ฝัง
- เผยแพร่เมื่อ 23 เม.ย. 2024
- In this video, I'll explain how to calibrate your 3D printer's steps for dimensional accuracy of the X Y and Z axis.
Digital Caliper: geni.us/digitalcaliper
eSteps calibration sheet: crosslink.io/downloads/misc/C...
CHEP Calibration CUBE: www.thingiverse.com/thing:318...
Video Chapters:
0:00 Intro
0:40 Why is dimensional accuracy important?
1:43 What are steps aka steps/mm?
2:26 Other factors influence dimensional accuracy
2:56 Step by step guide to do the steps calibration
5:56 steps calibration calculator spreadsheet
8:10 Saving new steps back to the printer
9:22 Validation of new values, finalizing the calibration
THE PRINTERS I USE: kit.co/danielcrosslink/the-3d...
MY MERCH: crosslink.io/shop
BECOME A PATRON: / crosslink
CHAT ON DISCORD: / discord
BUSINESS INQUIRIES ONLY (OR IF YOU WANT TO SEND SOMETHING) - DANIEL@CROSSLINK.IO
FOLLOW ME:
INSTAGRAM - / danielcrosslink
TWITTER - / danielcrosslink
FACEBOOK - / danielcrosslink
WEBSITE: crosslink.io
The Music I use: crosslink.io/es, AMAZING for TH-camrs!
GEAR USED
kit.co/danielcrosslink/my-vid...
Disclaimer: As an Amazon Associate I earn from qualifying purchases. - วิทยาศาสตร์และเทคโนโลยี
As @hunterisu999 stated correctly in his comment, I accidently used the term "esteps" for "steps", where the term "esteps" is really just used for extruder motor steps and not for the other axises.
That is a common mistake. I do that also most of the time by accident. However printing a cube to calibrate the printer is just wrong as you add so many variables in the mix that have nothing to do with steps and mm. steps/mm is just that. No slicer, no printing, no elephant food, no overextrusion that causes problems with the measurement. You just need to measure the movement and use that as a base for the calculation. If the cube is to big after the extrusion you need to fix that in the slicer to compensate filament expansion/shrinkage.
@@guidomersmann9744 I agree. This is the wrong methodology for calibrating X, Y and Z distances. This method can only achieve a first order approximation due to the uncontrolled variables and limited test cube size. It would only be helpful if you had a severely incorrect configuration. It's better to accurately measure the movement across the full length of travel. Yes, that is harder (especially Z), but if you're going for accuracy ...
@@karimvirani842 I don't think you need to measure across the entire axis. Sure the more you measure the higher the accuracy is as it compensates measuring errors. If you measure with an fixed apparatus like mine (www.thingiverse.com/thing:4599123) this is at maximum low anyway. Especially as you can repeat the test multiple times at the very same spot. If the printer has different results on different lengths on the same axis there is something wrong you cannot compensate with the motor steps. Like a broken belt missing a tooth or being unevenly stretched.
@@guidomersmann9744 any more info on this...tried the link but its gone, your method sounds far more accurate, but I'm not totally sure how I'd do it...
Thank you so much for this! Your the best! I nominated you for the 3d printing awards this year for influencer of the year!
Exellent video. I had problems with part accuracy and calibrated my printer now the issue is gone. :)
There are several videos on correctly setting the extruder steps but I have never seen anyone address the separate axis' steps, which is something that I have done with both my printers. Good job.
Just wanted to say thanks for this info and the spreadsheet.
This immensely helped. the M92 XYZ stuff made my cubes within +-0.01, Just have to tune the flow and I'm looking good.
Thank you!
Dude!
Great explanation, solved my problem right away
I used to print twice my pieces and use a formula to scale to the ritght measure, not anymore
thanks so much
Excellent! Thank you, just what I needed!
Isn't the steps for X,Y & Z functions of the physical properties of the motors i.e. Steps/rotation, pulley size, etc?
Love your 3d printer channel.
Thank you excellent video. The excel spreadsheet really helped make it easy.
You. Are. FABULOUS!!! Thank you so much!!
You answered all my questions!!! Thank You!!!!
Awesome thanks. Just got my anet a8 used and have been printing upgrades for it and fixing what the previous owner did to it.
Invaluable info! Cheers!
