Calibrating your XYZ steps using a dial gauge for maximum accuracy
ฝัง
- เผยแพร่เมื่อ 22 มี.ค. 2021
- If the dimensions of your printed objects are out, this video will show how to use a dial gauge to measure and adjust your steps per mm for the X, Y and Z axes. It also covers the wrong way to tune this parameter so you don’t make the same mistake.
Thanks to the viewers and my patrons for requesting this guide and video.
XYX steps calibration tab on my website: teachingtechyt.github.io/cali...
Previous video launching my website: • 3D printer calibration...
Purchase a dial gauge on Amazon: amzn.to/3fbw02B
My printed dial gauge holder on Thingiverse: www.thingiverse.com/thing:480...
Buy quality and affordable filament from X3D. Buy 3, get 1 free and a free sample pack with every order: www.x3d.com.au
Get Quality Resins from 3D Printers Online. 5% off storewide for Teaching Tech subscribers [Code: tech5]
3dprintersonline.com.au/
Take a look around and if you like what you see, please subscribe.
Support me on Patreon: / teachingtech
Some responses to frequent comments here:
- In the example shown fixing Z axis binding on the Ender 3 V2, the leadscrew was removed, hence the gantry's ability to move up and down freely. The full process is here: th-cam.com/video/vW4DaWCd0Xs/w-d-xo.html
- A few comments on plastic shrinkage contributing to the final inaccuracy of the parts. Very good point and I will add this to the webpage.
- Some people has critiqued the method, which is fair. Ideally, we would be measuring 100mm or more of travel, but the dial gauge doesn't really allow that. There's a note about this on the webpage.
i dont mean to be off topic but does anybody know of a method to get back into an instagram account?
I was dumb forgot the account password. I love any tips you can give me.
@Bode Moses instablaster :)
@Duncan Beckham i really appreciate your reply. I got to the site through google and Im waiting for the hacking stuff now.
Seems to take a while so I will reply here later when my account password hopefully is recovered.
@Duncan Beckham It did the trick and I actually got access to my account again. Im so happy:D
Thanks so much you saved my account!
@Bode Moses No problem =)
Thanks a ton for giving us this info, this is sadly something that has been neglected and something I feel is critical and missing in the community so hats off for getting us up to speed with accuracy!
Can't wait to try this out. I have been running my Ender 5 with only minimal alignments and have had very few issues. Might be time to see what this thing can really do when tuned up. Love the channel for its content and presentation. It's fun to learn new stuff,Thank you.
Outstanding tutorial. How and WHY clearly explained.
Ohh ball bearing tips?! Sold!
Looked at them a while back but just used my calipers instead. Now I need one. Thanks!
Thank you so much for a proper insight to setting up a 3D Printer and the best thing that no one else has done and that is have a website to enter the values of the E-Steps and get a proper result without any confusion.
now my Stepper Motors are set up correctly and now performs super accurate on my Enders 3 V
Your site is very helpful. Thanks so much for making it.
Being a tool and die maker in the past, once a year we would have a tech come in and check our cnc mills for travel and or backlash. There are belts between, stepper motors and ball screws. This would give the tech the condition of the belts, balls screws etc. We needed to do this because we worked a lot within +-.0005''
Also, they are a closed system. They can remove the backlash. 3D printers are open systems. In this case, temperature and humidity take a huge factor on the belts. Like guitar strings. ;)
So like many mention, Your parts of the printer are a constant due to being solid and the belts can change length with temperature and humidity. Like any object, they are affected by temperature and all have a different thermal linear expansion.
So stepper motor counts/mm should stay the same. Either dial in the belt tension with the dial indicator or use the thermal expansion in the slicing software.
Also previously being a plastic injection mould maker, you have to take the shrink rate per type of plastic you are using. For example, setting Z steps for PLA for that perfect 20mm will not be the same fore other types of material since the shrink rates are not the same. Again, using the thermal expansion in the slicing software.
I experimented with this and got better results with default step values according to pulley diameters, etc. I honestly don’t think 1” dial gauges have great enough resolution to make this practical. This is a better concept if it can be scaled, particularly for harsh, changing environments and wear.
You are 100% correct. Esteps have a lot of variables meaning that no "one rule applies". XY and Z steps are based on math. They are correct. Changing them is a fast way to losing dimensional accuracy.
