1.8 vs. .9 Degree Steppers Help or Hype?
ฝัง
- เผยแพร่เมื่อ 7 มิ.ย. 2024
- In this video we find out are .9 Degree Steppers Help or Hype? Many 3D Printer vendors are now including .9 degree steppers. .9 Degree steppers can also fix Vertical Fine Artifacts. Vertical fine artifacts look a lot like ringing in 3D prints. We test them head to head to see how the print quality improves. .9 degree steppers are an inexpensive upgrade to many 3D Printer models. Before you upgrade watch this as we put 1.8 vs. .9 steppers in head to head competition. The results may surprise you.
0:00 - Introduction
2:43 - What the .9 Steppers should do
4:02 - Testing Results
10:11 - Wrap Up
Please mash that like button if you enjoyed this video and leave your comments below!
.9 Stepper Motor Used In Video
amzn.to/3lZAY2k
Microscope used in the video:
Plugable USB 2.0 Digital Microscope
amzn.to/3kdwPY4
Please consider helping support the channel through Patreon or GoFundMe so we can continue to develop new content.
/ makersmashup
Follow my Thangs.com profile to find all the models I make!
thangs.com/user/Makers%20Mash...
Chat with us and other makers on our Discord: / discord
Follow us on Twitter: / makersmashup
Instagram: / makers_mashup
Visit our website www.MakersMashup.com
MakersMashup Uses TubeBuddy as their TH-cam assistant! www.tubebuddy.com/MakersMashup - วิทยาศาสตร์และเทคโนโลยี
![[Teaser EP.2] “MasterChef Junior Thailand Season 3” วันอาทิตย์ที่ 9 มิ.ย. นี้ 6 โมงเย็น ทางช่อง 7HD](http://i.ytimg.com/vi/KojBzrifbWg/mqdefault.jpg)
Hi Rob, thanks for sharing your experiment results. This is helpful for those who really is looking for upgrades that make a difference on printing results. I appreaciate, my friend, for excelent video, as always!
Being in the semiconductor industry and working on robotics, I can say that the resolution of your prints depend on the driver micro-stepping the motor. The driver that we use for each axis is about the size of a 24V power supply, and it takes about 150-200 micro-steps/pulses to move 0.1mm. The driver divides each step by hundreds or thousands to move a 0.9 (400 steps/rotation) or 1.8 (200 steps/rotation) motor a few micrometer or nanometer. Some our tools use electromagnetic linear actuators with infinite stepping at the cost of about $120 million a pop...*Cries*
Thank you for sharing this, and your work on this video! Really interesting results!
It is difficult to talk about comparative printouts if you still have the ghost effect on the texts. In my opinion, this test model shows the problems with tuning your printer more than the difference in quality.
Moreover, some artefacts may suppress or hide others. This is a complex subject and the fact that you can see the internal structure does not make a direct comparison easy.
I would think more about how to print patterns for comparison.
Edit 31/12/21 - I've been using these 0.9 motors for almost a year. I have not done such experiments, but I have a comparison of the various test prints before and after. I must say that subjectively I see a significant difference in favor of 0.9.
Hmm curious. The visible infill can be drastically improved by unchecking "infill before wall" in Cura. It took me a while to figure that out. I'm not sure if that changes as structural integrity but it really helps removing infill artifacts and cleaner edges.
Thank you for this input. I had forgotten that I had this activated since it helps when changing colors
Great video! Thanks for the comparisons!
Thank you for taking the time to do these experiments.
I believe stepper drivers make the largest impact on print quality.
Great and interesting video 👍
Nice to have a actual test, to see if there are any difference in the prints
Thanks for sharing 👍😊
Interesting. I'm curious to see the difference between the 0.9 and 1.8 steppers when combined with different stepper drivers. Thanks for the video!
I noticed some time ago when i started printing around 2012 that the melt zone acts as a shock/buffer for VFA if you print at just the right speed where the plastic is melting very well and is also printing fast enough.
