Unsolved Mystery. Weird Z0 behavior around build perimeter.

[KAS]

It's not pointless. Tower angle and radius offset can both have an impact on that. The printer thinks its towers are somewhere they aren't, and that's what causes that problem. If it only knew where they really were, it would lay down a perfect 1st layer every time.

Even if they are all high opposite the towers? I guess I'll have to fool with it a bit, but even after reading this whole thread and others it just seems like people are randomly moving things around until it fits. Is there even any math involved or steps to take to fix this in some sort of logical manner? Either way I guess I'l mess with the rotations a bit.

The issue is how to check, change, and verify with the same repeatability in order to gauge if the changes made an acceptable difference. That's the hard part without probes and/or an exact method to follow when first getting into this "sport" of chasing down precision. I went through 3 carpenter squares only to realize I wasn't getting the towers square enough to the bed in a single axis. The manual doesn't explain how to correct any left/right lean to the build table at each tower. It wasn't until I bought a digital angle finder that I was able to measure and verify front to back and left to right lean on each tower. The next issue I'm working on is the tower rotation, how to measure and correct the actual tower twist is proving harder with manual loots and my lack of experience.

I've graduated from a piece of paper to feeler gauges to adjust the nozzle height. Although, you can get a good idea on how far each nozzle is off by taping a piece of paper down to the bed and run all 4 macros to press the nozzle to the paper. Measure the center of each towers nozzle impression to the center of the bed nozzle impression. In a perfect world those three numbers would be exact, and I believe that would show a perfect rotation??. It helps to have a little oozing filament to make the marks visible.

Anyway, what these folks are doing above are way above my level, I'll just continue to manually tweak and test to the best I can until then.

[Moeparker]

Ok so I printed the calibration thing from thingiverse. After some adjustments to the tower angles (within 0.5 degrees), checking belts, skates, etc.

This is what I have.

The areas between Z and X is high (not much adhesion) and the area between Z and Y is Low (squished plastic) . Areas at the towers seem to be ok. Any ideas on what my next step should be?

Still stumped. . .

At this point I think you can fix this with a rotation tweek in EPROM on the tower between the two, Z in this case. If you change the rotation value slightly it will go down on one side and up on the other side.

Like if you change rotation to be 0.5 degrees towards X, it will go up/down on X and down/up on Y. I can not remember which is which so change it slightly and test Z.

[bauf]

So i came over this video https://www.youtube.com/watch?v=E-92PSNINO0
And then i tried to print that calibration circle,and it was just a big mess.. Up until now i've only used the middle of the glass,but now i have several pieces to print. But only the prints closest to the center turns out ok.
So i tried to find the carriage offset value,but i cant find it in the configuration.h tab.
Tried to search for this line: #define CARRIAGE_HORIZONTAL_OFFSET, but nada.

[Jimustanguitar]

It's in there. It's the section right below the little character picture:


/** \brief Horizontal offset of the universal joints on the vertical carriages.
*/
#define CARRIAGE_HORIZONTAL_OFFSET 38.4 // Theoretical is 38.4, but actual may be slightly diff. due to plastic injection molded parts shirnkage



Maybe search for the word "shirnkage" I'm sure it's only in there once

[bauf]

huh.. I couldn't find it on the firmware i had on my pc, so i downloaded it from github and now it shows up
Thought it was supposed to be 35, so if it's already changed to 38.4, there's no need to change it,or?

[Jimustanguitar]

huh.. I couldn't find it on the firmware i had on my pc, so i downloaded it from github and now it shows up
Thought it was supposed to be 35, so if it's already changed to 38.4, there's no need to change it,or?

Back when this happened on page 5 of the thread (wow, that was 11 months ago!) the guys included the update into the newest firmware release on their Git, so that proposed fix has already been baked in.

[bauf]

oh lol.. guess i just need to keep calibrating and see
Thanks

[RollieRowland]

I have tried every possible solution to calibrate my V2, however, I decided to write a program(just javascript) which will CORRECT for the displacement between the towers, also if your bed is slanted, or your plate is warped, or if your towers are not square.

I had the area between the towers off by around 0.5-2mm, this program leveled my bed entirely giving a tolerance of less than +/- 0.015 millimeters. It has not had a lot of testing yet but it should still work properly.

