M2 Accuracy -- What's reasonable to expect?

General discussion topics
BBID
Posts: 36
Joined: Sun Jun 04, 2017 4:47 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by BBID » Tue Jul 25, 2017 10:54 pm

Thanks for all the great feedback Zemlin!

I took your advice and printed four cubes at once. The lines are still there, but like the 0.100 mm (z-res) print I ran, there are more of them and they are closer together. I'm still waiting for Makergear support to get back to me on this problem. I am currently running calibration cube prints are two other printers, each made by a different manufacturer, to see if this issue is typical for the FDM process on all commercial FDM printers, or if it's just showing up on the M2. I'll post back when I get more info.

User avatar
insta
Posts: 2007
Joined: Tue Sep 16, 2014 3:59 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by insta » Wed Jul 26, 2017 3:49 am

are you using the same slicer on all 3 printers?

do you have a black Z nut, or brass one? post the output from M503 to help answer :)
Custom 3D printing for you or your business -- quote [at] pingring.org

BBID
Posts: 36
Joined: Sun Jun 04, 2017 4:47 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by BBID » Wed Jul 26, 2017 8:01 am

1. I'm using Simplify 3D for the Makergear M2 Rev. E, Flashforge Creator Pro, and the Ultimaker 3, but Eiger.io (Markforged's proprietary cloud-based slicer) for the Markforged Mark II.

2. My M2 Rev.E's z-nut is black, not brass. I'm assuming this is the latest iteration of this z-nut for the M2?

3. M503 output for my M2 Rev. E:

Steps per unit:
M92 X88.88 Y88.88 Z1007.70 E471.50
Maximum feedrates (mm/s):
M203 X200.00 Y200.00 Z25.00 E25.00
Maximum Acceleration (mm/s2):
M201 X900 Y1000 Z30 E2000
Acceleration: S=acceleration, T=retract acceleration
M204 S2000.00 T3000.00
Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s), Z=maximum Z jerk (mm/s), E=maximum E jerk (mm/s)
M205 S0.00 T0.00 B20000 X4.00 Z0.40 E1.00
Home offset (mm):
M206 X0.00 Y0.00 Z7.76
PID settings:
M301 P20.56 I1.42 D74.52


The quality and accuracy of the Mark II print is unbelievable. Dimensions are dead on, no PID lines, solid horizontal surfaces are completely flat, strength of parts is astounding. I can't even find the print seam/zipper on most parts. I can upload pics of the Ultimaker 3 print later...although its looks very, very similar to the Flashforge Creator Pro print.
CalCubeCompare.jpg

User avatar
insta
Posts: 2007
Joined: Tue Sep 16, 2014 3:59 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by insta » Wed Jul 26, 2017 4:22 pm

Yeah, the black nut is one thing I'm not a huge fan of -- because of this part of your M503 output: Z1007.70. It is imperial threading (1/8in per turn), but the firmware is metric. This causes a nasty 25.4 error, which microstepping can't entirely fix. The brass nut (no longer available for various reasons) uses a clean 400 steps per millimeter. "Round number" layer heights get even numbers of full steps. With a decimal step-per-millimeter number, there is no easy way for the firmware to divide it by a round-number layer height, and you'll get step errors piling up every dozenish layers.

Try "0.1905" as your layer height. That cleanly divides.
Custom 3D printing for you or your business -- quote [at] pingring.org

Gwhite
Posts: 372
Joined: Tue Jul 11, 2017 3:38 pm

Re: M2 Accuracy -- What's reasonable to expect?

Post by Gwhite » Wed Jul 26, 2017 4:26 pm

I must say, I'm jealous of the Flashforge results. My M2E is probably somewhere between your M2 and the Flashforge, but I'm still tuning things. The Markforge is amazingly good, but with a custom slicer and a $13.5K price tag, you'd expect that.

BBID
Posts: 36
Joined: Sun Jun 04, 2017 4:47 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by BBID » Wed Jul 26, 2017 10:05 pm

I must say, I'm jealous of the Flashforge results. My M2E is probably somewhere between your M2 and the Flashforge, but I'm still tuning things.
To add insult to injury, the Flashforge cost $1k less and printed that cube at 80mm/s, twice the speed of the M2. If Makergear would add a proper blower-type cooling fan to their machines, their printers would most likely surpass the Flashforge in terms of speed and quality.
The Markforge is amazingly good, but with a custom slicer and a $13.5K price tag, you'd expect that.
That quality can be had for much cheaper, the Markforged Onyx One is $3500 IIRC.
Try "0.1905" as your layer height. That cleanly divides.
I'll give this a try, but I'm not sure what 0.1905mm cleanly divides into...1007.70 steps/mm? To lower the build platform one layer height would require 191.96 steps, another decimal step amount. I'm confused. :?

mathisyourfriend
Posts: 100
Joined: Sat Apr 19, 2014 8:34 pm

Re: M2 Accuracy -- What's reasonable to expect?

