How to use Linear Advance in Marlin
ฝัง
- เผยแพร่เมื่อ 7 พ.ค. 2018
- Prusa research has recently updated the MK3 firmware (3.2.0 RC1) to include linear advance. This is an advanced feature of the Marlin firmware that aims to account for the pressure of the molten plastic inside the nozzle. Using LA should allow vastly faster printing with little reduction in quality amongst other benefits.
Please note that ideally you should perform steps one and two for each brand of filament and determine a K value for each different material. This video serves as a guide to getting started. Be prepared for some toiling to get ideal results.
This lesson features a Prusa i3 MK3 and Simplify3D but this can be applied to any 3D printer using Marlin firmware.
*edit: I made a mistake in the part where I measure with the ruler. This was actually done a couple of weeks ago and then refilmed for this video. I should have said 21mm left, not 19mm. This would mean 99mm extruded instead of 100mm.
The pages featured in this video:
Matt's Hub extruder calibration: mattshub.com/2017/04/19/extru...
Ian Johnson's setting the flow rate blog post: solidoodletips.wordpress.com/...
Matt's Hub linear advance: mattshub.com/2017/10/02/linea...
Chris Warkocki's MK3 linear advance enabled Simplify3D profile: / 853639768153839
Wii Racing Wheel: www.thingiverse.com/thing:346604
Gearbest affiliate link (saves you money and helps support the channel): www.gearbest.com/?lkid=13807527
Take a look around and if you like what you see, please subscribe.
Support me on Patreon: / teachingtech
Excellent video! Thank you for covering linear advance in sufficient detail and how to go about tweaking the Marlin firmware settings as well. Great information! Definitely saving this video in a playlist for future reference.
I read the video description after noticing the math blunder.
Congratulations, you got another sub.
Great tutorial!
Thanks for linking this all together and running through the steps, well done I think Ill give this a shot!
Greetings Teaching Tech! I would address you by your first name but I don't know it. This is my first Teaching Tech video & I instantly subscribed. The quality of information and presentation is just aces man. I took some notes to improve my own videos/channel. The straight to the topic video style is so refreshing & welcomed! I was so impressed by this video, I showed my mate here in the office & now we are analyzing your channel stats on +VidIQ. We are trying to figure out why you only have 3400 subscribers when 600 i3 mk3's are shipping every day. Of all the mk3 videos that I have watched, yours are hands down the best (sorry Tom S.). I'm sure I will eventually find out your name as I absorb more of your videos this evening. That will make you more relate-able to me as a human. Cheers & thank you.
Hello, my name is Michael. It flashes up briefly after each intro. Thanks for taking the time to watch and give detailed feedback. The channel is still pretty new, but growing nicely. Original goal was 1,000 subs in the first year but I've revised that to 10,000 after a good start.
Excellent tutorial - thanks. You’ve just earned another sub.
You literally are the best. I LOVE YOUR VIDEOS THEY'RE SO HELPFUL THANK YOU
Can’t wait to put this into practice on my Ender 3. Thanks sooo much!
Thank you for this video, nicely explained, thorough!
been looking for a tutorial on this! nice!!
Great content and video! Want to see definetely more such content. I am printing the walls for setting up the correct extrusuon multiplier with 3 perimeters (3x 0,4 = 1,2mm). I personally recommend to print (not only calculate) with different extrusion multiplier because there is some small area (gaussian courve) where different multiplier will result in optimal wall thickness. It is best to use highest saturation so that walls are sticking best (which result in highest layer adhesion). For my Prusa MK3 0.95 multiplier shows best results.
As always man, GREAT job! Love your videos.
Thanks for watching.
Awesome tutorial, thanks! One thing you forgot to mention is to ensure that the Coast and End and (I think) Wipe Nozzle in Simplify3D when doing the calibration, particularly for the K factor. Once Linear Advance is enabled those workarounds aren't needed anymore.
I agree I should've mentioned that in the video. The profile I used already had it sorted. I read that retraction should also be off but when I tried it I got terrible stringing.
Great video. Thanks for the tutorial!
Gained a sub mate. Brilliant and straight forward
Note that M503 doesn’t necessarily tell you the settings stored in the EEPROM. It reports the current settings in RAM that would get stored with M500. To get the stored values, use M501, and then M503 will show the values just retrieved.