For those of us who provide technical support for printers, we're preparing macros to ask "Did you change the default steps other than Extruder?" essentially. This is gonna end up with long, 24 hour+ prints failing due to messed up steps.
Yes, this may be an advanced method, however the average person or even kid who got a 3D printer for their Christmas/birthday/whatever and has access to their dads toolbox/Amazon can and will do this. And it's gonna screw up a lot of things.
Damage control is underway.
@@Guffy1990 Yep. You definitely nailed it. Factory e-steps for XYZ don't need to be changed.
@@Guffy1990 i agree with you. But i wonder why it is not like that with the e-steps? The gears are pretty accurate i think
@@certified-forklifter there's a lot of variability with esteps. We know they should be 80, but because the gears aren't standard, the moving axis isn't one-dimensional, the pressure that is on the filament, how soft the filament is, the setup that a person is using... Yeah, as you can see, there's a lot of variables to take into consideration haha there's a reason tuning esteps is common, whereas this method hasn't really been talked about, because it's a moronic technique.
You likely don't need an exact estep setting, as evidenced by stock machines coming with their esteps set to 93 usually, to deliberately underextrude. A slightly liney print is better than a horrific blobby mess. Even then, flow can be adjusted in the slicer in order to compensate for the stock defaults.
To put it into context, the default esteps calculated mathematically should be 80. The best esteps are likely around the 100 mark for a non-geared Extruder. You can see how bad the underextrusion would be here I hope
@@Guffy1990 thanks
This just what I needed. I just printed a dial mount for the CR-30
I’ve been doing this for years so this isn’t anything new to 3D printing but was great to watch 👍
Although a good idea, going of the math for the pulleys and belts (default steps/mm) is probably better. I found if my parts are a bit to big or too small, using the expansion compensation in the slicer is a more reliable.
I need my parts to be as accurate as possible for work, this is a HUGE help. Thanks!
You can see my other more detailed post on why, but I promise you it's not a good use of your time for work. You'll get far better rewards for your time not cheaping out on your extruder + hotend, doing direct feed if possible, dialing in your print profile and pre-heating your system+ controlling your ambient temperature. If math derived X Y Z values aren't giving good results, there's something mechanically wrong in your system. Fix that instead.
The recipe is: Check your belts, set screws, mounting screws. Use math derived X Y Z steps. Set e-steps by extruding free (no hotend). Determine maximum material flow rate @ given temp based on this no-resistance e steps value. Focus your machine firmware settings & print profile on maintaining maximally consistent volumetric material flow rate. Pre-heat your system and control ancient temperature and humidity when printing. That's how you maximize part consistency.
And as a reminder the best fdm tolerances in the world right now are the greater of +/- .09mm or .0015mm per mm.
If you're serious about your work, don't waste your time.
I should've also probably mentioned checking axis skew and binding when you're checking the mechanics. Just to be thorough
I am on your website now. I'm up to a Slicer Flow Calibration cube its printing. then you upload a video about it. I just want to say Thank You Michael. The amount of times your videos and website have saved my Bacon. I just joined the Patreon its the least I can do.
I used this all the time!! THANKS
Use a 360 degree rotation of the stepper for movement calibration. That eliminates some periodic errors.
This is very helpful because the typical way is to print a xyz calibration cube but if your z steps are too short then your nozzle will not be raising enough making your print fail so you won't have the dimension to calibrate the correct steps/mm.
Also since I didn't have the dial gauge mount I just used legos
This is awesome! Thanks so much
Love your videos! Especially since you're Australian and when it comes to getting tools and supplies you point me in the right direction every time! With the calibration method, would the bed heating/cooling have an effect?
Good video, easy to understand explenations.
I think you should have mentioned material shrinkage. Especially because no two materials shrink by the same amount.
In the beginning this may not be as important especially when printing with pla, but on some projects where part dimensions are key. this needs to be considered
Keep on going!
Great point!
I needed dimensionally correct scaled parts in a resin version. But this is a neat calibration for the old Wanhao i3 FDM with upgrades. I'm becoming a Patreon in hopes you'll figure out a resin dimensional calibration how to for your website. Cheers - NJ
resin should not need any machine calibration if your slicer has the correct screen specs for your printer. you need to scale and/or size the parts correctly before slicing.