Specifically noticed when i took out the PTFE tube out on my Replicator Dual, that made the filament pool up and create said buffer/shock area. Later noticed the same happens when you have a long melt zone, like on a Volcano hotend. Other thing i noticed is older stepper motors seemed to bleed off a lot of heat, eating up ruff stepping on older stepper motor drivers like smoothers do.
With all that said, 1.8 degree printing at higher speeds (150mms+ on outer walls on my printer) seems to remove most of the VFA's. But also when i print very slowly with a 1.8. So 0.9 degree might work for you, and/or you might never really benefit from it while loosing a bit of holding torque
Curious about the dimensional improvements between 1.8/.9. motors
it'd be cool if you tested some of those closed loop servo steppers (mk2b or something?)
checked if they improve ghosting
Maybe a noob question, but what orientation were the "square" test prints in? If that was a CoreXY, then walls aligned to the axes are actually diagonal motions, running both motors. Could the VFAs be caused by being a half-step out of sync?
How about changing the pulleys too?
great comparison video!
would it affect tolerance though? well visually yeah it may not be a big deal, but would this turn a hobbiest machine to a one of production horse level one?
I'm currently working on motorizing my telescope mount based a project called OnStep. I've got two nema17 0.9 motors on the way, and for astrophotography it will make a big difference (especially for my setup which is not using a guide camera to keep everything dead center).
Glad to know that they're not that important for 3d printers (likely my next project, maybe the x201) because the price difference is significant if 5 motors are needed.
You did not say it, but did you switch X, Y, Z, and E motors or was it just the E motor? How does Maximum Resolution, Maximum Travel Resolution, Maximum Deviation play a role into how it works? Also, the infill sticking out could be a problem with not enough walls. Does adjusting Acceleration, Jerk, Junction Deviation, and Linear Advance help with the 0.9 over the 1.8 motors? Would Arc Support benefit from the more steps from the motors?
I think the holding torque is where the 0.9 deg steppers make a difference. They are a bit worse. I would only use 0.9 deg for the Z axis, where speed and torque don't matter so much.
it always depends on other variables and you could just gear a 1.8 to 2:1 instead. most typical machines purchased like ender/CR10's have slow outer shells for better quality. the slower you print, the more chance of VFA's. A Stepper's full step is smoother at faster speeds, there are some demonstrations around for this. i've done this experiment nearly a decade ago using s3d and just printing outer shells faster removed them. further improvements can be had by lower printing temperature at the expense of strength. reason is PLA and PETG being the most common are fairly molten and pick up way more artifacts from vibrations or stepper steps. with ABS or ASA for example the print quality in terms of surface finish nearly doubles. that's the magic to good prints, different filament. either matte or filled filaments produce great results since they're more viscous when extruded. a CF-Nylon looks almost like it was SLS printed.
Thanks for sharing, I use 0.9° steppers on my (delta) printers (and also use 1.8 for messing around ;) ) . I'm not sure about your testing methods here, I would not suggest 0.9° steppers if someone asked me how to improve there prints, going from the reference prints you show. I would say say to further tune the print profile.
The 0.9° shows its advantages best when printing slower in my experience, not faster. Steppers tend to fall back towards full steps when moving too slowly, which makes for more accurate positioning, like probing and homing using a 0.9°. (while delta's print fast, the arms don't always move fast)
While 0.9 have a lower torque, but they have steeper torque curve, which some say is better and worse, but torque aint really an issue on printers...
But overall, don't expect a 0.9° to be twice as good as a 1.8°, A geared extruder is not -ratio- times as good as a non-geared one, and either should only be used to push the quality of well build machine, not to solve an issue. And yes, first the board and the drivers to upgrade before you think of the motors.
Would better belts make a difference?
Can they produce thinner layer heights? I never actually considered them for fixing artifacts.
The other thing I'm looking at is closed loop steppers, namely those from btt. Teaching tech already did a vid on them but would be great to see your opinion.
If you want thinner layer heights youre better of changing the lead screw to something like 2mm like we did on the X301. I've tested closed loop steppers. If you have frequently missed steps it could make sense but I've been able to address that problem in the past by adjusting the motor current. For most 3D Printer applications I don't know if CLS makes much sense. Just my opinion but if you feel you have a use case for it they do indeed work great.
