You have under extrusion because your line width is supposed to be 120% of you nozzle diameter. If you want to use 0.4 as a line width for a 0.4 nozzle, you need to have a machine that is calibrated very well. Not just e-steps and flow. Obviously your printer isn't calibrated this well.
But aren't you calibrating it for actually extruding material with the nozzle? In real use, the nozzle resistance is part of the problem you're trying to solve with the calibration. What purpose does it serve to calibrate without the nozzle?
@@HardFirmSoft do we recalibrate e-steps after nozzle switch or tube change? nope :) I would say nozzle resistance more about torque/physical stability of printer. And people do increase/decrease temperatures, which affects resistance, that's why it's better to exclude it. And imho, by calibrating with resistance, we add 1 more variable inside e-steps value, that affects quality.
I use a jig that holds my caliper in place and a special clamp to hold the filament onto the moving caliper jaw. Extremely accurate. Found the idea on thingiverse.
Even better use a silver marker pen for welding. It'll mark on any colored filament. Also, if extruding 100mm mark at 120. If you over extrude with tape marked at 100 you have no idea how much you overextruded by.
Tape, or a sharpie mark. Remember where on the tape. I use the bottom edge lined up at 120 or 150 and run 50 - 100mm through the heated nozzle at the typical printing temp.
YES, I have the same concerns as you. I had my flow rate down to 92% to achieve dimensional accuracy, but then every time I tried to print a bridge in my temp tower, it was under extruding.
That explanation and illustration at the end just blew my mind. Totally makes sense now why my theoretical target value for flow rate looked so bad. It doesn't account for lateral squish. Thanks for the video!
Got my e steps "right on the money" with this video. Watched other videos that went into crazy formulas and didn't even work. This did, plain and simple. Thanks.
Good observations, ability to draw independent conclusions, and trust those conclusions enough to publish the results--not automatically discard those conclusions because they don't fit the norm. That is science!
Thank you so much for posting this video. I just installed a direct drive extruder on my Ender 3 V2 and could not figure out why it was extruding so much filament. After watching this I was able to calibrate the E Stepper and thank God it worked..
Great explanation about flow rate vs. horizontal expansion. I've been wondering about this myself after experimenting with different ways to get better quality thin walled prints.
I like the fact that you investigate on the matter and you expose good arguments against it. But imo it doesn't make any sense to end the video saying there is a better way of solving this without showing or explaining anything about the proposed method. Good insight tho, will have to investigate more on the subject.
Yes! Wonderful logic and explanation, I am so glad that I found this video. I had exactly the same issues with under extrusion as you and now I have a better idea as to why. However, I am still left wondering how to calibrate flow rate? Could you point me in the right direction?
There is only one way to calibrate flow rate. Make sure your E-Steps are on par. Your printer has to push 100mm of filament into the extruder if you ask it to. After that leave flow rate at 100% and tune in horizontal expansion. But only on small parts like cubes of 20mm because once you print something bigger like 200mm your part will shrink as well. so you need to calibrate that too. In short: 1.) Make sure E-Steps are correct 2.) Print small cubes and calibrate horizontal expansion (basically scaling the part to it´s correct dimensions) 3.) Print a large part of 200mm a cross for example with 100% infill, let it cool down and measure shrinkage. Usually somewhere between .3% and 1,5% depending on the material After that you are good to go every part fill fit as long as you don´t change line width or materials
Conclusion: Line width (measured on a printed part with single perimeter) should always be greater than extrusion width set in slicer. Mathematically (min. line width = nozzle dia. + layer height) 0,4mm nozzle + 0,2mm layer height = 0,6mm line width. (Notice how you should have a smaller line width for smaller layers) Set line width to 0,5mm in slicer -->Print out single wall piece and measure 0,6mm line width, perfect settings. This will mean your layers will be bonded well, you will not experience gaps in prints and you will get reasonable well details with 0,5mm extrusion width. Now you can also easily tune your dimensional accuracy. You know that your set extrusion width is 0,5mm but real line width is 0,6mm this means you can set a XY comp. for -0,05mm. I must thank you a lot for this video, because I had always issues with my extrusions, it was either too much or underextrusion. Or the dimensional accuracy was wrong etc. I looked further into the issue after I watched your video. And finally solved it through my own conclusions and information on the internet. Here it is graphically explained why min. line width with a 0,4mm nozzle and a layer height of 0,2mm should be at least 0,6mm: dyzedesign.com/2018/07/3d-print-speed-calculation-find-optimal-speed/#:~:text=A%20general%20rule%20of%20thumb,bigger%20nozzles%20and%20layer%20height.
Ha ha, that will teach me to not view the entire video BEFORE i started to follow it step by step. DOH... at least I learnt how to get the flow rate perfect. My wall thickness was actually way out. Now it is PERFECT and consistent even though, it appears "its not the way to adjust it". Your video is really great, straight to the point, no fluffing around and at a pace one can follow. Thank you.
How you explained E steps was really helpful , i was really apprehensive on changing flow , just for the sake of changing something. I only modified Extruder E steps and happy with the fill that im getting now
In general you are right, but there is still one problem. At 80% I get the correct single line width (say for small details where I only use one single line) at 100% I get closed top shells but not the accuracy of the single shell. A 2nd adjustment one could make is at 80% increase overlap of peremeters from standart 25% to 50% this way the shells will be merged together. And decrease line width from about 0,45mm to 0,4mm the extra squish will force plastic beyond 0,4mm to say 0,43mm this way top shells will be closer together. Also increase the amount of top shells from e.g. 3 to 5. It´s worth a try as well because I get precise single extrusion lines. Please tell me what you think. What is the disavdantage of your method ? Is it more difficult to print details and single lines ?