Well spotted, thanks for sharing.
Thanks a lot for this video was really informative.
Thank you... Can't wait to try it.
My Ender 3 arrives tomorrow! So excited! I can't thank you enough for all of your help and continuing to put out great content.
Thanks for watching and have fun.
Ahhhhh....you got to this before my video came out! Hopefully, mine does it justice. Nice work!
Yours is better ...Teaching Tech copies everyone...
I've been watching your videos for awhile now and they're always very informative. As a side note, I'm always drawn to the lighted frame over your right shoulder. Is it something you've printed and if so do you have plans for sale?
I couldn't justify spending a 3rd of my printer price on a slicer, so how would you rate the ease of doing this on Slic3r: Prusa edition? ?
Hi thank you for this explainatons, even if I'm not realy good in english I understand the technicals reasons of LA but it's a little bit harder to do the complete settings specialy with simplify3D that I never used before, this will take me some time to read every think in description.
But you certainly know that last prusa firmware add some settings for LA directly into the lcd printer menu? do you thing we can enter the same values caculate into this video directly into the printer without simplify3D?
Have you try the prusa's methode with the test tower gcode for the extruder LA?
And last but not least do you plan to make a new video using this new mk3 menu settings for LA (please)
Thanks for the video. Is this firmware update with the K value something you can also do on the Cocoon, or in my case the Monoprice select plus? Enjoy your informative style. Thanks, Jim.
I haven't poked around in the Cocoon firmware yet but if I do I'll ebale this and make a video. Currently looking into which bed I'm going to use to upgrade.
Hello !
Thank you for those explainations :D
What speeds do you use to print the 25mm empty cube ?
I have a CR-10 and 65mm/s (3900mm/min) and 50% for outline underspeed and it looks like it's going about 1mm/s...
You should make sure to tag this video with the Linear Advance Version this example was for because the K values are different.
Fair point, this is for the Prusa version. I know their is another Marlin branch that has much lower K values.
I noticed you are printed much faster than I usually do (even when not using LA), for PLA I usually print at 60-70 mm/s (1920-4200 mm/min) without LA. Do you usually print these speeds on your other printers, such as Ender and i3?
Good job
Just an observation..... Ran the “ linear advance “ G Code noticed bed was heating very quickly... this is great no waiting. Third/ fourth time of G Code bed would not heat.... Eventually found SSR had fused and wire melted on bed plate. Then found your Video on installing “KEENOVO “ heater......turns out the manufacturer had installed a cheap copy SSR . Off to JAYCAR for a genuine SSR .
Super Video ! Weiter so !
Thanks for a good tutorial! But I was wondering is some part of this calibration something that should be done again when you switch to a different type or different roll of filament?
Step 1 no. Step 2, possibly. Some slicers have a setting to input filament diameter. Adjusting ths after measuring with calipers may suffice. The K value should be fine for changing rolls for the same type . Changing materials needs the K value updated for sure, although ABS and PLA is meant to be quite similar..
One thing about calibrating your extruder - generally speaking the steps are right-ish. But the problem with extruding through the nozzle for E step calibration is that then your motor is actually calibrated wrong. It will be calibrated for that filament with that nozzle, and that's it. Changing to another filament or another nozzle and your calibration is out the window if it doesn't interact the exact same way. You should calibrate into free air to get a correct measurement and calibration of the motor with only the tube being a factor in your calibration. Any other extrusion changes should come from calibrating your flow - which has to be done for every type of filament anyway.
While technically you can flow calibrate your way out of a poorly calibrated motor, it just isn't good practice. You also may have a slightly clogged nozzle during calibration without even knowing or it may be too cold.
Other than that, amazing video as all of your videos generally are :D
Hello, thanks for the video. I tried to repeat the test with my printer. Unfortunately, when it start to print, the extruder doesn't not work. the printed is moving according to the design without filament. If I do not activate this function with marlin, it works again. Any idea ?
thank you
So, it's sort of like an 'easing' function? I would imagine it has to calculate the ramp up, the increase needed in feed rate which I guess would be some algorithm that determines it at any given uSec to the point of max speed where everything would start to take place in reverse?
