mhackney's Real V3 Build Thread

Start your own build thread so others can see how it's going, and even help out!
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

I found the post on insulating the V3 bed. Seems logical although I haven't had a problem at PLA temps. Perhaps at higher temps it might be needed.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
geneb
ULTIMATE 3D JEDI
Posts: 5362
Joined: Mon Oct 15, 2012 12:47 pm
Location: Graham, WA
Contact:

Re: mhackney's Real V3 Build Thread

Post by geneb »

I suspect it's done with the assumption that something needs to be done in the absence of a reproducible need. :D

g.
Delta Power!
Defeat the Cartesian Agenda!
http://www.f15sim.com - 80-0007, The only one of its kind.
http://geneb.simpits.org - Technical and Simulator Projects
User avatar
DeltaCon
Printmaster!
Posts: 616
Joined: Sat Nov 14, 2015 5:01 am
Location: Wessem, The Netherlands
Contact:

Re: mhackney's Real V3 Build Thread

Post by DeltaCon »

Thanks Michael, I know this is your thread, but the question was to whom it may concern ;) . Actually a good point about the steadyness of the bed while removing parts. Hadn't thought of that yet.
I am DeltaCon, I have a delta, my name is Con, I am definitely PRO delta! ;-)
Rostock V2 / E3D Volcano / FSR kit / Duet 0.6

PS.: Sorry for the avatar, that's my other hobby!
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Well, I struggled with Matter Control for as long as I could tolerate! Between all sorts of hiccups and lockups (temperature display would oscillate and the hot end would never reach temp) that were directly MC (I tested with Pronterface and Repetier with NO problems) I decided to forge ahead with OctoPrint that I had set up on my Raspberry Pi at the start of this thread. It took about 2 minutes to hook up and configure the printer. I then had to open Cura to create a working profile for PLA and upload that to OctoPrint. Once that was done, I started a print (my BandAid test print) and away it went. The web interface is much nicer and having a dedicated controller over USB eliminates a lot of the USB jerkiness I was experiencing with MC.

OctoPrint comes with the Cura slicer built in so you can slice from the web interface. Here is a reasonable default PLA Cura Profile you can use for a stock V3. You load it into OctoPrint by clicking the Settings menu at the top of the page, then you all see CuraEngine under the PLUGINS header at the bottom left, click that, then click the Import Profile... on the Cura config page and load this profile (after removing the .txt at the end, it must be called DefaultV3PLA.ini).
DefaultV3PLA.ini.txt
(10.73 KiB) Downloaded 356 times
You do have to setup a printer profile too. Click Settings then Printer Profiles. Click Add Profile... and enter the info for your printer:

Form Factor - Circular
Origin - Center

Volume
X - 275 mm
Y - 275 mm
Z - 400 mm

Heated bed checked
Nozzle diameter 0.5mm

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

For those who bought the RaspberryPi kit from SeeMeCNC, did it come with the mounting standoffs - the ones like the RAMBo standoffs used to attach the board to the top plate?

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
travelphotog
Printmaster!
Posts: 398
Joined: Tue Jul 15, 2014 12:30 pm
Location: Huntsville, Alabama (Marshall Space Flight Center)
Contact:

Re: mhackney's Real V3 Build Thread

Post by travelphotog »

mhackney wrote:For those who bought the RaspberryPi kit from SeeMeCNC, did it come with the mounting standoffs - the ones like the RAMBo standoffs used to attach the board to the top plate?
I saw John mention that it does come with the same type stand off and I saw a pic of one last night on FB.
http://713maker.com/ Custom aluminum and carbon fiber hot end mounts for the Rostock Max and Orion.
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Ok, thanks. I guess I'll draw something similar and print a few. I couldn't find the CAD for these on the SeeMeCNC git.

Did you get my email last night (5ish EST) about the Cyclops mount?

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Sorry, I found your reply! For some reason it ended up in my Junk folder. Thanks.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
martins
Plasticator
Posts: 18
Joined: Fri Oct 07, 2016 8:29 am

Re: mhackney's Real V3 Build Thread

Post by martins »

mhackney wrote:For those who bought the RaspberryPi kit from SeeMeCNC, did it come with the mounting standoffs - the ones like the RAMBo standoffs used to attach the board to the top plate?
yes i got mounting standoffs with:)
martins
Plasticator
Posts: 18
Joined: Fri Oct 07, 2016 8:29 am

Re: mhackney's Real V3 Build Thread

Post by martins »

mhackney wrote:....