6:48 I'd love to see a video from you on how to set up Octoprint for the Anet A8!
Hi, great video, if I want to calibrate with a 100 mm tower (xyz) for a more accurate callibration to I need to multiply cell B8 by 100 instead of 20?
Love your videos. Keep it up. God bless you.
Thanks so much for this! So simple!
Great Video!! Thanks for the upload!!
This is masterpiece.. :).. thanks dear for the clear demonstration and working file.
You are welcome 😊
Thanks, this helped me get my printer dialed in.
you're the best
Well explained, thank you
Great video! I've read online a few different arguments about how you shouldn't calibrate the steps for the x y and z as they are already calculated based on the physical components of the printer ie belt pulley size, the specific stepper motors involved ect.. I think right now the popular idea is that these inaccuracies should be fixed in your slicer with horizontal expanse.. however this doesn't make sense to me when I move my x y or z 100mm whatever axis I decide to move should move exactly 100 mm.. and on my ender 3 pro it doesn't.. sure it's close but it's not perfect.. while I've always considered calibrating the e-steps for these axes I've never actually done it as I've never heard of anyone else doing it.. but I'm all now lol
True, many "engineers" on these 3D sites say what you have pointed out. What they never address is, if what they say is true, why does making these step adjustments correct some of the inaccuracies.
That makes no sense. Not all belts are perfect. Not all motors are perfect. Not all couplers are perfect. Sure they're 99%, but that little error adds up. My 20mm cube came out to be 20.23mm. slightly off, but off nonetheless. My "perfectly calibrated out of the box" extruder was extruding 91mm instead of 100mm. It's all fixed now and everything's how it should be.
@@joblessalex oh no.. for sure you need to calibrate the e steps coming out of the extruder.. And you're right, not all belts and pulleys are perfect.. however, I think trying to calibrate the e steps for your x-y and Z probably aren't worth any possible gains you could get..
@@danielrdugan I calibrated all of em and now everything's perfect. 20.0X error on all sides. Definitely worth calibrating every single setting on your printer.
@@danielrdugan The greatest gain would be the fitment of parts.
Hallo and thank you for the Video.
I wondered: Does this calibration stay accurate if I change the location of the printer? For example, for it to have access to the computer I have to place it on the table next to me, which is not perfectly even. If I mount the printer back on its usual location, does the calibration stay accurate?
nice video you explain so good. Medal of Honor for you. Appreciate your Excel. Greatings
Great video and thank you for sharing your knowledge. I have a doubt, when I send, in the terminal, M503, the Anet A8 answer me "Recv: epr: ..." instead "Recv: echo: ..." and the parameters are shown different as you. Is it possible that I have a different configuration of the Octoprint? or perhaps do I have sending an order to a different part of the memory of the Anet?
Thanks and kind regards. Lorenzo.
fantastic guide, thank you for the excell!
Loved the video, but, your link for the .xlsx estep calculator is not working? please advise when fixed. thanks
Other videos I have seen say the faces the caliper are touching are the axis being measured, like you did for the Z axis. Your video says what is on the face you are looking at is the value measured. Is the CHEP cube different than other calibration cubes?
Hello mate, thanks for the video. I downloaded octoprint to windows 10 which was a right nightmare but got it working eventually. Then, I followed your steps and the printer tried to run itself off the belt, kept printing long diagonal lines and messed up everything on the printer. Luckily I could reset the eeprom chip and now it is back to it's normal self. I have a biquB1. Surely this should be possible to use on every printer?
Thanks a lot ... this Video is spot on. Is it ok to have manual mesh bedlevel turned on?
Sure, not a problem
Thank you for the time and effort you put into this! I have been using an Ender 5 Plus for years now and decided to upgrade to the silent board. I never had to calibrate anything as I was happy with the results from the factory however I decided to calibrate everything to the new board and this guide along with your excel sheet was extremely helpful! Thanks again!!
Thank you very much I am grateful to you
This video is completely overlooked and does not have enough likes! Thank you so much for this extremely helpful content.
What slicer did you use to on the g.code that is on thingiverse? it came out perfect!!!