Excellent Video! Do not know if it was mentioned, I did not see it in the comments. When talking about the extrusion multiplier at `5:18, you mention 0.54 in the box. Did you mean 0.45? On the description flashed on the screen it does say 0.45. Just thought to share what I caught on to.
I think there was in issue with refilming for the camera here. From memory I have a note in the description about this.
What was the 3D benchy time before and after Linear advance was added?
What speed did you print the 3D benchy before and after Linear advance?
Where can I get that cool clock timer you got on your desk?
I didn't time them when filming, but S3D predicts 29 min at 8000mm/min with LA, vs 37 min with Chris' arlier profile, which was set at 7500. S3D is always wrong but you can make a relative comparison. Clock is a $10 cheapie from eBay.
Hey mate
I wondering if you could do a linear advance video using the Ender 3? I can't get the lines to work properly either by using the k value test you ran or by using the k value test compiler by marlin! The slow speed works fine but when it goes to the fast speed it doesn't lay any where near enough filament on any k value?
Please help!
I agree
Did you notice a reduction in the seams when you print a cylinder?
The way LA (v1.5) works is to compensate for elasticity in the filament and the extruders reactions to it. On a direct drive you gonna see much much much much less improvement then on a bowden setup, simply because the elasticity of the filament in a bowden setup is much higher.
And this makes linear advance an awesome feature to have on basically all bowden setups. And when i say awesome i mean.... insanely feraking awesome. It actually allows me to get up to speeds higher than with my flex3drive setup while maintaining almost the same quality. But at no additional costs and a simple install with very little fine tuning (as opposed to installing the flex3drive unit). as a major plus it also reduces stringing by quite a shitload. My retractions could be lowered by almost 60%.
I can therefore wholehearthedly say that linear advance 1.5 is a must on bowden setups. It always you to print past the 200mm/s mark without significant quality loss.
You messed up on your math. You OVER extruded by 1mm so you should have calculated 101mm not 99mm.
Otherwise, very helpfull. I got my PRUSA i3 MK3 dialed in quite nicely!!!
I realised after uploading, read the video description. Thanks for watching, glad it was helpful.
Ok I'm glad I wasn't the only one confused by that... I was like, wait he just made it worse and was hoping he was going to try it again to verify it pushed out 100mm and catch it, but didn't. That's also probably why he had to lower the extrusion multiplier by so much too. You really shouldn't go below .9 or above 1.1. If you do, you have an issue somewhere else that should be fixed first.
I try the linear advance on my hemera extruder. With the testpattern i get a result at 0.1. The testprint of a calibration cube came out super fine, but at a layer hight of ca 10 the extruder motor stop, did not know why. There is no overheat i think, measured the stepper with a laser thermometer at 45 degrees celsius. The driver is even colder. Do you have any ideas.?
Dos the K value change with Layer height , for example if you do your Test pattern at 0.2 layer height (where its Easy to pick winner) will it alter if you then drop your prints to 0.1 ??
Hey! my $14 caliper measures to the hundredths or the second decimal. :) Awesome videos!!
I was also confused about that. I've never seen calipers that are unable to measure to a hundredth.
Hey Michael, when i was tuning my linear advance and ive gone with the best looking line, basically folllowing your guide videos . But the result is i now have bulging at the first corner of my prints, the next corner has abit of radius.
Ive got an ender 3 on Bowden tube, extruder is calibrated, ive setup for 8mm retraction and 1.15 k value for linear advance, is there something ive missed, and help would be appreciated.
Your videos are always informative and helpful but i didnt have too much success one this one.
Thanks
same here bro
for some reason it is just doesnt working for my ender 3 (flashed the latest marlin)
i went up from 0 to 300 - no difference at all with the test pattern from the marlin's website
Hello Mr. TT i've read on serveral forums that you can better not mess with the steps/mm with Bondtech gears
because they need a specific number of steps to work optimal or is that not true?
Good info , however didn't you go the wrong way on the extruder calibration? You said default was 280 and you extruded 100 mm but you were left with only 19 which means you were over by 1. So you're estep should have been closer to 272? forgive me I see that other people have already pointed that out as well I didn't see the comments below. ;)
Hello, after type M503 command in the communication tab it shows SENT: M503 ; READ: ok without any information of X Y Z E. what do i suppose to do?