Worked for me! I had to solve extruder slipping first (nozzle temp was too low), but after that the measurements were precise and after a couple calibrations, my prints are coming out way cleaner now. Thank you!
This was the best vid I fond regarding flow and esteps and understanding the right way to calibrate. You explained it in such a better manner than some of the other tutorials. I hope to see tons of more content from you (2 years later).
What I do is I 1) calibrate my e-steps 2) then I print a single layer disc at 100% flow rate. If I get over-extrusion (I get jagged edges on the surface), I lower my flow rate to 70% then try again. If my surface is smooth, I bump my flow rate, etc. I do a binary search of the flow rate that will be smooth but barely making jagged edges. In other words to a point where it's not creating jagged edges but if I add 1% it will. I note this value for my first layer flow rate. Then I print a 2 layer disc, and do the same with the flow rate for the remaining of the layers. So in Cura I have 2 flow rates, one for the first layer and another for the remaining layers.
In my thought proses if your first layer lines are squishing out significantly more than the subsequent layers it has more to do with the first layer height than extrusion on the first layer. If your bed is to close to the nozzle your printer will be pushing out material to fill a gap that is slightly larger than is actually there, therefore causing the lines to squish out to the side as the material has no where else to go. Simply adjusting the flow to compensate for a bed that is either to close or to far away will do more to create a slight Z axis dimensional inaccuracy than truly fix the problem you are facing. I would aim for a first layer flow that is nearly identical to the rest of the model to keep things as accurate as possible. As for the possible "elephants foot" effect that can happen to the first few layers (especially on a heated bed) I use the initial layer horizontal expansion setting to compensate for the little squish. or simply design a vey very small chamfer into the bottom edge of your models to all but eliminate that concern.
Arriving on this video just after seeing one that explains only about de flow rate calibration, so I have a good complement. After trying calibration with the expansion parameter, I notice that the corners are way more visible, also on a surface when there is a perpendicular wall behind, we see it bumping out from the outer wall very much, so might touch other parameters because of the expansion rate to reduce flow at the angles or something
i need some help maybe you can explain it to me, i have my e Steps correct on my XYZ planes and when i extrude 120mm its spot on. i then started my flow rate calibration at the end of it in cura (calculated this using Flow Rate Calibration website) and it says that my flow rate should be 83% after doing this my walls were correct but what would i need to do to get my flow rate in cura to be 100%
damn ive been doing that stupid cube thing just for you to point out every issue I've had and that i shouldn't use it and i think your absolutely correct
I've been struggling with this as a newbie (though a scientist and engineer). I had found horizontal expansion before flow rate. I have the same thoughts as you, that it may be better to use nominal flow with negative expansion. Another issue using reduced flow besides strength. If you then need to use a positive horizontal expansion to account for undershoot on dimensions, on tight tolerance parts, Cura will expand them then decide that neighboring surfaces are one solid. And print those layers as one in the slicer! I didn't notice this until printing a clearance test, even though you can see the effect in the slicer. If you use 100% flow and retract surfaces to account for bulging, it's not a problem on print in place.
I'll add to our theory with geometry. If we compare the area of a square to a circle inscribed in that square, that's pi/4 or 78%. For the few filaments across two materials I've tried to calibrate wall thickness to using flow, the result was 75%-82%. All right around 78%. Suggesting that what we're really doing then is leaving gaps in the print in favor of dimensional accuracy over firmly bonded layers. Probably there's a better optimum in between. But I agree. Leaving parts very under filled must be bad for strength even though there's lots of tutorials suggesting that's how you calibrate flow.
Thanks for this video! When i use the single layer cube to set the "correct" flow rate, everything will be underextruded resulting in bad prints. That's why i usually print with 100% flow rate anyway. But now i have to print something dimensionally accurate so i have to look into horizontal expansion.
2:52 - I also have exactly 1.5mm differences... it's weird but what could cause this? I'm thinking that the stepper motor is missing steps or software bug. I have Ender 3 V1 stock bowden extruder. I was taking measurements without using the hotend, simply directly from the extruder module. I'm thinking the best way to get proper estep calibration is to take at least 10 measurements and then obtain an average. Then use the average value for the calibration calculation.
I have an ender 5 plus, and the touch panel (not a wheel) doesn't have a section for changing the e-steps. Do you know how I might perform this calibration on an ender 5 plus?
Thank you for he simple to use instructions but I am having a crisis with my ender 5 pro. it has worked fine for about 9 month but recently it has under extruded and clogged all the time. I have switched out almost all the printing hardware and still can't get the obstructions to clear out. I tried resetting the firmware and that has led to an estep calibration problem. I ran through the calibration steps you mentioned and only made the under extrusion worse. Its now printing less filament. I don't know if there is some step I'm missing but can't understand the errors.
İ am using double wall calibration method and for a 1 week i tryed many many o f settings yo correct first layer with %88 percent flow but i couldnt its very helpful video thank you
That seems to be if you have to reduce the flow rate, as in your case, but if after the calculations you end up needing to raise the flow rate above 100% you wouldn't have that undextruding gaps when compared to this case's 100%, would you? I'm asking because for most of my filaments I end up with a higher-than-a-100% flow rate and seem to get great results.
My thinking currently is that it doesn't really matter if you change flow rate or e-steps, both affect the same thing: how much filament is extruded. I would suggest making a best effort calibration for e-steps, and then varying the flow rate based on the material you're using, because that is a lot faster to change. It will also help to keep track of the values that you've found useful for different materials.