When indicating the Y axis, mount a target to the bed. Print it or clamp it lightly but rigidly. Also, mount the indicator base to a heavy steel block. These two things will reduce deflection errors.
Thanks,
John
Fantastic information, thank you!
I tried this method but wanted a longer travel than what was available thru pronter face.
I used Gcode to move 40mm in each axis. I found the actual travel varied up to 50% of the typed value.
Any idea why this would happen?
And here I thought I was doing something by printing the cube, measuring dimensions, then adjusting the steps based on over or undersized dimensions. I was able to get my cube within a few tenths of a mm in a couple prints, but I imagine I could saved the 6g of filament and the 20 extra minutes using the dial gauge. Seeing as I am building the MPCNC now I think I'm adding this to the cart.
I'm using Your site for every new roll of filament, great job and thank You!
I taped a needle to the hotend as dial and put a ruler on the printbed last time I calibrated e-steps.
XYZ are mathematical calculations based on pulleys etc, i never suggest to do XYZ steps unless your replacing parts with non stock ones that change those equations
Have you tried using a dry lubricant ( either spray or powder ) on the moving parts ? Especially in the Z axis .. wet lube and dust over time may gum up things . Especially if they sit for any extended amount of time .
Any way you can attach a bom with links to your ender 3 rear mount case? I can't seem to find the exact relay and voltage regulators you used.
is there any way with this method to calibrate a delta fdm printer like flsun q5? If not can you make a follow up video for core xy and deltas. i think would be intresting topic.
are you planning on doing a Video on the new BIGTREETECH E3 RRF V1.1 mainboard?
I have been on your site a lot lately and I am curious. Do you have your calibrations listed in the order in which they should be performed or do you have a specific order that should be followed for a new printer and testing new filaments??? I assume at the least, flow and temperature should be checked with every new roll of filament?
oh crap, this is a great idea!
Thank you
It's not, but if you find it fun, have at it. You're unlikely to make it much worse.
@@sneaky_tiki With such comment, I wonder if you even have or had a 3d printer at all loool
@@ralmslb I don't even know what a 3D printer is. I'm just another asshole on the internet, like everyone else
@@sneaky_tiki Thank you for confirming
I always had the problem that if I printed a calibrarion cube it was always 0.2mm too smal. Even when printing a 30mm instead of 20mm cube. Later I found out that this was from backlash in the servo I was using. Activated Backlash compensation and now everything is good to go again.
Out of curiosity, are you able to send these same commands through Klipper to tune it or is it still required to use something like Pronterface to connect to the printer to send the commands?
Such a good teacher. Many channels just talk and don't actually show any useful information. Not here.
Right, time to purchase a dial indicator
Hey would a print of something that rigidly mounts the dial to a section of aluminum extrusion work?
Could you have mounted your indicator 90Deg or horizontal to the bed and then move the indicator back and forth along z against something mounted to the frame?
I suggest ,for the Z - axis, it is easier to mount the indicator to the upper chassis of the printer and indicate the Z movement of the gantry
Measuring the Z displacement like that (up/down) will also include an error due to backlash in the measurements. I might suggest only driving the Z in one direction (probably upwards) and using the results of that for determining any changes to the steps/mm. Any backlash in the system should be handled separately.
I don't think you should have fractional steps per mm unless your belt is over stretched or really poor quality, because repeatability of the machine may not be the same all the way through the range of motion. What you can end up with then is a compounding error along the axis you change. IMHO steps per mm should be exact to the calculation of your pulleys and stepper motor or this error is amplified by the travel distance and not actually improving repeatable precision. On a typical 16 tooth pulley belt for example if you have 100step/mm and you decide to change this to 101, for a 100mm travel you will be 1mm too far, and for 50mm you will be 0.5mm too far.
Came here to say the same. Great, you can make perfect 20mm cubes. Not try 40mm and 60mm. Now you're off again.
You could use backlash compensation for a fixed value to compensate.
I love this but I no longer use Marlin. I think I recently saw you are getting into Klipper? Any chance you would do some videos on doing these calibrations in Klipper?
Can you use that Dial indicator with Magnet base to level the bed.. without having to mount the dial onto the hotend assembly?