Try ticking the 'verbose' button.
Missed a very important step, which is measuring filament diameter and putting that in your diameter settings. Only after that can you calibrate your extrusion multiplier
Quite right you are.
1. Hey do we need to enter the new extrusion multiplier value ?
Hello Have you ever tested linear advance on wanhao i3 plus using advi3++ firmware
Does the K value apply just to the 0.4mm nozzle or all nozzles?
Could you please make the same video for the ezabl from th3d on your Ender 3? I know that Tim put the option to enable it in his custom firmware. But no idea what it will do on a creality. Thank's & Nice job by the way! Keep up!
Already done: th-cam.com/video/jARw5_w0uGM/w-d-xo.html
Teaching Tech A video to activate advence linear mode in the firmware from th3d. Not a how to install guide. What happen if I remove the # for a creality product enter 3 or cr10? Thank you
6:50 I can't believe it. Are you sure step 3 is finished?
Would the parameters derived by the calibration be drastically different between filament types? For instance, between PLA and Flex (e.g. TPU)?
This post from Prusa has some suggested starting points for different materials: help.prusa3d.com/l/en/article/t5w9VsdVai-linear-advanced
does this aid any at lower speeds or is it only useful when going faster?
There is always pressure in the nozzle so it should allow more consistent extrusion.
For calibration, if you make a larger cube, won't it make the calibration more accurate as any miscalculations will be more greatly emphasized? would also save you the trouble of buying another caliper if for some reason you don't have one that goes down to two decimal places
I suppose it matters whether or not the perimeters are squished together or not. Eg. a three permeter 0.45mm width equals 1.35mm or actually less. Haven't looked into that yet.
Teaching Tech that’s a good point, hmm, so some slicers allow you to adjust perimeter gaps or come default squishing perimeters, in theory the outer perimeter should always be laid to achieve appropriate dimensions and then the supporting perimeters squish into it and reinforce the outer perimeter, I guess that just opens a whole other can of worms on calibrations, all way above my 2am brain farting, I’mma just buy a better caliper for my cheap scrougy ass
@@Bajicoy if you think about it, as the slicer thinks it's extruding 0.4mm you will not have 3x0.45... there will be overlap between the outer perimeters and the inner ones. The extruder will draw a line thinking it's 0.4 then move 0.4 and draw another line. But because they are 0.45, there will be a 0.025 overlap. So even with multiple layers you need accuracy as the error is the same (half of it on the outside, half on the inside).
I need to get some sleep and stop rambling...
Alex Cristescu true, that overlap makes things super complicated, dang, I can’t remember what this was all about but man am I staying away from calibration cubes that aren’t a single perimeter thick lol good stuff though
dont have a Prsa, but used this and Chris Riley's videos to tune my highly modded Anet E12 to see if LA would help/do anything. thanks for all your awesome work and sharing of info! Hey, you have twitter? postin there and just assume be nice to give ya some credit/publicity !
Great video, its helped me improve my prints. The one thing I cant seem to get right is print speed. Im using Cura, and entering the settings you mentioned. I see your print time is about 5min. When I slice it takes 13 min. Even if I increase the print speed the time doesnt change. Lost
It's because the printer can only move so fast. You can put a million in there if you want it won't change the actual print time.
I'm guessing cura has a maximum speed value for your printer in the profile so it's working off that when calculating the time.
Simplify 3D is quite behind on features and probably doesn't have maximum speeds in the printer profiles. Meaning it is just calculating based on the numbers he put in. The reality would be different though.
+1 SUB , and , I'v burned my laptop with all my profiles, parts, and most important my f...... marlin profiles for my printers, and he burned as I was uploading a new non-configurated marlin on my printer. Have nothing to do with your vid but I had to share Xp. I like your techs
Thanks for watching and subbing.
Was there any need to mark the first 20mm?
Seems like it was not needed since the filament would have been melted in the nozzle anyway.
In the case of over extrusion, a line marked 100mm along will disappear but you won't know how far in. If you have a 20mm buffer first, you will be able to still measure. Eg. 14mm left if it extruded 106mm.
@@TeachingTech then just mark the 120mm? The 20mm mark makes no sense :) Even you did not measure 100mm from the 20mm mark but 120mm from the coupling. Keep it up!
i couldnt find the file from Matts Hub - do you have a direct link?