@@HardFirmSoft - Exactly. Its good practice to use every parameter exactly to its purpose and not to mixed them up experimenting all over undesired side effects. Looking to Your Videos and conclusions, it seems that You are a very smart guy. Thank You for Your video!!
i did the 1 layer wall and got 95% , if i do all flow at that rate i do get tops under extrusion, so i just run my walls and infill at 95% and my tops at 100%, it would be interesting to see if height is same 95% as im guessing horizontal expansion wont help that...
Speed. I wanted to prent hot and fast with big nozzlles when I got mine and assumed something so simple as volumetric flow would be adjusted in cura if you changed your line settings, but nope.
You cant choose the view of the top layer cuz the extrusion width combines with layer under it and the one over it and on the top surface there is no layer over it.
I would calibrate Estep without extruding through the nozzle to be more accurate and avoid issue with temperature and hotend (but with DD configuration it’s quite hard). You should explain how to use this Expansion setting instead of flow multiplier though
But isn't the point of the calibration to account for the resistances that are in the printer? And yes, it would be a good idea to make a video about using the horizontal expansion setting.
@@HardFirmSoft In this particular case I disagree. ESteps Calibration is about the calibration of the amount of filament moved by the extruder regardless of all other impediments. Those other things should be taken into account on a filament-by-filament basis, as they will vary. Example: Printing PLA at 210 on my printer will have much less resistance than printing PLA at 195. Therefore my Slicer Settings should account for that. It might be that the printing speed needs changing, or something else, but it should not involve ESteps. You also say in this video how NOT to calibrate FlowRate, but fail to tell us how it should be done (while the video title indicates you will tell us how). I have asked in support groups, only to be told the only way this can be done is the method you say is wrong... Thank you for the video anyway.
@@HardFirmSoft Extruder motion is purely mechanical. If your printer is working properly, the extruded filament without the nozzle should be exactly the same as with the nozzle in place. If it's not, then you have a problem with your extruder (clog causing the extruder to loose steps, or your filament is slipping in the extruder gears). Either way, you're not calibrating for "resistances" in the printer.
@@cliveknowles7496 x-mm extruded is the command and if it does not extrude that much then the flow has to be cranked way up. If your extruder is slipping then you will not get the amount that you are expecting. E-staps has to be accurate for the slicer to "know" what to do. I plan to check E-steps with a couple of different materials and temperatures, but if it is significantly off then the base of the process is off, IMO.
This evening I tried to „calibrate“ my dimensions by adjusting the e-steps settings for X/Y inside Marlin firmware. And I earned very good (in terms of precise) results, at least for the test cubes. Let’s see if this will confirm also for larger objects.
My E-steps are calibrated, when I print at 100%flow 0.5 line width i measure 0.5 with a caliper but somehow when printing an object it looks overextruder how???
if i babystep z during the print will that change the wall thickness of the cube and therefore the flow rate? more squish makes a thicker wall right. so if i babystep and make less squish than the dimensions change
Why when doing this test do people use 100mm or in your case 90mm? Why not just mark off the curent mm that is set in the printer, in the case of the ender 3, 93mm so mark off 93mm on your filament then run the test. I dont get why 100mm
100mm is just easier to do math with. You can do 101, 120, 95, 20, whatever. The larger the number the more likely your math will be closer to what you need, but too large and you waste filament and time.
Very informative, thanks. Will try this on my CR10S-Pro. My outer dimensions are like within 0,1mm but holes on prints seem to be almost 0,5mm small. So "hole horizontal expansion" experiments it will be :D
I clicked this video not knowing E steps were even a thing. I'm leaving it convinced that is my problem. My prints always have trouble initially sticking to the plate for the first lalf inch of layed filament. Like it gets spat out, but doesn't stick to the plate and ends up curling. After that half inch, it lays and sticks.
I have 3 printers (CR-10S, Ender-3, Ender-3 Pro) and have done more calibration prints that I would like to admit. If you test different filament between the 3 different printers you end up getting different results anyways. If you want to change line width to 0.5 instead of 0.4 then this changes the results you get for the 2 wall calibration cube using the same flow setting. If you use different slicers you will get different results as well. I've come to the conclusion that you will never get it dialed in 100%. It's an endless rabbit hole, so my advice is don't go down it like I did, lol. I think printing something that is meant to fit together (print in place) to test tolerances and also has some top surfaces to look at would be a better test. Then adjust the flow based on the results of those prints.
Hmmmm Very interesting. Many are touting Luke's flow rate cube. I just calculated my eStep on my direct drive Ender 5. o wonder I'm having printing problems. My new Estep is 133. Yikes.
Right idea, but only extruding 90mm is the same as only looking at 90% of the test. Wouldn't it be more accurate to do 100mm for a linear average? Everyone I've seen do this test does 100.
Yes you could extrude 100 mm, but there's nothing special about that number. You could also extrude 1000 mm but how hard would that be to measure? Why I extrude less than what I marked on the filament, is to be able to take a measurement even if the printer extrudes more than what's requested. Hope that clears it up for you.
@@HardFirmSoft actually, the test is mark off 100mm, then 10mm past that and measure what doesn't extrude. It's on UA-cam, and thats pretty much how EVERYBODY else does it. But you do you.
Either way works. If you only want to make one mark and like the number 100, then you can mark at 110 and extrude 100. Technically, more will give better accuracy.
Would have been nice to see how to do this without the direct drive. I don't have that and have zero idea how to go about this. I've struggled with this damned Ender 3 for 8 months now and still don't have a working printer.
