#z_tilt issues with unklicky

127 messages · Page 1 of 1 (latest)

oblique stream
#

Hello all, I am having an issue with z_tilt_adjust.

Context; Using Unklicky, the toolhead picks up the probe, moves to its starting point at top right corner of bed, probes twice with ~2mm between probes, drops the bed to around 5mm below the probe and moves to the next probing point, probes twice, drops 5mm and moves again, doing this two more times in all 4 corners. If the tolerance is not met, (usually not, frames jacked) it starts the process over again, but this time it probes twice and drops to 25mm before moving to the next probe point, probes twice, drops to 25mm and so forth, dropping to 25mm before every move on each set thereafter.

I am at a loss, research has been attempted, questions have been asked and here I am. I can supply the link to cfgs on github if need be.

This started when I changed to the unklicky and adjusted all the z things from 15 to 25 for clearances.

I have since went back and reverted the settings back one by one and I am not able to find what is causing this.

Any help would be appreciated. Thanks!

vestal stone
#

Is unklicky the same config as klicky with different hardware?

oblique stream
vestal stone
#

What’s weird it only happens after the third probe.

#

I’m not familiar with Z tilt…yet. Is it supposed to start on back left?

oblique stream
vestal stone
#

Oh. Ok. That makes sense. The video, not the concern. That doesn’t make sense.

#

Are you sure you adjusted everything back?

oblique stream
#

There are a few that I cannot adjust back or the probe hits the bed and gets knocked off. Thinking about it, I guess for testing sake I can put the normal klicky back on just to do the probing if the off sets will let me. 🤔

#

At least then I can definitively figure out what setting is causing the issue. 🤷🏻‍♂️

vestal stone
#

Can you disable Z tilt and see if that stops?

oblique stream
vestal stone
#

What about the console read out? What’s the call? I know it’s Gcode, but it may give a clue.

oblique stream
vestal stone
#

The log may not reveal much. I don’t have access to my computer to check the macros to see if there is something you missed. This is the klicky Z tilt?

oblique stream
# vestal stone The log may not reveal much. I don’t have access to my computer to check the mac...

Klipper has z_tilt built in, just give it the parameters and it does the rest. Here are my config files. These are my pre changes files, I have been referencing back to them. https://github.com/TheAZtech/Mercury_One.1

GitHub
GitHub - TheAZtech/Mercury_One.1: Configuration Files for ZeroG bui...

Configuration Files for ZeroG buildout. . Contribute to TheAZtech/Mercury_One.1 development by creating an account on GitHub.

#

Tho there is a klicky-z-tilt-adjust.cfg in the klicky files but I am not that great at coding and not sure if it is messing something up.

vestal stone
#

Im trying to find where it commands it to run the second probe.

#

I wonder if this is running two Z tilts one from klipper and one from klicky. So it’s dropping another 15mm.

oblique stream
vestal stone
#

No. The whole second go around. Since that’s when it dropping the bed.

oblique stream
#

ohh, its the tolerance in the the z_tilt.cfg

#

it the first one doesn't meet the tolerance it retries up to 10 times till it meets it.

vestal stone
#

That makes sense. Is the klicky Z tilt active?

oblique stream
vestal stone
#

Is that the one you’re using currently, or the klipper version? Sorry my mind is like a steel trap, I know.

oblique stream
#

No worries, lol. The one that is in the Klicky folder only references the klipper one, or adds the docking/undocking commands to it rather.

supple wolf
#

so its probing the font of the bed just fine (lowers a bit) and lowers the back of the bed what seems to be double

vestal stone
#

Ok. And the only thing you changed was klicky to unklicky?

supple wolf
#

did you resquare the frame in the past few days?

#

what's the bed mesh look like?