Thank you for your instructions.I have 3 ender 3 v2 . I use your calibration cube instructions. 2 of them worked perfect . The last one no luck . I can't get a constant measurement. I suspect the motherboard is defective. I adjust the vref voltage for all the motors. Same problem. I give up ... I wait for new motherboard upgrade from 4.2.2 to 4.2.7
Nice thank you for your video. I setup my printer accordingly and get the 20 mm on all sides.
After I try a 100 calibrating cubes, i was very frustrate with my 3 D printer. I saw your Video, download the sheet. Make 2 Probes.....and Super i get it!!!!!! A Thousand, a thousand and a thousand thanks!!!!!!
Glad I could help
hi. great job, thanks. how can i do this without octo?
LIFE SAVING VIDEO!!!1
wow, thank you. You explained that so well. I was looking for a video just like this, and the spread sheet was icing on the cake. Thank you.
Thank you very helpful
Are these gcodes for any printer? I have a anycubic kobra go but don’t want to do something that’s going to mess it up.
AWESOME video! I have printed that cube on each of my printers but never saw how to fix them, till today. Thanks!! I have calibrated my extruder, but am very much looking forward to how you do it now.
Glad I could help!
the sheet is not working for me... am i missing a program to open it up with?
@@Rage123m It needs Microsoft Excel or another spreadsheet application to be used.
Great video this did solve my issue since i was experiencing this. However, the part that was printed on my build plate has now been printed towards the bottom right instead of the center of the origin. When I changed my steps the values given were X106.67 Y103.23 Z634.92. The original steps were X80 Y80 Z400. The print is perfect in dimensions but I want the print to be printing in the center of my build plate instead of the bottom right because I think the increased values caused this shift in direction. Do you know a fix to this? Again, thank you for the guide as this saved me tons of time and frustration!
Let’s start with, where does the printer go if you send it to 0/0? Is it really in the corner?
@@Crosslink3D Well I haven't gotten to do that yet so tomorrow I will make sure to do so. Ill update you about what happens. Thank you for helping me to diagnose this problem
@@Crosslink3D Ah i fixed this issue. Turns out cura has a slicer setting where you can change the x and y offset so i just measured how far it was off then input those values and bam, it prints in the middle now
Vielen Dank, Daniel! Ausgeseichnet Video. Mit Freundlichen Grueseen von Romeo, Michigan, USA
Thanks gute Arbeit Bruder
Hi. In my experience there always is a "elephant foot". So if you put the caliper over the whole height, you are measuring the elephant foot too, which is misleading. You should start measuring only above ~5mm height - don't you agree?
If you are using Cura, you can change Initial layer expansion to -0.1 to reduce elephants foot.
Your offset for your nozzle is too low if you get elephants foot
Save yourself time and headache and measure the movement of the print head, not the prints.
great share, thanks.
Daniel, great video thanks. I've got as far as trying to find what my current values are but when i put the m503 code into the console i just get an error message. I have a ender 3 v2 linked to a Creality Sonic Pad. Any ideas?
It might be that your firmware doesn’t have that enabled. There is a way to disable M503 in Marlin and maybe for some weird reason Creality chose to do that. Maybe get a fresh firmware installed first from the firmware factory and then try again. Let me know if in doubt
@@Crosslink3D I think I've found it now using the printer.cfg file on the sonic pad.
Thanks again
Thank you, I love you
This is the first video I've found to explain how to make the changes. The spreadsheet is an added bonus!! Thank you
hi man, could you make a video about how to fix the Bltouch homing problem? I have marlin 2.0.x and my Z homing doesn't stop when the sensor detects the hotbed and it keeps going. I know that there is a fix for marlin 1.1 but nothing about marlin 2. Thank you
i had the same issue on my E3P (with an SKR E3 Mini board) after installing it. tried over a dozen different firmware builds too. After VS code drove me over the edge, i used Wordpad (!) to edit the files, then VSCode to build the firmware, worked perfectly after that.
any way to calculate mm per step instead? Thanks man
I put in the new esteps with Pronterface, but 503 returns the old value. My printer board seems to be locked. Does that make any sense? Thank you for any help.
Would I be correct in assuming that the formula you used, only works for a 20x20x20 cube. I actually printed quite a few cubes and was able to get very close to the actual settings. However I tried to use the same sheet to try and calibrate a 100x100x100 print in order to try and get closer still, but all the values I received, were nowhere close to what they should be. As an example, the current e step for the X-axis is 100.0 the actual measured dimension was 99.57, but the spreadsheet is indicating that the new e-step should be 20.09. What is going on here???? Thank you for the time you will take to address my concerns.