Only thing I've found is a reduction in Vertical Fine Artifacts. Which you hit on right away. Great video
The timing of your video was perfect, i was giving. 9 deg steppers serious consideration.... Well, i think I'll just stick with the 1.8's.
In my experience with CNC I observed that the microstepping is some sort of spring, then no way to set the exact position in the time. The motor 0.9 have the double positions than 1.8 then the spring zones are lessed due that the motor arrive a their positions with much more garantee than the positiond intermediate by microstepping that are more uncertain. Then these effect of spring of microstepping can be the resonance of VFA, been minor in the 0.9 motors due to minor angle to freely spring between real steps of motor.
A little late to the party (thanks youtube!). The results you found make sense. Moving to a spot and stopping will be more accurate with a .9 (assuming you have the juice to hold a microstep properly) but in theory, once in motion, either stepper will barrel on through positional data, letting inertia carry it if there is no need to stop, so there should be little difference in the overall layer quality. Still, nice to see proof that this idea holds water. (puts wallet away)
Hi Rob ;) good video! It is because of the weight of the bed! Install the motor with more torque and everything will be fine!!!! I did test and video half year ago and no difference once I have install on the table motor with torque 70 n per s- everything is good! Try it!
Do you have a link for the motor you're talking about? Thanks!
could this be related to your firmware settings? Should your firmware settings be adjusted for the 0.9 steppers?
The firmware was appropriately updated.
very well demonstrated. I'd just get a cheap resin SLA printer if details are required. It works better than 0.9 fdm ones at only a fraction of cost.
WHat if you upgraded from 16 microsteps to 32 microsteps?
did try any other slicers other cura? i would be curious to see what results you would get with prusa slicer. i much prefer it over cura Imho
I haven't but it is a good question. There have been a number of good suggestions in the comments so I will likely do a follow up video on this.
@@MakersMashup that sounds awesome. i started off using cura, but have much preferred prusa slicer / slicer, i believe prusa slicer is a derivative of slicer, and from my experience the vase mode in pslicer produces better results. also pslicer performs better on my system. looking forward to seeing an updated video. 👍
cool video but I was hoping you would do some actual testing of the motors not just the print results.. Like do 0.9 steppers run cooler than 1.8? and how does the holding torque affect it...
I think most educated people know that you aren't getting a huge increase in quality from less than 1degree difference cuz the length of the belt and the pulleys practically negate that..
I have good experience with TMC2130 and 0.9° steppers. It allows me to very fine tune the z offsets. They aren't so expensive, so why not increase resolution. Use a fast microprocessor, not an atmega 8
I did a lot of tests with 0,9 and 1,8 so my conclusion is new motors with more torque ;) using tmc 2209 ;) works good!
What Motors did you use with the 2209?
I think the best thing of the 0.9 is that you can lower the microstepping to 8 in order to get much higher torque while maintaining the resolution
I thought about that while I was doing the video. The steppers I got were pretty beefy but if I run into any skipped steps I will certainly give it a shot. The 1.8's are 45Ncm while these .9's are 44Ncm so I think they should be fine running as is but if do run into problems i'm certainly going to give it a shot.
I even try to lower to 1!!!! Better install new motors with more torque!!! On my bed I have installed one with torque 70 n p s !!!
It makes a big difference in bed slinging printers, it’s not a hype at all. In prints with long lines along y and diagonal lines I rather have a 0.9 degree stepper on y axis.
A difference might have been easier to spot if there wasn't the infill showing through everywhere...
update ur firmwares segmenting to take advantage of the new steppers too also 0.9 on extruder makes a bigger difference
In my case I'm not sure a .9 stepper on the extruder will make any difference. The X301 uses a geared extruder so the resolution is already 415 steps/mm. I think its unlikely i'll see a visual difference in extrusion.
Frankly your prints are not high quality enough to show the difference. If you still have infill and ringing showing through you have much bigger problems to concern yourself with than upgrading to a 0.9 stepper.