With dial gauge is problem that it only apply to 1cm, and that error could not be linear (especialy Z with screw) unfortunately.
If you have a variable error depending on Z high it means your lead screw pitch is variable -> hardware problem, not related to settings.
@@net2CoCo However you will not know this unless you test across the entire range of motion so I think his point stands.
And in addition it is needed to take plastics shrinking into account for dimensionally accurate parts.
I have been trying to find this info forever! Thank you!
This is, in fact, highly useful! In part because it is explained without intense use of 3D Printer jargon. Real methods of measurement, Excellent!
If we've calibrated the flowrate before, can we use calipers and a testprint to check accuracy?
You are just awesome good sir!
This morning (in germany) I was thinking - maybe my printing problems are related to e steps on Z...
6 hours later boom - video from teaching tech about it
Thank you so much 💖
I really have to say a caliper is not a bad tool to use in this case. There are very good constraint points for the Y and Z axis, and the X axis requires some visual aid to make it perfectly parallel, which is not ideal but I was still able to calibrate.
Good video. Well done.
Being an Aerospace Machinist, I'd like to point out to you that it's a Dial Indicator and is properly known and called such. This is not a Test Dial Indicator, but a Plunge type Digital Dial Indicator.
That is great but if were going for accuracy then what about backlash?
Of the steeper+ belt/lead screw.
I know marlin have the M425 but unfortunately Klipper don't have this feature
Question how would you calibrate XYZ steps for a Delta Printer (using Klipper).
That would probably be a good idea
I think you should add encoder for the motor and change lead screw into ball screw. Even though error is still occuring in smaller number. It is a robust way than changing your step/mm. Using dial gauge is a smart way up to a point but does not eliminate the problem of dimension shifting over time due to temperature and disturbances working on the motor and printing different material. But if you use encoder. Dial gauge is used for setting the flatness of your bed
perhaps a cnc digital readout might be a better tool for this. you can usually pick them up on amazon for around 30 bucks for a single axis. It would give you a lot bigger range of motion than the 1 inch digital runout gage your using
Tell me, what about the values themselves? I've heard that only integers can be used as values. Is this true or can I put 92.29 if it fits? Thanks!
Does the slicer take into account the width of the extruded material? It seems from your tests it doesn’t do this which means the specified dimension of any shape will be incorrect by the width of the extrusion, bigger for outside dimensions and smaller for inside dimensions. This is generally not a problem for my prints until I make holes for screws or a lid that has to fit on a mating surface. I wind up with tolerances learned from experience on my printer or from printing test holes for screws with several dimensions above and below the theoretical size. This is an ongoing issue and maybe I need a chart showing tolerances pinned up above my pc.
hey nice...i just grabbed myself a dialguage..guess i have to find the USB cable Now! ;)
I want to ask if you disable stepper motors your z-axis is falling down the table, like we saw in the video?
The leadscrew was removed for that testing so I could test it moving freely.
there is a ball bearing on it's tip. but it doesn't rotate. it's just there to be a hardened perfectly rounded tip.
So there is this technique. And then there is the horizontal expansion technique. How would i know which setting is causing my failure? As a technician i can confidently say that there are usually 3 or more different incorrect ways to make adjustments or corrections to COMPENSATE for poor machinery performance however there is only 1 correct repair. So how would i know which is the real issue?
Thanks for all your amazing effort. What would be handy is a conversion to rotation distance for Klipper users.
I been watching your videos for a long time now and I have learned so much. Now I am stuck. I need advice, help,suggestions. I purchased a Biqu b1 a few months back and it prints great out of the box. I wanted to try idex printing so I found there idex upgrade kit on Amazon. I received the kit installed all components and upgraded to the FM bigtreetech suggested. Now comes the problem. I try copy,mirror and idex mode and it all just went Bazar, the original X axis (X0) and the added X axis (X1) move totally different than each other. When in copy or mirror. I try a cal cube and X0 prints at Y= 20mm Z=20mm X=20mm however the added X (X1) prints Y=20mm Z=20mm X=10mm
It seems that X1 prints at half of the x0 (I feel it maybe in the steps) the board is a skr2
Is there away to adjust the (X1) steps or anyway to text? If you or your viewers has any suggestions I would be greatly appreciated and thank for your time
Hi, I have a CR-10 Smart and like it very much. But is lacks calibration setting on the touchscreen display or a USB to change in terminal mode. Any suggestions?
one other question, I am was getting roughness on my first layer and I reduced the flow 83% and raised the temperature to 220C, PLA+ , but I am seeing some roughness on the second layer. Ideas?
