Calibrating a Core XY

[Hayden Pulley]

Does anybody have any tips or a good guide for calibrating a core xy machine? I'm trying to improve my dimensional accuracy. I understand that the two motors need to use the same steps/mm, so I presume that belt tension is important so the x and y movements are the same. How do you know which belt to tighten or loosen? If both axis are off, do you average the error for the steps/mm? Thanks in advance!

[David Mason]

So first up I hate the fact people really recommend adjust xyz and e steps per mm for any design these days. These should all be mathematically calculated based on your components and then left u touched. The differences you will get from the variations in those parts will be just washed out by those from filament variations etc. Obviously check this movement is correct using a ruler but do NOT get the information from measuring a printed part..

Last 3 printers I've built and I've never once had to modify these from the calculated values. 

Now onto actual calibration. 

Accelerations - too high and you will get overshooting leading to oversized bits. Key identifiers are ringing in the parts and not very clean edges. Belt tension is quite important here too high and you will get more ringing, too low and there will be play in the system (essentially backlash) . Easiest way to check they are equal (assuming your kinematic system is incredibly rigid) is to see if both sides of the x beam touch whatever the physical end of the y axis at the same time. I'e the x axis is perfectly square to the y. There are a bunch of ratrig and voron assembly videos covering that. 

Extrusion multiplier - check your filament dia most brands with tend to err on one side of the tolerance band specified. If all your filament measurements are 1.78 then turn your extrusion multiplier down a tad. If it's less than 0. 8 or more than 1.2 then either your filament is really bad or something else is wrong. Check your optimum filament temperature, you extruder gear tension and watch out for lost steps if you are pushing speeds! 

Shrinkage  - this is a massive one! Some filaments are more affected than others. There is no easy fix to this one. The best way is to incorporate for the shrinkage jn your design. Oversize holes a bit to account for it etc. 

Elephants foot - squidge the first layer too much and it will stick out affecting the part. If you need that much squidge to get adhesion either use elephant foot compensation in the slicer or make sure you have cleaned the bed etc. 

[LukeH]

The belt tension in a CoreXY is fairly forgiving, and doesn’t need to be exactly the same on both sides, as long as the gantry is square when you put them on. The way I do it is to push the Y gantry all the way to the front or back of the machine, clamp the Y gantry to the frame of the printer, then put the belts on. If you just let the Y gantry float while the belts are going on, it may crab slightly, which is going to result in your prints also being a little out of square, which you would have to correct either mechanically or in firmware (there is a correction factor to square up your prints, at least in Marlin, and presumably in the others). 

At full tension the belts should be just tight enough to not slip on the pulleys, and depending on the size of your printer might have an inch or more of play in them. Tighter isn’t better, especially for the little 6mm wide belts. For print speeds up to about 150mm/s, if you pluck it and it sounds like the low string on a bass guitar, it is probably too tight. It is not a car fan belt.

[LukeH]

The firmware fix for crabbing on a CoreXY is to use Bed Skew Compensation. This does involve printing and measuring how out of skew it is, since you can’t figure it out mathematically (because it shouldn’t be skewed at all). You can also correct in the other directions (not just XY, but XZ, and ZY), to compensate for a wonky frame.

[bron...@gmail.com]

this video is for a Rail core style

but it might give you some info

https://www.youtube.com/watch?v=9-1PNzsmrZM

just restung my core xy this week

[Hayden Pulley]

Thanks! some good info. I'm getting some layer shifting, and I'm pretty sure its from the acceleration. I'm using Reprap firmware, but I'm honestly confused about the codes. I've read about them on duets webpage but I feel like they could be explained better, especially M566 and M204. Currently I have

M566 X100.00 Y100.00 Z6.00 E110.00 P1                    ; set maximum instantaneous speed changes (mm/min)
M201 X1500.00 Y1500.00 Z100.00 E3000.00               ; set accelerations (mm/s^2)
M204 P700 T2000                                                              ; (mm/s²) 

Are these reasonable values?