I think you must have missed key portions of the video commentary. I said all of these were printed at higher speeds to show the VFA's . Would be pointless to show a 3D Print without the VFA's in it. The test cubes clearly show them but when you're printing at the same speeds under normal use you'll never see the difference. Which was the point I was raising specifically. I invite you to share you're personal results.
@@MakersMashup You can have vfa's regardless of print speed. Also what someone considers to be normal use will differ from person to person. It may be the case that 0.9 steppers do not make a difference for your printing needs but they do make a tremendous difference for my printing needs. If the only use case is for functional prints and or high speed print then I would not recommend most users spend the money to upgrade to 0.9steppers as it simply won't make a difference for them. In addition I watched the entire video.
@@MakersMashup I don't know that youtube allows the sharing of links but I'll give it a shot Here's the last print to come off my printer (please zoom in and pixel peep) cdn.discordapp.com/attachments/684578465139392556/767543974377685022/Chase_Daigle_DSC6879_2020-10-18_20.11.09.png
And here's a typical miniature for me: cdn.discordapp.com/attachments/684578465139392556/752690111199576104/Chase_Daigle_DSC6655_2020-09-07_20.24.08.png
Both photos were taken on a 200mm macro photography lens on a full frame sensor
I was wondeting why the VFAs did not show up in your practical tests, but your remark here on printing at high speed tells me all i need to know.
You actually have it backwards; VFAs are typically showing more on slow to moderate speeds.
The reason why you didnt see much change means you were testing mostly outside of the bandwidth where it matters. And the issue beung replaced by your other artefacts.
Especially at high speeds you can not only run into ghosting but also demand too many steps per mm, stepper drivers overheating or resorting to spreadcycly instead pf stealthchop, or even the steppers skipping steps at some point.
I ordered my 0.9 degree steppers today and cant wait!
Is there a difference in the noise between 1.8 and 0.9 stepper motors?
Nope. Most of the noise on steppers is the drivers not the motors themselves
Well, how fine details we think we can print? If we use 0,4mm nozzle and 0,2mm layers. If we have 80 steps per mm with 1,8° and 160 steps per mm with 0,9°. If we print straight, like cube wall, only length can change and if we use only one motor. That difference could be a0 - 1/80 mm. What is 0 - 0,0125mm. I think, because if our positioning was 1/80 mm and we got one step between originals, it give us double accuracy. So, if you ask wall length to 20mm, then both motors try make that 20mm. Well, of course there is 80 steps more with 0,9° motor, but that means 0,00625mm steps instead of 0,0125mm steps.
So what I think, if we need or want finer details, we must change at least our nozzle and desing much finer to get different results.
If we print stuff what is "designed" to 0,4mm nozzle and 0,0125mm accuracy is enough, we do not get much different results if we add more steps to our motors. =)
Maybe, if we print with 0,1mm nozzle and layer high with 0,05mm... and 0,2mm high text or other details. Maybe then we get some difference?
I agree. You might see finer details with a .2 nozzle. My point was that .9 steppers are largely hype. While there are minor improvements other changes are necessary to get significant results. Absolutely agree though a .2 nozzle.
Exactly.
I didn't meant to point of your video with my first comment. Maybe more like a support it.
I start thinking what is the difference between those motors.
And I do not address to you these thinkings, but all who might think difference between 1.8° and 0.9° motors. And think if they need better positioining.
What I try to say, is maybe easier understand with following:
If I have perfectly tuned printer and 1mm is exactly 80 pulses with 1.8° stepper motor.
That mean, if I want move my nozzle 20mm, then 1.8° stepper motor got 20 x 80 pulses = 1600 pulses and stop exactly 20mm point.
Same with 0.9° motor, needs 20 x 160 pulses = 3200 pulses and nozzle stops exactly same spot.
Well, no difference.
If I want move my nozzle 20.0125mm, then with 1.8° motor got 1601 pulses and 0.9° motor 3202 pulses. and spot is same like a last time. No, difference.
(1mm / 80 pieces = 0.0125mm and 1mm / 160 pieces = 0.00625mm)
If I want move my nozzle to 20.00625mm, then 1.8° motor perhaps travel 20.00000mm or 20.0125mm (1600 or 1601 pulses), but 0.9° motor can travel 3201 pulses and spot on 20.00625mm.