Thanks
Larry
Much like Raspberry Pi videos have gotten me to pronounce it "razzbury" like they do in jolly ol' England, I'm quite sure TeachingTech will have me saying "zed" in short order.
Another way to measure X axis and Y axis movement is to securely tape a ballpoint pen or pencil to your hotend with point resting a sheet of clean paper taped to the bed. Use the machines controls to move the hotend a given amount and then use a pair of calipers or a metal ruler to measure the length of the line on the paper. Cheap, simple, accurate and easy....
So once I do this, if my prints are still inaccurate? What do I adjust to make them smaller or larger?
Would you be willing to add rotational distance calculations for those of us using klipper?
I just updated Klipper last night and decided to go the route of rotation_distance in my config file. Looking at the docs (and knowing a little about my printer) I have "standard" stepper motors, so microsteps of 16, full rotation count of 200, 80 steps per mm. Everything perfect world would mean rotation distance = 200 * 16 / 80. 80 (the steps per mm) is the number here you would be looking to update using the teaching tech calibration. Once you get requested movement vs actual movement, plug it all into the calculator on the teaching tech page. Then take the newly calculated number and use it as the denominator in our formula: 200 * 16 / . The result of that calculation is now the rotation distance value for that stepper in Klipper. Im not sure how precise it supports, but rounding to the nearest hundredth is probably fine.
I'm going to try to do this with a Laser distance measure. No contact, larger movement areas and ... Well, lasers!
this here, next level
How accurate is the device your speaking of and how much does it cost? If it is accurate enough and affordable then I would be interested in getting one because it would easily allow for a much larger measuring distance of the full travel of the axes.
Does anyone know about a model which you can measure and includes a couple of possible accuracy errors? Like "if this wall is off you have to do x" "if this dimension is smaller do y" "if this circle is off do z"?
What lube u used with T8 rods? Ive used silicon or vaseline but those are going pretty meshy after while and eatc movement displaces lube.. So now Ive got some PTFE spray, should I use that
Try replacing your lead screw nuts from brass to POM (Delrin). No need to lube these at all. I’ve had a set of these installed for about 5 months now and they’re still working great - with no lube! You can find them on Amazon. The ones I found came with 6 nuts (3 sets) and they were pretty cheap
@@Mali-jm4lm Already ordered those and wayting to install
Also as you were saying this would be better over longer distances why not mount the dial further away and have the subject of measurement meet the dial at the end of the 100mm travel?
If you think about what you just asked you would realize that makes no sense whatsoever. That would be like me standing what I think is 12 ft away from you and tell you to walk towards me 10 ft and hope that you touch my outreach hand and then I'm supposed to tell you somehow how far you traveled. Let's say my arm moved 4 in when you walked into it, I have no idea how far you went.
Hi I think I encountered a problem. I followed the steps with an ender 3 v2. The results: X steps changed from 80 to 84.12, Y changed from 80 to 83.32. Z changed from 400 to 426.7. Printed the 20mm cube and the dimensions were accurate to about 0.2 mm. However, now when homing the printer, the extruder assembly collides with the right side of the X axis and wants to move further, the bed collides with the belt tensioner and wants to move further. Each time I can hear the stepper motor trying to push forward. I think the increased steps threw off the bed limits. I'm afraid this will destroy the motors. Help please. Thank you!
Hey I have a problem doing this with my printer. every time I try to save to EEPROM it fails to do this. I recently bought the new skr e3 turbo you featured recently and put the version of marlin that came with it (with slight adjustments for my specific printer) on it. it's running just fine but saving to EEPROM via any way (octoprint or printerscreen) doesn't work and I get failed to save to EEPROM error. do you have any idea why this is happening?
There might be a thing about doing a clear eeprom from the menu once.