My post goes into more detail on how it works, how to use it, etc. -
http://forum.seemecnc.com/viewtopic.php ... 698#p72789

If the link doesn't work just go to my page, it's my first post.

[Jimustanguitar]

I have tried every possible solution to calibrate my V2, however, I decided to write a program(just javascript) which will CORRECT for the displacement between the towers, also if your bed is slanted, or your plate is warped, or if your towers are not square.

I had the area between the towers off by around 0.5-2mm, this program leveled my bed entirely giving a tolerance of less than +/- 0.015 millimeters. It has not had a lot of testing yet but it should still work properly.

My post goes into more detail on how it works, how to use it, etc. -
http://forum.seemecnc.com/viewtopic.php ... 698#p72789

If the link doesn't work just go to my page, it's my first post.

Thanks for the xpost, it's definitely relevant. Can't wait to see your video.

BTW, what do you think of MHackney or 626Pilot's code? They've been working on the same thing for (I think) longer than you have. Just curious.

[lumpy_potato]

So 99% of this is right over my head, but I do have one question:

What Dial Indicator model are you guys using? The ones I've been able to find seem like the probe would be too short to go all the way to where it needs to be?

I'm doing the paper-calibration for now at least to get back to printing normally again, but it seems like the consensus for calibration is:

Do the towers X Y and Z

Do opposite points A B C

Calibrate Horizontal Radius / screw adjustments until everything is more or less 'correct'

Check that travel from XYZ to ABC is consistent

Mess with tower rotation and steps per mm

Do your Z = 0 adjustment and get to printing (for the most part)

Does that sound somewhat right?

[Jimustanguitar]

What Dial Indicator model are you guys using? The ones I've been able to find seem like the probe would be too short to go all the way to where it needs to be?

I've been using a cheap-o digital gauge in the TrickLaser holder.

http://www.harborfreight.com/1-inch-sae ... 93295.html
http://tricklaser.com/Pen-Tool-holder-f ... RM-PTH.htm

You take off the hot-end to mount up the gauge, otherwise you've got to rig up some other sort of long zprobe to go next to the hot-end.

[lumpy_potato]

Thanks - does the procedure I described more or less sound right? Right now I'm still doing the method prescribed in the build manual (at least in the short term)

[Jimustanguitar]

Thanks - does the procedure I described more or less sound right? Right now I'm still doing the method prescribed in the build manual (at least in the short term)

The build manual does an excellent job of calibrating the machine. Definitely do that first. If you're still having issues afterwards, then dig a little deeper.

[RollieRowland]

I have tried every possible solution to calibrate my V2, however, I decided to write a program(just javascript) which will CORRECT for the displacement between the towers, also if your bed is slanted, or your plate is warped, or if your towers are not square.

I had the area between the towers off by around 0.5-2mm, this program leveled my bed entirely giving a tolerance of less than +/- 0.015 millimeters. It has not had a lot of testing yet but it should still work properly.

My post goes into more detail on how it works, how to use it, etc. -
http://forum.seemecnc.com/viewtopic.php ... 698#p72789

If the link doesn't work just go to my page, it's my first post.

Thanks for the xpost, it's definitely relevant. Can't wait to see your video.

BTW, what do you think of MHackney or 626Pilot's code? They've been working on the same thing for (I think) longer than you have. Just curious.

I actually have not seen there code, I am going to go look into it. Thanks

[Jimustanguitar]

I actually have not seen there code, I am going to go look into it. Thanks

http://forum.seemecnc.com/viewtopic.php?f=36&t=7542
and
http://forum.seemecnc.com/viewtopic.php?f=82&t=7640

[d1rron]

So I seem to be having this issue. I thought at some point I had fixed it through a normal recalibration, but it would seem I was mistaken since my towers and center are all calibrated, I'm just having the issue between towers. Anyway, I checked my towers with a speed square and they all seemed to be perfectly flush at 90°. I've read through this thread, but haven't found a solution. Maybe I overlooked something? My CARRIAGE_HORIZONTAL_OFFSET is already set a 38.4. Any suggestions on where to go from here? Or do I have to upgrade to a Smoothieboard just to resolve this? I just really want to get back to printing and being mesmerized. Lol