The formula is made for a 20 inch cube. By the looks of the formula, if you change the '20' multiplier to a '100' (since youre doing a 100 mm calibration), you should be good to go. But I agree, a 100 mm calibration will be much more precise.
Stop reading my mind of the information that I am in need of at the time :-) 👍👍
Hi Crosslinnk, thank you for this video, i have come into a problem. I have a Anycubic Vyper that I have never had adhesion issues with, I created the initial cube dropping my bed temp to 50c instead of the usual 60c I use with PLA +, the cube printed well and the measure dimensions were x 20.14 y 20.10 z20.06, I made the changes suggested by the spread sheet and reprinted the cube, This time 4 times in a row the cube lifted from the bed at about the 25% mark. Do you know what would have caused this?
I would assume that changing the steps for x/y/z will maybe require adjusting the z-probe offset a little. Did you try to go back to the original values? Does that fix the problem?
ANY PLANS FOR A Delta CALIBRATION TUTORIAL?
Regarding your advice at 3:20-- How would I control how far I squeeze down the filament?
Use the Z offset.
ALL THIS TIME and I never knew there were "steps" along with "esteps". I thought that only esteps could be adjusted. Hopefully this fixes the issues I'm having with printed holes not being the right dimensions for magnets and other such items.
super!! thanks
Question: can you read my mind? I just got a digital measure device (forgot the name for a bit) and my 3d printer is off by 0,07mm.
Thanks!
Would you have to take shrinkage into consideration? Calibrated for pla. How about for other materials?
Great video, thanks! I also recently calibrated my steps (XZY and E), but using a different approach (measuring the axis movement directly on the sled with a caliper, which I think is more accurate). But now here's my question: how does the change in the step values affect the magic numbers for a 3d printer, or does it at all? I've been looking for an answer to this in many videos and texts but haven't found one yet. Would be awesome if you could help out with that!
The value has to do with the gear ratio and your micro step value. there are plenty of calculators on line for "steps" per millimeter (which are actually micro steps per millimeter). The larger the pully(more gear teeth) the less steps per millimeter. but the method above is using a relative ratio to calibrate. You are definalty correct measureing the actual movement is more accurate. getting acurate measurements with callipers can be difficult its best to use a dial guage.
Would you recommend xyz steps first or esteps first?
It really doesn't matter. One doesn't affect another.
One thing I NEVER hear anyone talking about when calibrating a 3D printer is something I fell into completely by accident, and that is temperature. I would print things off that fit well right away but the next day they didn't fit. Couldn't figure it out and I kept running test cubes which all came out correct. Then I measured one I had done the day before but one that was now completely cooled down to room temp and it was measurably smaller than the freshly printed, and still warm, ones. Depending on the amount of infill or number of walls I would see variations from between 0.1 and 0.4mm in each axis. No amount of recalibration would ever get the parts to an accurate dimension...until I let them cool to ambient and then remeasured. I also found that using my calipers at ambient temperature made a difference as well....as much as an additional 0.1mm with just threw off the cube measurements even more.
Interesting observations you have there. Absolutely want to verify those for sure. I am also curious what happens to parts of they get warmed up again. Do they expand and how much
I understand that steps/mm are factory set and normally they don't need to be adjusted, I found this relevant as I've swapped my idler pulleys to 20 tooth pulleys from the standard "Top hat" bearings fitted face to face. The bearings are 1.5mm bigger on the diameter and thus with the new pulleys the print size is reduced in size both in X and Y. So thank you for a great video which I found very informative.
After calibration, i have elephant foot. Can u tell me why? i'm using cura and i set initial horizontal expansion to -0.15
Adjust the Z offset.
How do you correct if different sized objects are off by differing percentages?
By measureing the actual mechanical movement of the hotend, not the prints. The way in the video is not proper method.
Link to spreadsheet is giving me an error that location doesnt exist?
how much infill to use in the cube when printing it
15% or more
What about the effect of elephants foot? Wouldn’t that bias some of the measurements?
Sure, the z can be affected by this so you’ll have to make sure you’re not squeezing the first layer so much. Printing a much higher part for measuring makes more sense also
@@Crosslink3D All other measurements are affected by this also. But only if you measure across the foot. People need to realize this and measure above the foot.