You do have to setup a printer profile too. Click Settings then Printer Profiles. Click Add Profile... and enter the info for your printer:

Form Factor - Circular
Origin - Center

Volume
X - 275 mm
Y - 275 mm
Z - 400 mm

Heated bed checked
Nozzle diameter 0.5mm
thx for this but howto setup Axis? in octoprint for RostockMax v3
[img]http://bbmail.co/semecnc/octoprint.png[/img]
geneb
ULTIMATE 3D JEDI
Posts: 5362
Joined: Mon Oct 15, 2012 12:47 pm
Location: Graham, WA
Contact:

Re: mhackney's Real V3 Build Thread

Post by geneb »

You can leave those defaults as they are.

g.
Delta Power!
Defeat the Cartesian Agenda!
http://www.f15sim.com - 80-0007, The only one of its kind.
http://geneb.simpits.org - Technical and Simulator Projects
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Some tweaks and tips...

Post by mhackney »

Once you have your V3 built and probing and printing here are a few things you can do to improve auto-calibration and reproducibility.

Tweak the carriage screws to get all three tower endstop offsets to within 4 steps/mm (or less). Travel is 80 steps/mm so 8 steps is 0.1 mm. The target 4 steps is 0.05mm. As you can see in the screenshot, mine are within 4 steps.
Screen Shot 2016-10-13 at 2.57.38 PM.png
Here's how you do it:

Calibrate to get your baseline. After you calibrate, look at the eeprom settings for the tower offsets as shown above. In OctoPi I use 2 plug ins that make this really easy: 1) EEPROM Repetier Editor Plugin and 2) Custom Control Editor. I created a custom button that runs the v3Probing.gcode script, making it really easy. Here's a screenshot on how to set that up:
Screen Shot 2016-10-13 at 3.05.19 PM.png
Once you have the control created, you need to put a copy of the script in: /home/pi/.octoprint/scripts/gcode/custom on your RasperryPi. The control automatically looks in the custom folder.

Ok, so you calibrate and then look at the eeprom settings using the EEPROM plugin. I've found that a 1/4 turn counterclockwise is about 7 steps. Start with your tower with the largest offset. Take the difference in the number of steps between the towers with the greatest values - so if Z was 21 and X was 8 and Y was 14 - the difference between Z and Y is 7. Simply turn the Z screw about 1/4 turn counterclockwise then re-run the auto-calibration and check again. Continue this process until you sneak up on <4 steps for all 3 towers. Note that you could also do the opposite of what I described above and increase the steps of the lowest tower value.

Once you get to a low step difference, run auto-calibration several times to check repeatability. If the steps change significantly from run to run, you likely have some sort of mechanical slop somewhere. Loose endstops, pulleys, steppers and belts are prime candidates. Hunt the problem down and fix it.

While you are doing the repeatability check calibration runs, also jot down the delta (horizontal rod) radius and Z Max length to make sure they are reasonably stable too. If delta radius caries greatly, you may have an issue with tower rigidity, lean or angle or the bed is not rigid. If the Z Max length varies the bed may not be firmly clamped.

NOTE ON BED CLAMPS: I prefer to use every other bed clamp instead of all 6. With 3 clamping positions, since a plane is defined by 3 points, there is no chance to warp or distort the bed. With 6 there is a chance that the bed could warp if you are not careful. It is very difficult to calibrate a delta with a slightly warped bed!

Another thing you can do is add a line to your v3Probing.gcode script to turn off the hotend while probing. This allows you to get everything heated up and then probe while things are hot but with no issues with current running to the hot end (the accelerometer probe won't work properly if there is current flowing to the hot end. Here is my modified script:

Code: Select all

M104 S0 ;turn off hot end
G69 S2 ;Endstop Calibration
M117 ENDSTOPS CALIBRATED
G68 ;Horizontal Radius Calibration
M117 HORIZONTAL RADIUS CALIBRATED
G30 S2 ;Z height calibration
M117 Z Height Calibrated
M500 ;Save values to EEPROM
G4 S2
M117 SAVING CALIBRATION

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Some Calibration Run Data

Post by mhackney »

Here are the results of 4 back-to-back calibration runs with a cold hotend and bed.

Code: Select all

X endstop steps:      3              0              2             0
Y endstop steps:      2              2              2             0
Z endstop steps:      0              1              0             0
Horizontal radius:    143.424        143.465        143.437       143.451
Z Max length:         398.937        398.950        398.950       398.975

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Auto calibration findings

Post by mhackney »

An experiment that I and user Th0mpy did last night was to rotate the HE280 120°, I rotated mine counterclockwise (looking down from top) and he did clockwise. The experiment was to determine if the position of the accelerometer on the board had anything to do with the out of level issue a few of us are seeing after calibration. In both our cases, the out of level stayed in the same place - high at the Y tower base and low between the X-Z towers. Very interesting result. Here are the calibration results:

After calibration in the new rotated position:
Screen Shot 2016-10-14 at 11.24.49 AM.png
after rotating back to original position:
Screen Shot 2016-10-14 at 11.36.47 AM.png
Interestingly the tower offsets were greatly affected with the rotated HE280 but they returned to "normal" when I moved the HE280 back. Note the 80 steps is 1 mm, so the rotation introduced nearly a mm of error. Just an observation for now, need to think through the math and devise some more experiments. The next thing I plan to do is a manual calibration to see how it compares to auto calibration and, more importantly, how it prints.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Ok, I simply swapped the Y and Z probing points to basically cause the firmware to do the probing at the base of Y (rather than Z) to calculate delta radius. After auto calibration, the results all looked reasonable and in line with the above. Most importantly, the bed tilt remains unchanged and in the same direction - from Y to X-Z.