I know im late but I think there is a much better way of doing this. I have had a lot of success calibrating my X and Y steps by using a program like full controll gcode to write gcode that simply draws a one layer thick 150x150x5 cube. The gcode has the printer travel exactly 150mm, with each turn, so (while the part is still stuck to the bed) the inner diameter and outer diameter of the box for X and Y can be measured with a caliper, averaged, and then compared with 150mm to calibrate the new esteps. This removes problems due to overextrusion and errors in mounting the dial gauge.
digital indicators are cheap and easy, but if you get tired of swapping batteries, dial indicators can be better.
(most people will probably call them both dial gauges/indicators, but technically a dial is an analog/mechanical device. so i try to differentiate by calling them a dial indicator versus a digital indicator.)
an analog dial isn't as easy to read, because it basically requires a judgement call or estimation. but if you don't want to be hunting for a battery a 1am some night when you just want to get a measurement and move on, then i would encourage you to get a dial indicator instead (and learn how to use it effectively), and cut that potential issue off at the root.
same deal with calipers.
if you're organized enough to keep spare batteries in a place that you'll be able to find when you need them, then this advice isn't for you. but personally, i don't have enough room to keep everything that organized in my studio aptmt. someday, but not today. and regardless, i just like the peace of mind from knowing i don't ever have to worry about it.
So - I have 5 printers, 5 more being built. Most of my parts don't need accuracy, just quality. And I'm at the point, I can tune things by "ear" and by "eye" - hope you revisit the XY calibrations re: printing a 20mm cube, calibrate to that, but that is just a scaling factor, so now your 100mm cube of off. Seem like a week link in your "Methodology" as they would say.
what about the fact that printing inside out compared to outside in changes ur measurements
Nice video, but it does not work for my printer Anet ET4 pro (modified to direct extruder and z-axis stabilization) due to it‘s closed software. I did it this way, using Simplify3D:
1. I adjusted all parameters for the printer so that my prints are almost free of „bad“ effects that may have an effect on measuring the dimensions of a xyz test cube.
2. I designed (by TinkerCad) a 20mm test cube with x, y and z imprints on the respective sides to identify the axis
3. I measured the printed cube with a calliper
4. By calculation I identified the „scale“ of each axis in percent
5. In Simplify3D I added this scale values in the preferences. My values are z = 100%, y = 102.5% and x = 100.5%.
6. Printing a new 20mm xyz test cube showed that the accuracy is at 20mm, 20mm and 19.9mm. Good enough for me.
But using a dial gauge to check the evenness of the print bed is a good idea, I will do so in the next weeks.
Funny thing is, I have everything calibrated but my Y dimension is always bigger than the X, a 20mm cube measures 20.1 on Y and 19.9 on X, so it actually averages exactly 20, only it's bigger on one side and smaller on the other.
The only explanation I can think of is a combination of filament shrinkage, tool path, linear advance and etc, since the nozzle has to slow down for layer change in certain areas and not others, maybe it affects the shrinkage amount differently.
What grease should be used? Bicycle chain grease?
I like PTFE grease. Check out my video on maintenance for more info and a link.
I don't understand what improvement the Dial gauge will make if you still need to change the XYZ steps in the end based on printed cube ?
Generally I wouldn't recommend doing this. While it's true that tolerance stack up can affect the mathematically derived X Y Z steps, if you're only measuring in one location, you've only calibrated to that one location. And by location I really mean the interplay between your belt/lead screw, toothed pulley/nut, and eccentricities in your rotating components. Oh, and stepper angle tolerances. So unless you're able to measure across the entire length of travel, or painstakingly measure segments along the length and average, your far more likely to make it less accurate than more accurate. And that's not even getting into the realities of inherent variance with fdm extrusion, and the complicated thermodynamic and fluid effects at play. Did you do your tuning with your machine at steady state printing temperature? Well then you're no longer accurate. Not too mention belt and nut wear over time.
If you find this fun, by all means, have at it, you won't make it that much worse. But I doubt you'll ever actually make it better, and just generally will have wasted your time.