So, difference is 0.00625mm.
That was theory and math/electrical accuracy. And if I have mechanical accuracy of 0.00625mm, I wondering =) (you could try if your printhead not move any direction when you touch it)
If there is huge gap at 0.5mm, it is the same if I loss 40 or 80 pulses, because it is 0,5mm.
If there is 0.5mm zig zag, my nozzle could go straight.
Littlebit more math or theory =D
Where that 80 pulses coming?
Well, many times printers has half inch gear at top of stepper motor. That is about 12.7mm. So, 12,7mm times π is about 39,8982mm at full round.
When that is divided by 200 full steps, then one full step is about 0.19949mm (360° / 200 = 1.8° or 360° / 1.8° = 200). Well that is not 0,0125mm...
Ok, there is so called micro steps. And there is 16 of them (usually). If that 0.19949 is now divided by 16, then we got wanted 0.0125mm. =)
So, if those micro steps are actually steps, why there is different step types. And could some drivers actually use 256 micro steps. If so, can I make electrically positioining at 0.20 / 256 accurasy? (0.00078125mm). Or is that only audio level thing and actually there is always 16 micro steps, but other drivers just drive smoother them. But stops only every 16th step and drive that step its own 16 step... =)
That was just thinking. I do not wait answers =).
so i found the .9 degree steppers were only helpful for my delta 3d printer. the higher the accuracy/precision of the z-probe, the better the delta experience. the .9 steppers help a lot with that accuracy of values.
My theory is that most of these artifacts are caused by inertia, not by stepper resolution. The printhead is bouncing around with quite alot force i guess take away wheight from the head will improve most prints.
just did tests with titan and bmg extruders! titan is smother!
The "vertical fine artifacts" for the most part aren't caused by the steppers, it's caused by the mechanical aspects of the gantry and drive system. The biggest improvement you can make is to change to Japanese or German steppers because if you pull a Chinese stepper apart and mic the armature poles you'll finfd they varyin width by several thou. This translates into irregular movement, with some steps being bigger than others. This is usually the biggest contribution to banding outside of the mechanical system.
before addressing VFA you should have eliminated all the other artifacts :D
What do you mean that you increased the resolution from .5mm to .25 mm?
In cura you can adjust maximum resolution so the minimum segment is smaller than the .5 default
@@MakersMashup Thanks for the reply. I don't remember seeing that feature. I'll have to have a peek around for that.
@@1QKGLH It's under Mesh Fixes, and is not visible by default. You have to go into "Settings", "Configure setting visibility" and turn it on.
What about the noise. I'm really mainly looking for noise reduction
The stepper drivers are your source of noise in most cases. These stepper are just as silent as others with trinamic drivers.
@@MakersMashup oh I meant to ask, are 1.8 degree steppers louder/quieter than 0.9?
Same level of noise. Most noisy printers are the drivers not the steppers themselves. The .9 work as silently as any others with TMC drivers.
@@MakersMashup my printer makes a lot of noise even with TMC2209 TBH. I was tempted to decouple the stepper from the belt path and use a dampener for my next design
I don’t see how 0.9° stepper could help with VFAs. For a straight edge aligned to either the x or y axis on the bed, there should be no difference along the faces. The accuracy of the stepper motors will be the same. Only the precision is different. VFAs are an accuracy issue. Only precision is improved with the smaller step size. The accuracy of the hotend motion won’t be improved.
If you had a course hotend motion that only could move by 1mm increments, but that motion was accurate to 10nm, you would have no VFAs. You’d have other issues for sure.
Conversely, if you had a motion system that had 10nm step precision, but only 1mm repeatability (low accuracy), you’d have terrible VFAs.
People confuse precision and accuracy all the time.
How would I increase accuracy?
@@Nick-wl4qr microstepping, or by decreasing the number of teeth on the motor pinion. But is motor accuracy really an issue on 3D printers. I would argue that other factors that frame rigidity, belt stretch, linear rail smoothness, and motor mount stiffness are the limiting factors on the motion system.