Thanks Michael, but I can't get a single wall cube to print in Cura - it doesn't slice if the top and bottom layers are 0 (unless it does, and my self-modelled cube is trash.
You dont want to test a single wall anyway. a better test is 2 walls.
I have a strange problem about this calibration process after I measure the cube wall by my callipers, I do the calculation, then decided to change the flow rate to 93%, the print result was good, the wall thickness is right, but the height reduced by about 5%, do you know how to keep the Z height when reducing the flow rate? thank you.
That shouldn't happen, it should be unrelated. My guess is a coincidental other issue.
@@TeachingTech ok, maybe, then do you know which parameter I should adjust to calibration the Z height?
How does this interact with the "extruder linearity correction" in the release 3.2 firmware?
I have no knowledge of that but am keen to learn more.
Should have no relationship or any adverse interaction on one another. I find it honestly quite dubious that extruder motor can be linearised altogether, as positional accuracy is secondary to torque, and the single step calibration suggests that only XY motors will be linearised. It stands to reason that visibility of the feature is benign to near negligible anyway and you certainly won't get more speed out of it.
Linear Advance is the modification of the planner algorithm which tweaks the acceleration of extruder on filament bead speed changes to compensate for built up pressure in the nozzle; while motor linearisation corrects the positional microstepping error, where particularly fine microsteps end up being skewed towards or away from the nearest half-step due to peculiarities of the motor construction. Effectively a successful linearisation may allow the existing planner decisions to be executed more exactly.
why all utubers show LA on prusa like printers with DD, can you make one bowden setup LA setup like Ender 3 ?
I was just about to ask the same - is LA even usefull on bowden-style setups?
whats your calliper brand and model?
Does this work for any filament or is it specific to pla I'm curious about petg
It works for any filament. The K value will be different, that's all. You can edit the temperature in the linked test gcode to suit other materials.
As far as I understand it, it works for every type, but you need to dial each type of material separately and store the "K" value in the slicer profile of that specific material.
The M500 command on the CR10 seems to restore to previous settings rather than store the new settings. Any ideas?
The firmware needs to have EEPROM saving enabled in the source code. I'm not sure if th CR10 version does, I'd be surprised if it didn't. Now that it has recently been released you cancheck and reflash with the change if neccessary.
Is this video just for prusa printers?
What is the name for this in cura?
8:36 so is it better or just same.so why?
I think your math may be off. If you had 120 marked, but had 19mm left, that would be 101mm fed through, not 99mm. 21 left over would give you 99mm
must this be done with every single nozzle size i own?
I'd say that would be a good idea for results. You can store the extrusion multipliers and K values in the slicer profile to save time later.
can you use Linear Advance with TMC2208_STANDALONE mode?
Great Video... Matt would be proud. One comment, you don't have Vernier Calipers, you have Digital calipers. The Vernier system is based on the logarithmic changes of decreasing incremental measurement. More here: en.wikipedia.org/wiki/Vernier_scale THANK YOU!!
Indeed you are correct! I used to have a nice set of vernier calipers but I had to surrender them at an airport. Poor planning :(
Calipers??? omg
After inputting my k value my printer motors speed up and slow down like it does in the k value test. Is this normal?
That's what it does. Yes.
If I remember correctly, when using linear advance, you must disable any retractions on slicer
I read that too and tried it but it gave a super stringy result.
You must not - you may however decide to reduce retraction, as linear advance may perform a small retraction on abrupt extrusion stop, which may however be borderline insufficient, as the calibration value will not be exact for every circumstance and the actual pressure behaviour is likely to have nonlinear components that are not compensated for. Note that official K factor calibration script also performs retraction! The final length of filament extruded is the same either way, but dialling retraction down can help avoid heat creep.
What is more likely to interfere is Cura's tricks like coast and extra restart after retract. Wipe is harmless and can be left in, however is deemed useless by Marlin team - however with PETG for example, i have my doubts.
Interesting. I'm using linear advance as well. However, not for speed but for increased quality at lower speeds. Which is in my opinion the most important thing.