There are some mistakes in your conclusions. I did similar experiments .On 73% you get 0.4mm thickness on single wall .On 80 % double wall you get 0.8mm so the 2 walls are already squished overlapped ,so the picture is not accurate the walls at 80% are not in point contact are already overlapped. So the calibration has to be with more than 2 walls in order to be accurate and get accurate dimensions. Regarding the perceived under extrusion on the top that can be mitigated with other selections that not involve wall flow number of top/ Bottom layers, like skin overlap and top/bottom line width and top/bottom line direction ,choosing monotonic pattern.
So are you saying that you think flow rate should be adjusted using the wall thickness method? Then make other changes in settings to prevent under extrusion?
I know this is an old video but I am afraid this is not how fine tuning works. The 2 settings (flow rate and horizontal expansion) serve 2 different purposes and conflating the two is not very solid advice... Sorry! Flow rate is different for each filament. Each filament reacts different to different temperatures. All filaments have different viscosity and other different mechanical properties, making them react differently to the exact same extrusion rate. You need to test (with a single wall width) if your flow rate gets you an extrusion size identical to the nozzle width. People may ask, what about squish? This doesn't matter because on a level bed the nozzle is always the same distance (approximately) from the bed and on subsequent layers the nozzle is always an equal spacing away from the previous layer (with variable layer height the slicer will compensate with extrusion pressure). This baseline of a predictable extrusion width (measured as a single line) is important. Especially because it comes into play when setting horizontal expansion. With flow rate you are dialing in the properties of the filament, whereas with horizontal expansion your are adjusting the properties of the printer itself. If I instruct the firmware of a mechanically well-tuned printer to move a 10th of a mm more to the left, regardless of the flow rate, I can measure that difference. Whether the flow rate is spot on or 10mm off makes no difference. The difference between 2 extrusions (set with expansion settings) will always be the same. But if my flow rate isn't predictably dialed in, I can no longer say if I am measuring the effect of horizontal expansion or flow rate. Setting horizontal expansion changes the actual dimensions of the object you are printing! If all you have is an outer shell with infill (say you are printing a figurine), great you will not have much issues. If however you are printing something that has an inside and an outside dimension, you are changing both. Let's say you have a model consisting of 2 concentric rings that have a clearance between each other of .2mm and you are adjusting your horizontal expansion. It takes an adjustment of .1mm and your parts will be fused...
Bro... I appreciate that you are trying to help, but that whole explanation made zero sense. If there was any conflation going on, it's in your explanation above. However, if you actually believe you know how to properly set horizontal expansion and flow rate, please do make a video explaining how, and reply here so we can watch.
@@hamishbeaven7921 I have tried to reply twice before but I keep getting caught in YT's spam traps... Or whatever you call them. There is a link on Reddit that clearly explains the difference. I also think it's what keeps setting of the "alarms", so alas...
He went to; I "measured" 1mm, then suddenly solved math shows up. This in school is called cheating. How did you get from 1mm to 0.4/0.5 x 100, how does 1mm turn into 0.5mm?
I'd like to hear from you, what was the best calibration or tuning you've done that resulted in better printing quality?
You have under extrusion because your line width is supposed to be 120% of you nozzle diameter. If you want to use 0.4 as a line width for a 0.4 nozzle, you need to have a machine that is calibrated very well. Not just e-steps and flow. Obviously your printer isn't calibrated this well.
And tbh, e-steps cal should be done without going it though nozzle, because you calibrate length and not efficiency of extruding...
But aren't you calibrating it for actually extruding material with the nozzle? In real use, the nozzle resistance is part of the problem you're trying to solve with the calibration. What purpose does it serve to calibrate without the nozzle?
@@HardFirmSoft do we recalibrate e-steps after nozzle switch or tube change? nope :) I would say nozzle resistance more about torque/physical stability of printer. And people do increase/decrease temperatures, which affects resistance, that's why it's better to exclude it. And imho, by calibrating with resistance, we add 1 more variable inside e-steps value, that affects quality.
@@BubonicPestilence you should absolutely calibrate with the nozzle.
Using a piece of tape to mark the filament height for extruder calibration is the BEST tip I've seen in months. Thank You!
I use a jig that holds my caliper in place and a special clamp to hold the filament onto the moving caliper jaw. Extremely accurate. Found the idea on thingiverse.
Even better use a silver marker pen for welding. It'll mark on any colored filament. Also, if extruding 100mm mark at 120. If you over extrude with tape marked at 100 you have no idea how much you overextruded by.
Tape, or a sharpie mark. Remember where on the tape. I use the bottom edge lined up at 120 or 150 and run 50 - 100mm through the heated nozzle at the typical printing temp.
The top is a true “why didn’t I think of that” (-8
exactly what i was thinking!
YES, I have the same concerns as you. I had my flow rate down to 92% to achieve dimensional accuracy, but then every time I tried to print a bridge in my temp tower, it was under extruding.
That explanation and illustration at the end just blew my mind. Totally makes sense now why my theoretical target value for flow rate looked so bad. It doesn't account for lateral squish. Thanks for the video!
But where is the video on how to use the expansion?
Got my e steps "right on the money" with this video. Watched other videos that went into crazy formulas and didn't even work. This did, plain and simple. Thanks.
Good observations, ability to draw independent conclusions, and trust those conclusions enough to publish the results--not automatically discard those conclusions because they don't fit the norm. That is science!
Thank you so much for posting this video. I just installed a direct drive extruder on my Ender 3 V2 and could not figure out why it was extruding so much filament. After watching this I was able to calibrate the E Stepper and thank God it worked..
Great explanation about flow rate vs. horizontal expansion. I've been wondering about this myself after experimenting with different ways to get better quality thin walled prints.
I like the fact that you investigate on the matter and you expose good arguments against it. But imo it doesn't make any sense to end the video saying there is a better way of solving this without showing or explaining anything about the proposed method.