All this test is showing is the inconsistency quality in your linear motion system not the steppers them selves imo.
Try changing the stepper on your extruder, not XY.
Come on, there is definitely a difference in quality. It is not that huge of a difference, but it is definitely visible!
you need 0.9 stepper on XY and E plus a 32bit board if you want to see real results. 0.9 steppers doesnt work on 8bit boards and you indeed need to increase the resolution in the slicer. Cura is a bit shit for this. they all use 8bit boards in ultimakers. most of the artifacts I see in your video aren't stepper related. well most of them tbh.
Ah yes every 1/160 of a mm movement is very crucial coming out of a 4/10 of a mm wide hole
I think it has more to do with the quality of the steper
I'm curious. In what way would you think the quality of the stepper would affect it? while I didn't use something high end, Stepperonline steppers are fairly decent.
@@MakersMashup my sidewinder X1 came with ok Stepers but after a few months they began to cog a bit introducing artifacts using a better steeper than the 5 dollar bulk stepper seems to have corrected the issue and improved accuracy
This probaly not the case in your printer as you have fairly good one but might be intresting to swap out them steper online ones with some craptastic generic ones liek on a stock ender 3
Also might be worth trying driving the .9 steppers at 1/8 stepping insted of the 1/16
I'm sorry but your vase mode prints look very rough... I would not accept this quality from my Ender 5 Pro
It's not that .9 is better than 1.8. There are better 0,9 degree motors and there are better 1.8 degree motors. It helps to double the torque dents on the motor on a 0,9 vs the 1.8.
Like you noticed yourself, on spiral vase this is pretty critical, and you acknowledge improvement. It's a combination of motor induction and the torque dents a motor has. If you lmeasure a VFA defect with a caliper, it's not a wobble of the head, but it's a difference in inconsistent speed. (the line measures thicker and thinner points on a spiral vase printed object that has VFA). On a CoreXY the Y axis is most likely the affected axis. Lowering your current also helps. In my experience; when you compare a stepper by turning it by hand; a stepper with less obvious torque dents will show less VFA's. Higher induction doesn't help with VFA either.
On my Ender 3 when I converted the Y to linear rail the VFAs were so pronounced I switched back to rollers. Maybe I just needed to reduce the current. It was hot hot hot that's for sure. But my plan it to get some LDO 0.9 steppers and try it again.
heum... microstepping? lmao
0.9 LDO motors just sing and make noise -.- 2k on a voron 2.4 and Once all assembled: funny not funny with tmc2209 + 0.9 LDO. The Ender 3 standing beside the voron just laughs and prints xDDD
This comparison doesn't make any sense unless all other artifact sources are removed: Ringing (tightening belts and using input shaper with Klipper), pressure advance, better steppers (e.g. TMC2209) and a better extruder (one that pulls from both sides of the filament).
The steppers are 2209s, I'm not going to swap firmware for a video but you're welcome to test pressure advance and .9 steppers yourself. The belts are plenty tight.
@@MakersMashup I will once I finish building a Voron. I won't even bother with my current Prusa i3 because it has A4988 drivers (16 microsteps and really old tech at this point). Doesn't the TMC2209 have many more microsteps than 16?
256 microstepping through interpolation but the firmware still sends 16. The 2209 handles the interpolation on the chip.
I think what's been happening is the industry creates a problem for itself to solve and make money. What they do is lower quality over time with cheaper parts and then when people start to ask why prints are not looking good, they come up with "you need to upgrade to a better one". Thus, company makes money. It is called planned obsolescence.
Will you be putting .9 Degree Stepper motors in your 3D Printer?
No
if I ever do, it would be X & Y, maaaaaybe Z? for weird layerHeights especially with smaller nozzles
No I did did it and no difference! I did video ;) with nozzles 0.6 to 1.2 no difference :)
Install motors with mitre torque ;) so the difference will be ;)
Definitely, minor improvement is still improvement. There's no way to argue that twice the step per mm resolution is a bad thing, especially considering the same or better torque. I happen to dislike VFA's, and I live for improving not only my printer but the world around me!