I do have severe issues with this function. I should be using a K value of 105 with PETG filament. However, linear advance seems to ignore all jerk and acceleration limits of the extruder causing massive amounts of skipped steps on the extruder in realistic parts. I'm already down to a K factor of 45. That seams to be the maximum value.
My printer is a design of myself. Not a normal printer you can buy.
I've read that some branches of Marlin have buggy versions of LA. Also, there is a newer implementation that has a very different scale for the K value.
Frank Franky I was having the same problem with my printer, if you use the current marlin bugfix-1.1.x it uses the new version of linear advance v1.5 that respects Acceleration and jerk of the extruder by limiting the other axis. Or if you want a fully functional version of marlin just wait a little, Lin adv v1.5 will be in marlin 1.1.9 I believe.
In that case, I will be waiting for the new Marlin version. I assumed the error was going to be fixed at some point. I am surprised that they released it in its current state though. Skipping steps is a very obvious problem.
But nevertheless, it seems to work fine for other materials with lower K values. Definitely improving print quality!
Frank Franky it was designed by someone with a direct drive printer for direct drive. It works really well for bowden, but you have the problem we experienced with high K values needed with a long tube (or really with any bowden). They should release the next version soon, they postponed its release because of a intermittent bug that some people are experiencing. In Lin adv v1.5 you use k values lower than 2 for the most part (it represents mm compressed) , I needed a value higher than 240 in v1 but now I am using around 0.5 in v1.5.
Just for reference I think Prusa has developed their own branch of LA so the issues seen in the mainstream release aren't present in my machine. Fingers crossed for the new release of Marlin soon!
Not possible with 4.2.7 board TMC2225 ?
Neither of those were vernier calipers lol, but excellent video. Verniers have a vernier gauge on the side, it interpolates a 1 tenth scale of the divisions on the rule without using a dial or readout. 9 divisions on the gauge take up the same space as the 10 divisions on the scale giving you a different number that lines up for very tiny changes. Pretty cool way to get that extra decimal point, but they have to be made well.
Hi, where to find your nice clock?
It's just a $10 eBay item.
At 4:11 you held up a digital caliper and called it a vernier caliper. Vernier calipers can be read to two decimal places when measuring millimeters, and three when measuring centimeters. Then, there are dial calipers, not mentioned. However, you do have pay for quality calipers, but your not paying for digits, you are paying for accuracy, and repeat-ability.
Once again, at 4:52 digital calipers, not vernier.
However good info. thanks
Hi - where can I find 'extrusion multiplier' and 'extrusion width' settings in Cura 4.6.1 please?
Its called "flow" in Cura, and there is also an option for flow on the initial layer. If you do a google search you can search "calibrating Flow in Cura" with a 2 layer wide cube.
Peter John Joseph - thank you ever so much!
K L no problem. Had the same question myself at one point.
did you try higher acceleration values?
Not as yet.
Yeah, 20k.. and print a long (20cm long, 1cm high, 1cm wide) rectangular "box" with 1mm thick walls, open top, solid base, no infill. Maybe one of the long walls with a zig zag in the middle. Benchmark quality vs time for different speeds. Judge quality by box edges and zigzag. Time with a stopwatch. The design of the box is to make it cheap for you regarding filament and electricity costs, with a low print time.
Do you need to do this calibration for each filament? Or can I calibrate with pla and then print petg and abs with the same values?
In the next school holifdays I might pursue this, it sounds worthwhile. Especially in terms of eletricity usage.
geoBruce the Prusa blog says you are meant to vary it for each material. ABS and PLA are the same though. The test file from Matt's hub should only need the temperature changed to work with other materials.
Hey, time to make a video on Linear Advance v1.5? :)
I've been messing with linear advance on my Ender 3 (Bowden still). I spent hours tuning the K value with the calibration print, and honestly, all it really seems to do is overwork my extruder. I was seeing marginally (very marginally) better results at speeds above 100mm/s, but I typically only print at around 60-70mm/s and at that speed If anything my prints looked worse.
I was able to reduce my retraction to like 1mm though and I got no stringing, so there's that.
Is this really meant for direct drive? I've seen conflicting info. It seems like with a bowden printer it doesn't really do much since bowdens already print fast really well. Also there's a lot more "slop" in a bowden, so it seems to me like it really can't manage the nozzle pressure as well as it probably could with a direct drive setup. It also just stresses me out seeing the extruder going crazy the whole print.