Good insight tho, will have to investigate more on the subject.
Thanks for your feedback, I agree that it could have been done better.
Yes! Wonderful logic and explanation, I am so glad that I found this video. I had exactly the same issues with under extrusion as you and now I have a better idea as to why. However, I am still left wondering how to calibrate flow rate? Could you point me in the right direction?
There is only one way to calibrate flow rate. Make sure your E-Steps are on par. Your printer has to push 100mm of filament into the extruder if you ask it to.
After that leave flow rate at 100% and tune in horizontal expansion. But only on small parts like cubes of 20mm because once you print something bigger like 200mm your part will shrink as well. so you need to calibrate that too.
In short:
1.) Make sure E-Steps are correct
2.) Print small cubes and calibrate horizontal expansion (basically scaling the part to it´s correct dimensions)
3.) Print a large part of 200mm a cross for example with 100% infill, let it cool down and measure shrinkage. Usually somewhere between .3% and 1,5% depending on the material
After that you are good to go every part fill fit as long as you don´t change line width or materials
Huge help with Esteps, and thank you for warding me off of wall thickness.
Conclusion: Line width (measured on a printed part with single perimeter) should always be greater than extrusion width set in slicer.
Mathematically (min. line width = nozzle dia. + layer height) 0,4mm nozzle + 0,2mm layer height = 0,6mm line width. (Notice how you should have a smaller line width for smaller layers)
Set line width to 0,5mm in slicer -->Print out single wall piece and measure 0,6mm line width, perfect settings.
This will mean your layers will be bonded well, you will not experience gaps in prints and you will get reasonable well details with 0,5mm extrusion width.
Now you can also easily tune your dimensional accuracy. You know that your set extrusion width is 0,5mm but real line width is 0,6mm this means you can set a XY comp. for -0,05mm.
I must thank you a lot for this video, because I had always issues with my extrusions, it was either too much or underextrusion. Or the dimensional accuracy was wrong etc.
I looked further into the issue after I watched your video. And finally solved it through my own conclusions and information on the internet.
Here it is graphically explained why min. line width with a 0,4mm nozzle and a layer height of 0,2mm should be at least 0,6mm:
dyzedesign.com/2018/07/3d-print-speed-calculation-find-optimal-speed/#:~:text=A%20general%20rule%20of%20thumb,bigger%20nozzles%20and%20layer%20height.
Ha ha, that will teach me to not view the entire video BEFORE i started to follow it step by step. DOH... at least I learnt how to get the flow rate perfect. My wall thickness was actually way out. Now it is PERFECT and consistent even though, it appears "its not the way to adjust it".
Your video is really great, straight to the point, no fluffing around and at a pace one can follow. Thank you.
How you explained E steps was really helpful , i was really apprehensive on changing flow , just for the sake of changing something. I only modified Extruder E steps and happy with the fill that im getting now
Great video, well explained and no unnecessary talking, thank you! :)
In general you are right, but there is still one problem. At 80% I get the correct single line width (say for small details where I only use one single line) at 100% I get closed top shells but not the accuracy of the single shell. A 2nd adjustment one could make is at 80% increase overlap of peremeters from standart 25% to 50% this way the shells will be merged together. And decrease line width from about 0,45mm to 0,4mm the extra squish will force plastic beyond 0,4mm to say 0,43mm this way top shells will be closer together. Also increase the amount of top shells from e.g. 3 to 5. It´s worth a try as well because I get precise single extrusion lines. Please tell me what you think. What is the disavdantage of your method ? Is it more difficult to print details and single lines ?
Worked for me! I had to solve extruder slipping first (nozzle temp was too low), but after that the measurements were precise and after a couple calibrations, my prints are coming out way cleaner now. Thank you!
Nice to hear it helped you out!
same for me... got horrible underextrusion by adjusting flow rate based on wall thickness... but nearly every tutorial on YT suggests this
Exact video i was looking for, it solved my issue with model, it is realy good and logic explanation.
Great to hear!
Your comparison picture with several test cube surfaces along with their according flow rate was very insightful, thank you.
wow best calibration video. The only one that helped my print quality !!!!!
This was the best vid I fond regarding flow and esteps and understanding the right way to calibrate. You explained it in such a better manner than some of the other tutorials. I hope to see tons of more content from you (2 years later).
What I do is I 1) calibrate my e-steps 2) then I print a single layer disc at 100% flow rate. If I get over-extrusion (I get jagged edges on the surface), I lower my flow rate to 70% then try again. If my surface is smooth, I bump my flow rate, etc. I do a binary search of the flow rate that will be smooth but barely making jagged edges. In other words to a point where it's not creating jagged edges but if I add 1% it will. I note this value for my first layer flow rate. Then I print a 2 layer disc, and do the same with the flow rate for the remaining of the layers. So in Cura I have 2 flow rates, one for the first layer and another for the remaining layers.
In my thought proses if your first layer lines are squishing out significantly more than the subsequent layers it has more to do with the first layer height than extrusion on the first layer. If your bed is to close to the nozzle your printer will be pushing out material to fill a gap that is slightly larger than is actually there, therefore causing the lines to squish out to the side as the material has no where else to go. Simply adjusting the flow to compensate for a bed that is either to close or to far away will do more to create a slight Z axis dimensional inaccuracy than truly fix the problem you are facing. I would aim for a first layer flow that is nearly identical to the rest of the model to keep things as accurate as possible. As for the possible "elephants foot" effect that can happen to the first few layers (especially on a heated bed) I use the initial layer horizontal expansion setting to compensate for the little squish. or simply design a vey very small chamfer into the bottom edge of your models to all but eliminate that concern.
got my printer up and running again thanks to this video, much appreciated
This is the best explanation of both calibrations I have seen!