And does anybody have a good explanation for M566 and M204?

Thanks again!

[bron...@gmail.com]

M566 X100.00 Y100.00 Z6.00 E110.00 P1                    ; set maximum instantaneous speed changes (mm/min)    this is your Jerk setting in Marlin its how fast your printer can stop and change direction basicly

your setting are pretty conservative and should be plenty safe

M204 P700 T2000                                                              ; (mm/s²)   these are the max acceleration setting for while your printing and while your traveling

so if your extruding plastic it will not accelerate as fast as it can it will limit it to 700 mm/s^2  but if your not extruding it will accelerate at 2000 mm/s^2

these are total acceleration between all axis combined so if it is moving x and y at the same time it could move them at 1000 each

or 1500 for X and 500 for Y depending on what the movement requires

the M204 line is not absolutely  required I don't have on as my config is clearly older and it doesn't casue any errors

I may add one now that I know about it.

M566 X1000.00 Y1000.00 Z60.00 E1000.00         ; set maximum instantaneous speed changes (mm/min)
M203 X18000.00 Y18000.00 Z400.00 E3000.00      ; set maximum speeds (mm/min)
M201 X2000.00 Y2000.00 Z16.00 E8000.00         ; set accelerations (mm/s^2)
M906 X1200 Y1200 Z1000 E1000 I30               ; set motor currents (mA) and motor idle factor in per cent
M84 S30

these are Mine but please don't take them as ideal they are higher

and I have worked up to these by testing and incrementing them up

so I am sure some people are going to say my setting are way to high

but they seem to work for me

there are some tuning Macros you can use on the Duet forums increment acceleration while a print is running so you can real time changes

https://forum.duet3d.com/topic/6181/tuning-macros-menus-accel-jerk-retraction-pressure-advance

they are not well documented I have just started testing them but they could be helpful

[bron...@gmail.com]

your layer shifting could be a current setting you may not be sending enough power to the motors

On Saturday, March 19, 2022 at 11:43:04 AM UTC-7 hayden...@gmail.com wrote:

[Hayden Pulley]

I just upped the amps and that fixed it! I followed rat rigs recommendations for the set up, and it seems like it was half baked. It was set at 40% of my max current. I didn't think that would be the issue, but it makes sense. I'll take a look at those macros. Thanks!

[david mason]

The default is quite low but it surprises me you got skipped steps? I've configured 2 vcore3 one with klipper and one with RRF/duet and haven't had an issue with the defaults.

I would definitely consider checking your axis for any binding or mechanical issues as well. 

Get Outlook for Android

---

From: 3d-printing-...@googlegroups.com <3d-printing-...@googlegroups.com> on behalf of Hayden Pulley <hayden...@gmail.com>
Sent: Sunday, March 20, 2022 10:15:35 PM
To: 3D Printing Tips and Tricks <3d-printing-...@googlegroups.com>
Subject: Re: Calibrating a Core XY

 

--
You received this message because you are subscribed to a topic in the Google Groups "3D Printing Tips and Tricks" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/3d-printing-tips--tricks/4mLQyrUofzM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to 3d-printing-tips--...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/0ed7c565-d91b-4981-a298-db46121e76d0n%40googlegroups.com.

[Luke Hartfiel]

It used to be an issue with the older stepper drivers where they would come with the manual potentiometers set at whatever random setting they left the factory in. These days though, with most boards allowing stepper driver currents to be set in firmware, if you are using a kit or established design it is less likely to be a problem, but variations in stepper motors between models and brands can sometimes require tuning.

For scratch builds though, there is still a bit of trial and error involved, although you can make some educated guesses.

On 21 Mar 2022, at 9:22 am, david mason <slzer_...@hotmail.co.uk> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/DB8PR09MB4122CA530FAE4A8630BEDB90B9159%40DB8PR09MB4122.eurprd09.prod.outlook.com.