Maybe I'm doing something wrong.
What stepper drivers are you using? The TMC2208s need to be in Spreadcycle to handle linear advance, and while I can't say for sure, I doubt the stock power saw-like stepper drivers can handle it at all
@@Anarasha I'm using 2208s with the extruder in spread cycle and square wave stepping enabled so that the extruder works at all with linear advance. I've recently tried it again with a bmg extruder and I got similar results except this time the bmg really didnt like the jolt of being run back and forth really fast like that. I was scared it was gonna wear out the gears and cause backlash. I think that the K value required for bowden systems just makes the extruder work really hard to maintain control over the pressure correctly. My bowden tube is less than 350mm long and I'm using a capricorn tube so I was able to get a K value of only .45 to work well, but it was still stressing me out watching my poor bmg slam back and forth everytime the nozzle rounded a corner. It was especially bad with gyroid infill.
@@WrexShepard Well, if an extruder frame can't handle the stress of 3D printing at speed, it's not worth using. I get the stress you speak of, but you just have to relax and trust your hardware
@@Anarasha Fair enough. I shall trust my little trianglelab bmg clone. You're right, if it cant handle it without wearing out it's a piece of doodoo and I need to go back to the non-geared bowden style extruder. The big ass stepper motor I have driving it has more than a enough torque to give the bmg a beating though lol.
I might go direct drive soon anyway, since I recently added a 40mm wide extrusion on my x axis and a 2nd z stepper so I think I might have the ender 3 ready to be DD without causing massive ringing issues. This should allow me to drop the K value significantly. I just have to design an elegant way to mount it with a pancake stepper for weight.
Thanks for reassuring me that I shouldn't be able to break my printer with basic functionality lol.
@@WrexShepard I managed to make it work on TMC2208 by the way. You just need to use the following two values in your firmware:
#define MINIMUM_STEPPER_PULSE 1
and
#define SQUARE_WAVE_STEPPING
With those, I have no skipped steps on a stock Ender 3 extruder with the metal frame
Is this Linear_Advance 2.0?
No. The Prusa specific version.
I feel like you should have stated the time difference on some things like that boat and stuff for an idea if it is worth it
The slicing software gives an estimate which allows back to back comparison without even printing. Unfortunately Simplify3D is not accurate at all with this, although relative comparisons can still take place.
4:15 you can get 4 0-25mm micrometer for that and more acurate. digital caliber is 0.02 tolerance you never know is it read correct unless test it,. and more you push less you get. less you push more you get.you cant get correct value and then there is that +-0.02.thats how calibers work even
Extrusion multiplier and Esteps have EXACTLY the same effect (See the documentation to confirm this). Had the math not been messed up when calculating the Esteps, there would have (or should have) been no reason to change your extrusion multiplier.
Incorrect. Although both settings result in the same effect, they are being used for different reasons. Esteps take care of how much filament to be driven through the extruder. Different filaments have different characteristics and expand differently when they leave the nozzle. Then, by experiment you choose the correct extrusion multiplier for the different filament. Otherwise you will need to change the esteps on your printer every time you change the filament.
5:49 cooling is blowing x direction. why nobody has cooling 4 direction.maybe your sides would match. think bout it lol.its cooling different on x and y in that thin wall
this causes my extruder to skip a bunch
100 dollars for calipers??? I bought my 2 decimal place calipers for around 15 dollars
Once again no cr-X
After the comment about callipers, I did some research and there is a great vid (th-cam.com/video/WvszAb0Y0Ec/w-d-xo.html) by AvE. I thought my £9 callipers were OK until watching his video and they look like mine with different branding!
5:17 you changes extruder steps even it effect this one lol.maybe your caliber was 1mm off.it extruded 100mm you measured it wrong. why so complicated. you can just change that nose size keep extrusion multiply 1 keep original correct 280 step. what if you do this step reverse order lol. yes they all effect same thing. just change one
Interesting, can you give an example?
It grinds me when people call the extruder an axis. It's not. Stop it.
I understand it's not but the explanation needs to be easily understandable. If I had said stepper motors or similar it would have been confusing for those with dual z steppers.