Arriving on this video just after seeing one that explains only about de flow rate calibration, so I have a good complement. After trying calibration with the expansion parameter, I notice that the corners are way more visible, also on a surface when there is a perpendicular wall behind, we see it bumping out from the outer wall very much, so might touch other parameters because of the expansion rate to reduce flow at the angles or something
excellent explanation of flow rate with diagrams and charts!
If you choose accurate wall dimension. Cant you just compensate it by increase infill overlap or something?
i need some help maybe you can explain it to me, i have my e Steps correct on my XYZ planes and when i extrude 120mm its spot on. i then started my flow rate calibration at the end of it in cura (calculated this using Flow Rate Calibration website) and it says that my flow rate should be 83% after doing this my walls were correct but what would i need to do to get my flow rate in cura to be 100%
damn ive been doing that stupid cube thing just for you to point out every issue I've had and that i shouldn't use it and i think your absolutely correct
Such big brain! I use "extra wall between layers" or whatever and over-extrude for strength, gonna do the direct drive soon hopefully
Omg this video was so informative. I thought I was going crazy. I was about to buy another printer.
Great video. I am interested in any Neptune series calibrations and tricks you may use.
I've been struggling with this as a newbie (though a scientist and engineer). I had found horizontal expansion before flow rate. I have the same thoughts as you, that it may be better to use nominal flow with negative expansion.
Another issue using reduced flow besides strength. If you then need to use a positive horizontal expansion to account for undershoot on dimensions, on tight tolerance parts, Cura will expand them then decide that neighboring surfaces are one solid. And print those layers as one in the slicer! I didn't notice this until printing a clearance test, even though you can see the effect in the slicer.
If you use 100% flow and retract surfaces to account for bulging, it's not a problem on print in place.
I'll add to our theory with geometry. If we compare the area of a square to a circle inscribed in that square, that's pi/4 or 78%. For the few filaments across two materials I've tried to calibrate wall thickness to using flow, the result was 75%-82%. All right around 78%. Suggesting that what we're really doing then is leaving gaps in the print in favor of dimensional accuracy over firmly bonded layers.
Probably there's a better optimum in between. But I agree. Leaving parts very under filled must be bad for strength even though there's lots of tutorials suggesting that's how you calibrate flow.
Very much needed this video. Thanks buddy
Glad to hear you found this helpful!
Good video.
Definitely will be researching horizontal expansion vs flow control. Thank you.
6:36 Hats off👏👏👏 for your explanation
Thanks for this video! When i use the single layer cube to set the "correct" flow rate, everything will be underextruded resulting in bad prints. That's why i usually print with 100% flow rate anyway. But now i have to print something dimensionally accurate so i have to look into horizontal expansion.
Yes! When I’ve done esteps, then tried the single wall cube method. I get the walls to correct thickness but the walls are very broken and brittle.
I did teh same upgrade to my ender 3 v2... have you tried TPU? had any issues?... I'm still working on it..
Interesting and well presented. I have a new Ender 5, there is so much more for me to learn, I’m looking forward to it. Thanks from Orlando Florida.
Thank you so much! the information was clear and easy to follow. This is exactly what i needed.
amazing. really dense information. not a wasted moment. thank you
Thank you for the great explanation! Really appreciated the graphic at 6:44 that clearly explains the issue, cheers! :)
I tried tuning flow once and ran into the same issue with under extrusion. Will just stick with horizontal expansion for now.
👍👍😎👍👍. First video I have seen of yours. Definitely deserves a subscription from me. Looking forward to watching your previous and future content. 😀
hey i have some issue , when im lower my flow rate 100% to 73%. i have under extruison ? how can i fix ?
2:52 - I also have exactly 1.5mm differences... it's weird but what could cause this? I'm thinking that the stepper motor is missing steps or software bug. I have Ender 3 V1 stock bowden extruder. I was taking measurements without using the hotend, simply directly from the extruder module. I'm thinking the best way to get proper estep calibration is to take at least 10 measurements and then obtain an average. Then use the average value for the calibration calculation.
So what we need is s detailed method to calibrate flow rate and horizontal expansion for best surface quality and dimensional accuracy.
Yes, that would be a good resource to have. I may revisit this in the future to make a more comprehensive guide on the matter.
I have an ender 5 plus, and the touch panel (not a wheel) doesn't have a section for changing the e-steps. Do you know how I might perform this calibration on an ender 5 plus?
This is an awesome video! Explains everything really well.
Thank you again and again!!! You just made my day. Working fine now.
Thank you for he simple to use instructions but I am having a crisis with my ender 5 pro. it has worked fine for about 9 month but recently it has under extruded and clogged all the time. I have switched out almost all the printing hardware and still can't get the obstructions to clear out. I tried resetting the firmware and that has led to an estep calibration problem. I ran through the calibration steps you mentioned and only made the under extrusion worse. Its now printing less filament. I don't know if there is some step I'm missing but can't understand the errors.
İ am using double wall calibration method and for a 1 week i tryed many many o f settings yo correct first layer with %88 percent flow but i couldnt its very helpful video thank you
That seems to be if you have to reduce the flow rate, as in your case, but if after the calculations you end up needing to raise the flow rate above 100% you wouldn't have that undextruding gaps when compared to this case's 100%, would you? I'm asking because for most of my filaments I end up with a higher-than-a-100% flow rate and seem to get great results.
My thinking currently is that it doesn't really matter if you change flow rate or e-steps, both affect the same thing: how much filament is extruded. I would suggest making a best effort calibration for e-steps, and then varying the flow rate based on the material you're using, because that is a lot faster to change. It will also help to keep track of the values that you've found useful for different materials.
