the exact same problem. Every other spot is perfect except for G.

Re: Hyperfine bed leveling?

I increased my max bed tuning to 500 and its getting closer but still not enough.

Re: Hyperfine bed leveling?

PJR your a savior, I just finished my thread on the Mechanical Calibration process and findings, and after hours of printing Jeffs square and Mesh bed Compensation tweaking came to understand that the 4 points are just not enough.

So obviously the 9 points theory as you mentioned is the solution.

Here's the final result of Jeffs square fine tuned to the max in all aspects mechanically and the default mesh bed correction minus your theory.

As can be seen in the picture it's self explanatory.

Will read up on your thread and fire questions away as I try to figure it out. :ugeek:

Re: Hyperfine bed leveling?

OK, this is getting interesting. It would be good to know what values people are using for their correction.

I have 2 printers and here are my values:

Printer 1: G80 A-10 B50 C150 D70 E50 F-15 G-190 H-25
Printer 2: G80 A-90 B50 C125 D75 E100 F40 G0 H-70

Printer 1 has some weird things happening. The other day I was printing a large model (220 x 190mm). On the first layer the perimeters printed fine and part-way through the diagonal infill, the nozzle lowered by about 100 microns and stayed low. I have not as yet been able to reproduce this.

Peter