I tend to use a 100x100mm square for x and y. The larger the object the more accurate your settings will be. Once x and y are good i would only do z if I thought it was seriously out.
I normally use a 20x20 cube, but when I saw your message regarding the 100 I tried it and after entering the new settings the bed became smaller and was banging into the front y axis tensioner. Soon as I reverted back to the 20x20 settings it reverted back to normal. Love to hear an explanation for this.
@@highroyds Did you adjust the formula to the biger test cube, or did you use the same spread sheet from the video? Last one would give you very wrong numbers.
@@nel2834 I havent used this formula for a long time, Once I got it all dialled in I left it a lone lol. Not had any more issues and all the prints have been virtually spot on.
One of the things that confuses me. About calipers. Which part you use? The tip and the main part (where you show measuring) are actually different. (I know slightly different topic … but yours are similar to mine. There’s a small step at the end part
ive come to a big problem lately i need to calibrate e step but with a bigger model since its allway off in a bigger print
so i need a new sheet but for a 100mm print its my need right now :S
still havent found a clue how to achieve it 4 month later
You saved my life. My brand new ender pro 3 was printing everything small, my first 20mm cube was 15mm. I’ll be calibrating my extrusion next. Thank you!
I know this is an old comment, but wow! That printer came out of the box WAY out of calibration. My CR-10 SE that I just got had some issues as well, but nothing like what you had from your brand new printer...
Does this work for Delta Printers?
Indeed a very good question. What does M503 return on a delta?
i can't type my values into the little box... what am i missing?
spool holder link please. at 1:27.
at 127
Stay have a problem with 90 digree ptinting i can't geat a sharp 90 digree if i want to print a cubic and i don't know how to fix the X Y acces!
Print slower or use pressure advance (Linear Advance feature in firmware).
@Siana Gearz do you know how? Because i did adjust the flow but still a same problem
@@BeetnaOurHome What printer?
@@SianaGearz Anycubic chiron
But how do you get that number
Why don’t they just have internal X Y Z offset values like a cnc mill or lathe??????
I've finally had it become an issue
I'm making larger mechanical parts
and everything is slightly elliptical and doesn't perfectly line up with other parts
I could loosen my tolerances
but I already know, for a fact, my printer isn't dimensionally accurate, so I should try that first
the calibration sheet is borked
HELLO EVERY TIME I TRY THIS I NEED TO RE LEVEL MY BED AND MY PRINTS COME OUT WORSE WOULD YOU HAPPEN TO KNOW WHY ??? :/
😍👌👍
where's the spreadsheet?
The link is in the video description
You should ALWAYS start with extruder calibration first!
Ok so this kind of calibration based on printed parts like a cube might work for small prints but what happens after you calibrate this way and then setup a print with 2 cubes with 60mm of empty space between them? Does the calibration still hold accuracy? I calibrate my printer with a digital indicator like a machinist would and so far I trust that more than any other method.
It seems, every time I measure, the caliper gives me different measurements for each side x_x
There is always little differences and this also depends on the precision of the caliper
@@Crosslink3D Should I go with a sort of average? Or maybe the lowest/highest measurement?
Ohhh noooo, this information for the X, Y and Z calibration is soo wrong and will cause issue on larger prints.
The pulleys are toothed, therefore we will ALWAYS get the same distance of travel per rotation.
The number of steps per rotation of the stepper motor is always the same. based on this the X and Y axis can be calculated from design and will be accurate for all sizes of prints.
The same applies for the Z-Axis screw and again it should be calculated from design and will be accurate.
An over dimension of a test part is always from over extrusion of the extruder.
If you change something on your printer like the size of the pulleys, and being a toothed pulley, it's diameter and circumference is easy to calculate, that's when the steps per mm for that axis needs to be changed.
My first printer I ever owned was a kit pre calibrated at the factory using this method and then locked and caused me no end of trouble printing ovals instead of circles and having bigger errors at larger sized prints.
Most (not all) standard printers like the CR10, the Anet A9 and the Wanhoa duplicator series printers all use the same size pully, based on that the X and Y axis steps per mm is 80
For the CR10 and Wanhoa, the Z is 400.
The Extruder will vary a bit from machine to machine.