@@HardFirmSoft - Exactly. Its good practice to use every parameter exactly to its purpose and not to mixed them up experimenting all over undesired side effects. Looking to Your Videos and conclusions, it seems that You are a very smart guy. Thank You for Your video!!
Can I use Horizontal Expansion if I am using the Arc Welder feature in Cura?
i did the 1 layer wall and got 95% , if i do all flow at that rate i do get tops under extrusion, so i just run my walls and infill at 95% and my tops at 100%, it would be interesting to see if height is same 95% as im guessing horizontal expansion wont help that...
Awesome video~
Speed. I wanted to prent hot and fast with big nozzlles when I got mine and assumed something so simple as volumetric flow would be adjusted in cura if you changed your line settings, but nope.
You cant choose the view of the top layer cuz the extrusion width combines with layer under it and the one over it and on the top surface there is no layer over it.
THANK YOU SO MUCH for this AWESOME VIDEO thank you
wish you would have told us how to use extrusion multiplier
I would calibrate Estep without extruding through the nozzle to be more accurate and avoid issue with temperature and hotend (but with DD configuration it’s quite hard). You should explain how to use this Expansion setting instead of flow multiplier though
But isn't the point of the calibration to account for the resistances that are in the printer? And yes, it would be a good idea to make a video about using the horizontal expansion setting.
@@HardFirmSoft In this particular case I disagree. ESteps Calibration is about the calibration of the amount of filament moved by the extruder regardless of all other impediments. Those other things should be taken into account on a filament-by-filament basis, as they will vary. Example: Printing PLA at 210 on my printer will have much less resistance than printing PLA at 195. Therefore my Slicer Settings should account for that. It might be that the printing speed needs changing, or something else, but it should not involve ESteps.
You also say in this video how NOT to calibrate FlowRate, but fail to tell us how it should be done (while the video title indicates you will tell us how). I have asked in support groups, only to be told the only way this can be done is the method you say is wrong...
Thank you for the video anyway.
@@HardFirmSoft Extruder motion is purely mechanical. If your printer is working properly, the extruded filament without the nozzle should be exactly the same as with the nozzle in place. If it's not, then you have a problem with your extruder (clog causing the extruder to loose steps, or your filament is slipping in the extruder gears). Either way, you're not calibrating for "resistances" in the printer.
@@cliveknowles7496 x-mm extruded is the command and if it does not extrude that much then the flow has to be cranked way up. If your extruder is slipping then you will not get the amount that you are expecting. E-staps has to be accurate for the slicer to "know" what to do. I plan to check E-steps with a couple of different materials and temperatures, but if it is significantly off then the base of the process is off, IMO.
This evening I tried to „calibrate“ my dimensions by adjusting the e-steps settings for X/Y inside Marlin firmware. And I earned very good (in terms of precise) results, at least for the test cubes. Let’s see if this will confirm also for larger objects.
What was the result with bigger prints? Your method working well?
Does this work for the anycubic Kobra neo printer model
Beautifully explained, thanks!
I have terrible stringing from extra oozing from my nozzle, retraction maxed out it doesn't change, should i bump flow rate down?
I think you should try the cube tests with 2 and 3 walls and let the printer tell you what to do with the flow rate.
Every time I do wall calibration 0.4 example I ended up with weak wall for underextrusion.
My E-steps are calibrated, when I print at 100%flow 0.5 line width i measure 0.5 with a caliper but somehow when printing an object it looks overextruder how???
I have had experience with some filaments that might need the flow decreased a bit to look good. Is this happening with all materials?
@@HardFirmSoft yes, with all my PLAs, the holes are smaller. To improve this I need to set a negative horizontal expansion
nice video, good explanation
Very cool! Thank you for the info!
if i babystep z during the print will that change the wall thickness of the cube and therefore the flow rate? more squish makes a thicker wall right. so if i babystep and make less squish than the dimensions change
No.
YES, THAT IS EXACTLY WHAT I WAS DEALING WITH
So using flow rate is not good, it is better to use horizontal expansion..and then you dont how us how to do it? Or you showed it and I missed it?
Can I just set my ender 3 on 80% then ? all ender 3 should be the same .. ish . What does it depend on ?
Not all the same. Do the tests. E-steps, then flow rate.
Why when doing this test do people use 100mm or in your case 90mm? Why not just mark off the curent mm that is set in the printer, in the case of the ender 3, 93mm so mark off 93mm on your filament then run the test. I dont get why 100mm
100mm is just easier to do math with. You can do 101, 120, 95, 20, whatever. The larger the number the more likely your math will be closer to what you need, but too large and you waste filament and time.
Very informative, thanks. Will try this on my CR10S-Pro. My outer dimensions are like within 0,1mm but holes on prints seem to be almost 0,5mm small. So "hole horizontal expansion" experiments it will be :D
I'm glad to help!
I clicked this video not knowing E steps were even a thing. I'm leaving it convinced that is my problem. My prints always have trouble initially sticking to the plate for the first lalf inch of layed filament. Like it gets spat out, but doesn't stick to the plate and ends up curling.
After that half inch, it lays and sticks.
How do you do esters on the kobra
I have 3 printers (CR-10S, Ender-3, Ender-3 Pro) and have done more calibration prints that I would like to admit. If you test different filament between the 3 different printers you end up getting different results anyways. If you want to change line width to 0.5 instead of 0.4 then this changes the results you get for the 2 wall calibration cube using the same flow setting. If you use different slicers you will get different results as well.