#

I feel so out of touch with klipper at the moment. I have not touched it in over a month now

vestal stone
#

Bambu is making you soft.

supple wolf
#

it really has

#

my concerns now are 1) do I have enough filament and 2) do I have something plated for when this print finishes in a few minutes

#

@oblique stream I think we need to rule out your frame and bed mesh. You may have already done this and I missed it or I could be entirely missing the mark

#

are the parts I sent intended to replace the parts you are currently using?

vestal stone
supple wolf
#

yeah, I feel like that is a common sentiment

#

my merc toolhead just arrived

vestal stone
#

Thats the universe telling you to get back at it

supple wolf
#

ive been waiting on all the haydra parts for ali

#

38 days

#

not everything is here yet

#

wow this is tiny

oblique stream
supple wolf
#

it does

#

hmm

oblique stream
oblique stream
vestal stone
oblique stream
vestal stone
#

Please hold

oblique stream
vestal stone
#

Yes. I wasn’t suggesting that more like hoping it was a value you changed but had not changed back.

#

Does it ever retry more than once?

oblique stream
oblique stream
supple wolf
#

so its the frame?

vestal stone
#

Ok. Anything over attempt 2 does it drop again?

oblique stream
oblique stream
vestal stone
#

So if it’s 25 on the first, 50 on the second, 75 on the third, etc?

oblique stream
vestal stone
#

Ah ok.

supple wolf
#

any luck?

oblique stream
#

Still at a loss with this issue. I cant find anything and I still don't know what is causing it.

spring vortex
#

if you try to get the deviation of your bed lower than what your system can reliably achieve, the range will start to increase again

#

interestingly enough, I am facing a similar issue on our trident where from time to time, it fails while doing z_tilt
but I'm pretty confident it's cause I never resquared the frame after building it and thus the system is not able to reliably achieve the tolerances I expect
Cause on my ender, I have no problem, getting the tolerance down to 0.0025mm with 5 retries
0.001 is also possible but for that I need 15 retries

#

(given the fact my unklicky probes both have a range of 0.001x and I am running a retry tolerance of 0.0025)

oblique stream
#

I get that, my frame is pretty messed up, I usually get it with in three retries, the issue tho is that it probes the first set, all 4 corners, droping to lets say, 25mm, and then any retry after that it drops the bed to 50mm after each corner.

spring vortex
#

(probe points)

oblique stream
#

4 one in each corner of the bed,

#

2 taps each

#

it only drops like a 1 or 2mm between the taps.

#

it drops to 50 before the move to the next corner.

spring vortex
#

that means the system can swing up(I dont know what the correct english word is)
which messes with the algorithm

#

that's math at play
Problem is, I dont know how well I can explain that in english but let me try

#

first of all:
if your frame is messed up, it's basically the same as if your bed was warped
Which means you need to run a high tolerance
now you measure each point and klipper tries to get it to the lowest possible deviation (tolerance does nothing to the adjustment code but is only a threshold for klipper to say: now that is good enough, it will try to get everything perfect on each try)
tolerance was off, so klipper retries
now since you loweder front left, back right came up more than klipper expected as klipper expects a perfectly flat bed and perfectly squared frame
now the deviation is even bigger cause klipper "overcompensated" due to your system not being in a way the algorithm expects
Does that make any sense?

#

I would always use 3 points and not 4, cause a plane is defined by 3 points, 4 points is overdefined which can result in errors

Think back to math class
if you had 4 equations with 3 variables, there is a pretty high possibility that no ssolution exists

#

(now everything I wrote is my personal understanding of it so take it with a grain of salt, I might be totally onto the wrong thing here)

oblique stream
#

correct, my tolerance is at 0.08 right now, which it meets, just takes some time to do it.