[ZakRabbit]

So I seem to be having this issue. I thought at some point I had fixed it through a normal recalibration, but it would seem I was mistaken since my towers and center are all calibrated, I'm just having the issue between towers. Anyway, I checked my towers with a speed square and they all seemed to be perfectly flush at 90°. I've read through this thread, but haven't found a solution. Maybe I overlooked something? My CARRIAGE_HORIZONTAL_OFFSET is already set a 38.4. Any suggestions on where to go from here? Or do I have to upgrade to a Smoothieboard just to resolve this? I just really want to get back to printing and being mesmerized. Lol

I don't know if it's a "fix" but as I understand it, this program is obtaining similar results on the stock software: http://forum.seemecnc.com/viewtopic.php?f=36&t=8698

[d1rron]

Thank you for that link, it does look promising, but I don't have a Z-probe at the moment. I'll have to look into what I need to get one.

[KAS]

So, I'm going to say no on upgrading to Smoothieboard to correct the calibration as the results would be the same if the calibration issues are mechanical. Smoothie does offer different options to aid in the calibration process, but in the end I believe it still comes down to mechanical adjustments over software correction.

I wont knock anyone's calibration program but I have better luck using a dial indicator mounted on the effector to calibrate everything mechanically. You'll be surprised how much the effector lifts when putting the slightest amount of pressure at the top or side of one tower. I've tried the following just out of curiosity, and while I consider my machine to be mechanically calibrated these programs/methods changed values that made my machine less accurate. In the end, I gave up on auto calibration as it didn't work in my case. Obviously your mileage may vary and work perfectly for you.

DejayRezme/DeltaAutoCalibration : https://github.com/DejayRezme/DeltaAutoCalibration
626Pilot/Smoothieware : https://github.com/626Pilot/Smoothieware
RollieRowland/Delta-Bed-Calibration-Tool: https://github.com/RollieRowland/Delta- ... ation-Tool


Good read: http://minow.blogspot.com/

[ZakRabbit]

Thank you for that link, it does look promising, but I don't have a Z-probe at the moment. I'll have to look into what I need to get one.

I believe the earlier iterations were set to do it in a "long hand" way where you would take a reading with a dial indicator or something similar, then input the values manually. Many, including myself, are going with the FSR setup.

[d1rron]

Hm. I'll have to look into an FSR setup. I did notice that after reflashing my FW and Eeprom, and then reinputting he values from before I reflashed, the towers are now out of level, which is interesting. I changed horizontal radius and endstop offsets back to their values and all that. Idk. I'm going to screw with it more later.

[geneb]

Surely you're joking, Mr Feynman!



g.

[d1rron]

Ha! Hell yes! He's my favorite physicist.

So anyway, I ordered a new product from TrickLaser which they seem to think will, or at least may, resolve the dreaded triangle issue. They're just longer CF rods. They offer 300mm and 325mm. You do give up some build height, but I plan to eventually get longer towers anyway.

So anyway, I'm not endorsing them yet, but I will report back after I get them tomorrow. They also mention that you can get the full 11" diameter of print area.

[KAS]

Ha! Hell yes! He's my favorite physicist.

So anyway, I ordered a new product from TrickLaser which they seem to think will, or at least may, resolve the dreaded triangle issue. They're just longer CF rods. They offer 300mm and 325mm. You do give up some build height, but I plan to eventually get longer towers anyway.

So anyway, I'm not endorsing them yet, but I will report back after I get them tomorrow. They also mention that you can get the full 11" diameter of print area.

I wouldn't say it will eliminate the between tower issue, but It should push that error towards the edge of the bed. As for trick laser, great company to work with and he made the longer arms from a few requests on the forums

[d1rron]

Ah, good to know. I realized that it may not solve the issue, but I'm still crossing my fingers, and at least it'll make printing potentially more consistent, especially at the 5.5" radius. If this doesn't solve the issue for me then I guess I'll be looking into an FSR mod. It is weird how this issue disappeared for me at one point, and then came back. My towers are as square as ever. /shrug

I guess we'll see what happens. Lol I just want my printer back! Half tempted to build a frame like TrickLaser sells, but a bit taller.