The part tolerance on my Stratasys machines at work is the greater of +/- .09mm or +/- .0015mm per mm. So on some of the best fdm machines in the world, your 20mm cube still has a tolerance of +/- .09mm. Just food for thought when you're evaluating how you spend your time
wouldn't it be better to attach the gauge directly to Ender's frame? I mean also for X and Y axis
yeah, the main advantage is that you can repeat measuring the axis again using the unchanged setup after adjusting the steps in firmware. This way you can check the result. I used Calipers and a small mount, to clamp them onto the build plate. This way I can measure X/Y steps without any need for manual holding. For Z steps my mount has holes, so I can screw them onto the 2040 frame of my printer.
www.thingiverse.com/thing:4599123
@@guidomersmann9744 thanks ;)
Can you do a tutorial using this method with klipper firmware?
Everything is the same, except you alter a value called rotation_distance for each axis, instead of X/Y/Z steps.
Hi Michael, did you consider printing one 20mm cube and 40mm cube with the same settings and then extract both in XYZ? This way extrusion parametrs will be eliminated. Might be usable when you don’t have depth gauge or when you are unable to get precise measurements with it.
That's a clever technique.
@@TeachingTech Woodworking style 😆
Nice guide! I have a CoreXY so how could I make sure I am adjusting the correct motor in a coupled system? I think in order to do this you have to assume both motors are identical with adequate belt tension. I guess another thing to test is the Y movement (should be 0) from just an X movement, and vise versa.
Place a sheet of metal on the base for your x and you can use the magnet. Clamp the sheet of metal to the bed for hands free operation.
You can't imagine how many online arguments I've gotten into with people about this very topic. I've had supposed engineers tell me that you should not do this and you should simply rely on the numbers from the belts and pulleys. My efforts to explain to them this process resulted in them in giving me the "I'm an engineer" line. Well, I'm an engineer too and I have done enough mechanical measurements to say with a high degree of certainty, "never rely on cheap crap from China to be dimensionally accurate." I will say that I have never used a dial indicator to do this though. You should use as much of the travel on each axis as possible.
While the idea of calibrating my machine this way kind of is attractive (and kind of not, because I'm lazy) how often would you need to repeat this process? I'm not (only) asking because of my laziness but because I'm cheap as well! Kidding aside, while 30 to 40 bucks for a good tool certainly is worth the money, it's only worth it if you actually need it more than once. I own a single printer and I will add maybe another one (with a direct drive) and that's it. If I was running a print farm, sure, but how often do you really need to calibrate a single printer that way?
Another problem with using that dial gauge is that a 3d printer is, virtual, a "no load" machine. Other than moving its own weight around. That spring in the dial gauge, however small as it is, is throwing things off.
I have dial gauge.Do you have dial guage holder slt for cr 10 smart ?. I findig every where.I dont know how is write 3d :(
I was wondering how to dial my ender 3 in better but guess I don't need to my 20 mm cube is 20.01mm on every side
9:41 woah woah woah woah wait wait... It's supposed to go down by its own weight?
Not with the lead screw in, connected to a stepper motor.
I'm crying in the corner with my delta
parts all come out scaled differently towards each tower :(
Which Delta do you have?
Easy fix for a Delta. Adjust arm Length.
Down steps towards the left, triple steps
Here we go again, every few years someone comes up with the idea of calibrating their 3D printers X, Y and Z and every few years people who have AT BEST a dial gauge and some digital calipers go changing the 80 steps per millimeter (formula Steps per mm = (Motor Steps per Revolution (200 for just about every 3D Printer) * Driver Non Interpolated Micro steps (16 for just about everything but DRV8825 which are 32) / (Belt Pitch (2mm for most people) * Pully Number of Teeth (20 for most people)) it’s 80. It’s always 80 because all you can do is change it to 81 or 79 which will affect your prints terribly this number is not fractional at the level that a lay person has the ability to measure, all they will end up doing is measuring it incorrectly or getting thermal expansion/contraction or measuring REALLY HARD with the calipers to force it to be some value they want it to be. Stop it, just stop it, stop trying to calibrate a device that prints things the size of wastepaper bins at most with errors on the order of thousands of a MM, the only real possible room for any wiggling is in the Z height due to compression of hot filament but over 20 centimeters you might get it right for filament A but be vastly off for filament B. If you regularly print large plastic multi thousand tone cargo container ships at home then this might come in handy but for everything else this is an extra confusion to people new to the Maker community who are already confused enough!
I was a machinist for 7 years. This is how I calibrate my printers too.