I saw Polygonhell's comment in the other thread about possible calculation errors in the offsets. I'll run that test next.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
Mac The Knife
ULTIMATE 3D JEDI
Posts: 1409
Joined: Sun May 11, 2014 6:18 pm

Re: mhackney's Real V3 Build Thread

Post by Mac The Knife »

My Eris also has the issue of a "sloped" bed along the Y axis. High at the Z tower, and low at the opposite end. Using Repetier, I watched the serial communication while it was going through the auto-cal routine, and at the X and Y tower, it would show two sets of numbers,,,, one for each tap, but at the Z tower, only one even though it taps twice. I have since just calibrate it in the eeprom settings, and don't even use the auto-cal.
R-Max V2
Eris
Folger Tech FT-5 R2
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

I auto calibrated and then checked the tower offsets at all 3 towers at the same probing point. I discovered that Y was 0.15 mm too high than what was saved in EEPROM. Z was just a bit high and X was right on. I did this 3 times with the same results. I am now manually compensating the tower offsets in EEPROM and also readjusting Max Z. This should result in a level bed and if the delta radius is good, should result in a good print.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Well, that did the trick. The first layer went down perfectly.

I think a few of us need to do repeat this test several times and report the values with auto cal and then the final manual calibration. I'll look over the firmware and see if anything jumps out.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
Polygonhell
ULTIMATE 3D JEDI
Posts: 2430
Joined: Mon Mar 26, 2012 1:44 pm
Location: Redmond WA

Re: mhackney's Real V3 Build Thread

Post by Polygonhell »

If you temporarily swap say the X and Y towers over electrically or in pins.h, and rerun the same test and the error is now in X, I'd pretty much guarantee the issue is software.
I glanced at the code, it's more complicated than it needs to be i.e. it computes a plane equation which because of the delta geometry and the fact it's measuring directly under the towers, it doesn't need to (it just needs to scale the measured offsets by the ratio of printer radius to probe point radius).
I didn't see anything obvious that would only affect one tower unless the Plane equation calculation has an issue and I didn't look at that.
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

As soon as this part is finished, I'll swap X and Y steppers and see what happens.

G69 is implemented so I can actually test just the endstop calibration separate from delta radius and Z height.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Oh, and if anyone tries the stepper swap, don't forget to swap the end stops too!

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

Swapped the X-Y steppers and end stops, auto calibrated and ran. Did the test twice to validate. In both cases, the tilt moved to the X - Y-Z direction. However, the thin side is now towards the X tower and the thick side across from it. This is opposite of the original problem where the thick side was near the Y tower.

I'm reverting back to normal config now.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
Polygonhell
ULTIMATE 3D JEDI
Posts: 2430
Joined: Mon Mar 26, 2012 1:44 pm
Location: Redmond WA

Re: mhackney's Real V3 Build Thread

Post by Polygonhell »

Then it's most likely software, it could be one of two things.
Either the computation of the plane equation or it's use are incorrect.
Or less likely the specified probe points have inconsistent signs with the internally tracked positions, which would result in an incorrect plane equation.
I don't have a delta with a probe on a RAMBO to debug this, but I'd start by logging the X/Y/Z offsets and the computed plane equation to see if they make sense.
Though I'd probably just get rid of the plane equation altogether, because it's unnecessary, you can do the entire calculation just using similar triangles, because you're probing under a tower.
I just drew this on a piece of paper and I take the above back the calculation isn't similar triangles, but it's still relatively simple.
Th0mpy
Prints-a-lot
Posts: 21
Joined: Thu Aug 11, 2016 1:35 pm

Re: mhackney's Real V3 Build Thread

Post by Th0mpy »

Just moving over from the other thread...

Where do we go from here? Do you think it makes sense to try Repetier 92.9?
User avatar
mhackney
ULTIMATE 3D JEDI
Posts: 5412
Joined: Mon Mar 26, 2012 4:15 pm
Location: MA, USA
Contact:

Re: mhackney's Real V3 Build Thread

Post by mhackney »

No, it makes NO sense to try 92.9. SeeMeCNC added code to a custom form of Repetier to support the accelerometer probe. ONLY the Repetier download from SeeMeCNC's GitHub will have that special code.

I'm working on it, just takes some time.

Sublime Layers - my blog on Musings and Experiments in 3D Printing Technology and Art

Start Here:
A Strategy for Successful (and Great) Prints

Strategies for Resolving Print Artifacts

The Eclectic Angler
Post Reply

Return to “The Build Zone”