[bron...@gmail.com]

they likely set them as low as possible in the recommendations to avoid burning up peoples stuff

Mine are pretty low around 50-55% of the drivers capacity and as you can see I am running pretty fast

I would look into excess friction or binding a little bit of light oil goes a long way on linear rails also if they

brand new they may take a bit of time to "break in"  I ran my machine at crazy speeds for a while with some hand written g code

before I even had it fully assembled just to get things moving a bunch and find out where the problems were if there was going to be any binding I wanted to find it before I Put a full hotend and stuff on in in case I had to take it all apart again

[Hayden Pulley]

I've checked for binding several times, not an issue for x and y. Z is having issues but only halfway down. Above and below the halfway point it's fine. I turned down the amps and tried a couple different prints and didn't get any layer shifting. I decided to try the print I was getting layer shifting on again. I was getting some curling on parts of the print. The nozzle caught on some curling pretty good and caused the layer shifting. Turned down the layer fan and that fixed the curling. I'm not used to having a layer fan work that well. I'm not exactly sure what to do for my Z axis though. Individually each motor assembly seems fine and it all moves smoothly. Its odd too because it's only halfway down that I get the issue. The way its built allows some variation in the plate for thermal expansion, and as part of that It allows some minor error in the placement of the back stepper motor. I would think its one of the front motors or a bent leadscrew but I don't see how either of those would make it bind only halfway down. I have kept everything well lubricated with light oil. I've been pretty busy so I have yet to look at it in detail.

[Rob Hall]

If the screws 

And the rails don't align well 

Your going to have binding

On Tue., Mar. 22, 2022, 7:56 p.m. Hayden Pulley, <hayden...@gmail.com> wrote:

I've checked for binding several times, not an issue for x and y. Z is having issues but only halfway down. Above and below the halfway point it's fine. I turned down the amps and tried a couple different prints and didn't get any layer shifting. I decided to try the print I was getting layer shifting on again. I was getting some curling on parts of the print. The nozzle caught on some curling pretty good and caused the layer shifting. Turned down the layer fan and that fixed the curling. I'm not used to having a layer fan work that well. I'm not exactly sure what to do for my Z axis though. Individually each motor assembly seems fine and it all moves smoothly. Its odd too because it's only halfway down that I get the issue. The way its built allows some variation in the plate for thermal expansion, and as part of that It allows some minor error in the placement of the back stepper motor. I would think its one of the front motors or a bent leadscrew but I don't see how either of those would make it bind only halfway down. I have kept everything well lubricated with light oil. I've been pretty busy so I have yet to look at it in detail.

--

You received this message because you are subscribed to a topic in the Google Groups "3D Printing Tips and Tricks" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/3d-printing-tips--tricks/4mLQyrUofzM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to 3d-printing-tips--...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/980fa33e-ecc4-448d-a2b9-6d1a12ef4e2dn%40googlegroups.com.

[Luke Hartfiel]

Same if the printer frame and/or the frame for the bed aren’t square. The bed will rack and bind the lead screws at about half way (where the lead screw deflection is at its maximum).

On 23 Mar 2022, at 3:12 pm, Rob Hall <bron...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/3d-printing-tips--tricks/CABWxTj-yzxxobT9uM-wSD-2Ay5LwVLTRu0-7py9G2ewmpZDa6w%40mail.gmail.com.

[Rob Hall]

Get some 123 blocks and a couple of clamps 

[Hayden Pulley]

Finally had a chance to look at it. I have extensively checked for squareness and leadscrew alignment. Everything seems pretty square, as all measurements were within .5 a degree. I added more lubrication and found a drastic reduction. I am new to using linear rails so I probably underestimated how much lubrication I need to use. Instead of the whole printer shaking only a couple leadscrews had vibrations. It was only vibrating on the way down, and not up (this was before and after I lubricated it some more). This suggested to me that something was loose, so I tightened the bolts attaching the anti back lash nut to the bed arms (not sure if that's the right term). With just half a turn of tightening it stopped vibrating! I had kept it loose to help align it but I had failed to retighten it completely.

[bron...@gmail.com]

hope its better now glad you found a few things hopefully that gets it sorted out