Post by mathisyourfriend » Wed Jul 26, 2017 11:06 pm

"Try "0.1905" as your layer height. That cleanly divides."

I also got 101.96. I did some math and got to .1905329 which was good to 5 decimal places (theoretically).

I realize 1/10 millionth of a millimeter accuracy is too much to ask. Are you saying that .1905 is close enough?

I'm not trying to be pedantic. It's my curiosity and desire to make this as easy as possible for my myself and my students (while "tricking" them into using math).

Getting continually better looking prints is a great incentive for learning.

User avatar
insta
Posts: 2007
Joined: Tue Sep 16, 2014 3:59 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by insta » Thu Jul 27, 2017 2:02 am

it's more that the firmware won't care after a certain precision. i don't know what it is, but .1905 will be close enough :)
Custom 3D printing for you or your business -- quote [at] pingring.org

Bratag
Posts: 438
Joined: Wed Jan 14, 2015 5:33 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by Bratag » Thu Jul 27, 2017 3:41 am

BBID wrote:1. I'm using Simplify 3D for the Makergear M2 Rev. E, Flashforge Creator Pro, and the Ultimaker 3, but Eiger.io (Markforged's proprietary cloud-based slicer) for the Markforged Mark II.

2. My M2 Rev.E's z-nut is black, not brass. I'm assuming this is the latest iteration of this z-nut for the M2?

3. M503 output for my M2 Rev. E:

Steps per unit:
M92 X88.88 Y88.88 Z1007.70 E471.50
Maximum feedrates (mm/s):
M203 X200.00 Y200.00 Z25.00 E25.00
Maximum Acceleration (mm/s2):
M201 X900 Y1000 Z30 E2000
Acceleration: S=acceleration, T=retract acceleration
M204 S2000.00 T3000.00
Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s), Z=maximum Z jerk (mm/s), E=maximum E jerk (mm/s)
M205 S0.00 T0.00 B20000 X4.00 Z0.40 E1.00
Home offset (mm):
M206 X0.00 Y0.00 Z7.76
PID settings:
M301 P20.56 I1.42 D74.52


The quality and accuracy of the Mark II print is unbelievable. Dimensions are dead on, no PID lines, solid horizontal surfaces are completely flat, strength of parts is astounding. I can't even find the print seam/zipper on most parts. I can upload pics of the Ultimaker 3 print later...although its looks very, very similar to the Flashforge Creator Pro print.
The attachment CalCubeCompare.jpg is no longer available

It looks great but a couple of things.

1) You need to use similar filaments when comparing look. CF looks amazing in general - there are a couple of screenshots of CF prints on the M2 and they look great.

2) You have something screwy with your settings on the M2 you shouldn't be seeing lines like that - certainly I don't.

This is a print I did yesterday with some ratty old PLA to test the variable layers in the new S3D. I did thisa at 75mm/s it goes from 3mm -> 2mm -> 1mm . It has some ringing there that I wish I could get rid of.
Attachments
IMG_20170726_224441.jpg

BBID
Posts: 36
Joined: Sun Jun 04, 2017 4:47 am

Re: M2 Accuracy -- What's reasonable to expect?

Post by BBID » Thu Jul 27, 2017 7:06 am

You need to use similar filaments when comparing look. CF looks amazing in general - there are a couple of screenshots of CF prints on the M2 and they look great.
The CF print from the Markforged printer was just to show what a commercial desktop printer is capable of. Some users were suggesting that the perimeter lines were inherent to the FDM process no matter what printer you use. If you want me to use the exact same filament, slicer, infill settings, layer height, etc....look at my Flashforge calibration cube prints. I think I made my point that this issue is specific to the M2, although I haven't figured out exactly whats causing it. I contacted Makergear on July 9th about this. They've asked for a couple bits of info on my slicer settings and my g-code, but I haven't heard back from them for a week.
You have something screwy with your settings on the M2 you shouldn't be seeing lines like that - certainly I don't.
This is exactly what I was worried about and what I've been trying to convey to Makergear.
This is a print I did yesterday with some ratty old PLA to test the variable layers in the new S3D. I did thisa at 75mm/s it goes from 3mm -> 2mm -> 1mm . It has some ringing there that I wish I could get rid of.
I don't know how you got that print at 75 mm/s without warping, but my M2 Rev. E can't print PLA faster than 40 mm/s with a 50% underspeed on the outermost perimeter if I want the print to be even remotely accurate.

Post Reply