I've come to the conclusion that you will never get it dialed in 100%. It's an endless rabbit hole, so my advice is don't go down it like I did, lol.
I think printing something that is meant to fit together (print in place) to test tolerances and also has some top surfaces to look at would be a better test. Then adjust the flow based on the results of those prints.
Hmmmm Very interesting. Many are touting Luke's flow rate cube. I just calculated my eStep on my direct drive Ender 5. o wonder I'm having printing problems. My new Estep is 133. Yikes.
Right idea, but only extruding 90mm is the same as only looking at 90% of the test. Wouldn't it be more accurate to do 100mm for a linear average? Everyone I've seen do this test does 100.
Yes you could extrude 100 mm, but there's nothing special about that number. You could also extrude 1000 mm but how hard would that be to measure? Why I extrude less than what I marked on the filament, is to be able to take a measurement even if the printer extrudes more than what's requested. Hope that clears it up for you.
@@HardFirmSoft actually, the test is mark off 100mm, then 10mm past that and measure what doesn't extrude. It's on UA-cam, and thats pretty much how EVERYBODY else does it. But you do you.
@@davidharris5045His way is simpler - only one mark needed. I like it!
Either way works. If you only want to make one mark and like the number 100, then you can mark at 110 and extrude 100. Technically, more will give better accuracy.
That intro reminded me of someone reading a sign with a gun pointed at them
how to calibrate flow rate then?
A well done video. :)
horizontal expansion is a setting on what slicer guys?
You can find it Cura
Would have been nice to see how to do this without the direct drive. I don't have that and have zero idea how to go about this. I've struggled with this damned Ender 3 for 8 months now and still don't have a working printer.
Make a video about octoprint!
Thanks for the suggestion! I've been thinking about it and probably will try that at some point.
hi, where can i find the file for the cube? thanks
The calibration cube can be found here www.thingiverse.com/thing:1278865 and a the simple cube here www.thingiverse.com/thing:477
@@HardFirmSoft thanks alot
There are some mistakes in your conclusions. I did similar experiments .On 73% you get 0.4mm thickness on single wall .On 80 % double wall you get 0.8mm so the 2 walls are already squished overlapped ,so the picture is not accurate the walls at 80% are not in point contact are already overlapped. So the calibration has to be with more than 2 walls in order to be accurate and get accurate dimensions. Regarding the perceived under extrusion on the top that can be mitigated with other selections that not involve wall flow number of top/ Bottom layers, like skin overlap and top/bottom line width and top/bottom line direction ,choosing monotonic pattern.
So are you saying that you think flow rate should be adjusted using the wall thickness method? Then make other changes in settings to prevent under extrusion?
I can't find the option in my printer to extrude like you do. Edit: I found it. Thanks for the video btw.
Thank you.
I know this is an old video but I am afraid this is not how fine tuning works. The 2 settings (flow rate and horizontal expansion) serve 2 different purposes and conflating the two is not very solid advice... Sorry! Flow rate is different for each filament. Each filament reacts different to different temperatures. All filaments have different viscosity and other different mechanical properties, making them react differently to the exact same extrusion rate. You need to test (with a single wall width) if your flow rate gets you an extrusion size identical to the nozzle width. People may ask, what about squish? This doesn't matter because on a level bed the nozzle is always the same distance (approximately) from the bed and on subsequent layers the nozzle is always an equal spacing away from the previous layer (with variable layer height the slicer will compensate with extrusion pressure). This baseline of a predictable extrusion width (measured as a single line) is important. Especially because it comes into play when setting horizontal expansion. With flow rate you are dialing in the properties of the filament, whereas with horizontal expansion your are adjusting the properties of the printer itself. If I instruct the firmware of a mechanically well-tuned printer to move a 10th of a mm more to the left, regardless of the flow rate, I can measure that difference. Whether the flow rate is spot on or 10mm off makes no difference. The difference between 2 extrusions (set with expansion settings) will always be the same. But if my flow rate isn't predictably dialed in, I can no longer say if I am measuring the effect of horizontal expansion or flow rate. Setting horizontal expansion changes the actual dimensions of the object you are printing! If all you have is an outer shell with infill (say you are printing a figurine), great you will not have much issues. If however you are printing something that has an inside and an outside dimension, you are changing both. Let's say you have a model consisting of 2 concentric rings that have a clearance between each other of .2mm and you are adjusting your horizontal expansion. It takes an adjustment of .1mm and your parts will be fused...
Bro... I appreciate that you are trying to help, but that whole explanation made zero sense. If there was any conflation going on, it's in your explanation above. However, if you actually believe you know how to properly set horizontal expansion and flow rate, please do make a video explaining how, and reply here so we can watch.
@@hamishbeaven7921 I have tried to reply twice before but I keep getting caught in YT's spam traps... Or whatever you call them. There is a link on Reddit that clearly explains the difference. I also think it's what keeps setting of the "alarms", so alas...
by the way nice video of DD ender 3
You are over thinking the flow rate.... it will change anyway depending on temp and material .....
A video saying how-to-calibrate-flow-rate which tells you how NOT to calibrate flow rate, and then goes on NOT to tell you how it should be done.
So my flow rate is 36% ? Thats weird but ok
I dont have those e step settings
Are you on Cura? If so, click the 3 lines beside the search box in the print settings and make sure "All" is selected.
What printer are you using?
@@HardFirmSoft E3 pro. I just had to flash new firmware and now it works
Okay, good thing you got it sorted.
He went to; I "measured" 1mm, then suddenly solved math shows up. This in school is called cheating. How did you get from 1mm to 0.4/0.5 x 100, how does 1mm turn into 0.5mm?
I always come back here because i always forget the formula😅