Over extrusion can also cause holes to be filled in more than they should be but in some cases, you may need to adjust the horizontal expansion to get dimensionally accurate holes, this is not supported in S3D but is in Cura.
To prove the point, print out your 20x20x20 test cube, then scale it up in the X and Y axis to 150 x 150 x 20 and compare your results.
If your method was correct, then that should both be accurate, but I'm gonna bet the 150 x150 x20 will be under sized on the X and Y and not even the same amount of under size.
Ray
So true.
Incidentally, E-step is short for Extruder Steps per mm. :-)
What you stated is true, however you are assuming all of your components are within 100% calibrated tolerances. The method described in the video calibrates for minor intolerances of hardware, wear, and filament. I wouldn't be surprised if only 1 out of 1000 pieces of hardware are being measured out of the factories in china with a passing measurement of 99.5% accurate. The design is a starting point of measurement not the finalized numbers.
@@Yakena1 You are most likely correct, however the tolerances of things like the stepper motors number of steps per rotation and the number of teeth in the toothed sprocket for the GT2 belts are going to be at a very high tolerance level as these are full unit inter values and do not change.
To get any value out of any calibration where there is a fixed limit such as the X, Y and Z travels, you need to calibrate over at least 80% of the full travel, not 10% as is done printing a 20 mm x 20 mm x 20 mm cube.
You also need to eliminate anything that can produce an error if possible such as the amount of filament that is extruded, and this one does vary a lot and is the most common cause of errors in printers.
There are 4 settings we are looking at here and this video covers the first 3 using the name of the last one.
We have the X-Steps, the Y-Steps, the Z-Steps and finally the E-Steps.
Typically, the error in the X, Y and Z steps are extremely low to almost non-existent.
E-Steps on the other hand does have a history of being out of calibration even between batches of filament. and the ware in the extruder feed gear.
If you are going to follow a video like this one.
Calibrate the E-Steps first.
Make sure your slicer is also calibrated, and the test your X, Y and Z on a much larger print so as to cover at least 80% of the axis travel, even if you use a rectangle test part orientated in each of the 3 axis.
@@CyberOne This gets back to assuming the tolerances are with in spec. I would not even bother doing a print. Write a custom gcode to move your x and y axis in a continuous clockwise square (10x) that is 90% of your bed space noting your exact starting location. Do this 10 times and measure. It will be off by a few mm. Now run it again counter clockwise and it will be back to the start point!
I think 20x20mm is too small for calibration, I would use at least 100x100mm.
I think your both wrong and should be calibrating based on a digital indicator like a machinist would. Wastes zero filament and you get an exact travel distance.
Daniel, sorry but I must disagree with your terminology: E-steps are strictly related to extruder steps ( short for Extruder steps per mm) and the rest axis are simply Steps per mm.
As for changing steps per mm for movement axis, there is more in it. One example: how you expect your printer to do 98.03 steps?
Calibrating movement of axis has meaning only if you use parts from an unknown source or printed pulleys. for example for Ender 3 Pro Creality uses cast pulley and genuine Gates GT2 belt that with every pulley teeth belt is "realigned" thus you should never change default steps per mm.
Thanks for the comment, you’re right about the terminology, I will probably update the video to correct this. On the floating point values. The digits are only relevant in movements that are multiples of that, otherwise they get rounded
I've decded to pin a comment to this video with the rectification. Figured it's very difficult to cut it out of the phrases, makes the video sound very weird at those occurences.
Every thing worked great till M500. Ender 3 v2, No eProm to save to. Now what??
You should look into upgrading your firmware first as this is a missing feature in Creality's default firmware.
@@Crosslink3D @Tom Gonzalez Can I put the M92 command in the start G-Code in CURA?
If you can't quite get it dialed in, a calibration trick is to halve the error correction amount. Example if the calculation says remove 1%, then remove 0.5%. Or here if it says change you X from 80.10 to 80.70, then go with 80.40 (M92 X80.40)
Daniel HELP You need to make a Video how to do the configuration of bed leveling on MARLIN 2.0 for ANET A8 .
Hey!
Your calculation goes wrong on esteps on that excel.
Instead of
=current_esteps*100/Extruded amount
you should have written
=current_esteps*100/(120-extruded_amount)
Have a nice day!