#

I have seen as low as 0.049 on a cold bed.

spring vortex
#

oh, another thing pointing towards my theory of the frame being out of square causing such issues:
When I try to z_tilt the trident while cold, good luck
hot it's mostly fine

oblique stream
#

So understanding all of that. I am trying to figure out why the first set of 4 points on the bed drops the bed to the set spacing, and all other moves from point to point after that are dropping the bed to a different spacing.

west nebula
#

ho, just seen this. you know, my printer does exactly the same, I haven’t looked like something that need fixing, but it’s annoying

#

if you notice also the values of the tilt change considerably after the first round of probing, could it be a relative coordinate problem, meaning the z tilt goes by relative coordinates after the first adjustment

#

it’s not much of an issue for me, as I left the tolerance at around 0.05, and unless I misplace the pei it goes by in 1 or 2 attempts. but I see it doing when it takes more

oblique stream
spring vortex
#

When I have time today I will dig into the code
Tbh I haven't looked into that
Maybe I can find something
If I find something would you be down to help me testing it @oblique stream @west nebula ?

oblique stream
#

Sure, let me know what you come up with.

spring vortex
#

Depends on what I can find
If you have a general understanding of coding we can also take a look at it together
4 eyes see more than 2

#

Another observation I just remembered:
When trying to get the 0.001(I do a more precise z_tilt when doing a bed mesh) I needed to add a feature to deactivate the "probe range increasing" safety check(I know the check has a reason but tbh, for a Hydra bed to crack, you need a lot of deviation, more than I could ever reach with 15 retries)
Otherwise it would trigger that
With that deactivated, it does some runs where the range decreases, then it increases again and then it settles in

oblique stream
west nebula
west nebula
# oblique stream <@261571630109556746> I confirmed what you are saying. I have this posted in a f...

it is annoying tbh, I was lazy to see this as a thing to fix, but it doesn’t look like a good behaviour. to confirm this, one should observe the z position of before and after the second round. if it changes of about the same amount as the height of the klicky probe, it could be that. the first round occurs after homing the nozzle on the z-probe (at least for me) that the first tilt at the 4 corners, and it lower the bed to some safe value+height of klicky

oblique stream
#

I am sure something, somewhere can be added to force a height from the probe plunger or nozzle even. The issue is I am not a coder in anyway shape or form and have the lease slightest clue on how to do any of that.

spring vortex
#

Btw
I looked at the code and the code seems to be pretty good
It does a coordinate descend to calculate the adjustments
Tbh I think those problems we are having are a result of a badly squared frame

oblique stream
#

This is what the calibrate_z looks like...
// Z-CALIBRATION: probe=20.560 - (switch=18.068 - nozzle=2.108 + switch_offset=4.485) --> new offset=0.115250 // HINT: z position_endstop=2.094 - offset=0.115250 --> possible z position_endstop=1.979

oblique stream
#

So I this was the issue. I have no idea how this ended up like this. I knew it did not look right but something changed it to this and thought it was supposed to be.

Screenshot_2023-12-17_at_2.28.10_PM.png
#

@west nebula Check your offset, if its this low, I am doing a writeup on how to fix it.

west nebula
oblique stream
#

In the save_config section on the printer.cfg?

west nebula
#

I hate patching the config with saved value.. so I don’t have any save config entries 😅

#

but I’ll have a look as soon as I can

#

you never know I pressed save with my sausage fingers

spring vortex
#

I have new information:
After looking at the x and y offsets again, I noticed that a mate of mine had 21.00 for y on his sb instead of the default 25.00
I changed mine to 21 as well and lo and behold, consistency improved quite a bit
then I took out the cad and noticed, that with unklicky, the x offset was -1.9
i added that as well and now I am consistently getting 0.001x tranges on z-tilt

oblique stream
west nebula
spring vortex
# west nebula wow

well to be honest, I always assumed klicky had the same offsets as the inductive one and unklicky has the the same as klicky
turns out: they do not apparently XD
Good thing for me: I designed my merc toolhead with the unklicky in mind so my x offset there should actually be 0

oblique stream
#

I found my issue was the z_offset. Some how, some way, it got set to 0.268 when it should have been 20.something. A gentelman on reddit gave me a list of steps to perform and it fixed the issue.

vestal stone
oblique stream
#

Ya, I have been working on a write up, I am just not very good at it.

vestal stone
oblique stream
vestal stone
#

What file type is this?

#

@oblique stream

oblique stream
#

just a text file

vestal stone
#

Ok. I’m on mobile. I will have to grab my laptop and take a look.