Voron v0.2 shaper issues

[frosty pagoda]

Hi, finishing up my Voron V0.2 formbot kit

Noticed some horrible ringing / ghosting, and generally bad input shaper results, especially on my X axis from what I can tell, attaching pictures that were requested in a separate message in here

Relevant mods:
West3D hybrid ceramic bearings
Fysetc CNC kit
Honeybadger CNC x carriage (MGN7h)
Hiwin MGN7H X rail Z1 preload
Rapid burner with Dragon UHF + Orbiter V2
Picobilical
LDO Leviathan
LDO 35STH52-2004

Printer famile: Voron V0(.2r1)

Which ADXL was used?
ADXL345 on Picobilical toolhead board

Measuring tension by Easy Tension phone app

Voron cube printed with 600mm/s @ 15000 mm2 accel (maxed out at 131mm/s on the cube)
0.5mm nozzle

Input shaper and belt comparison graphs, printer was on a wooden floor while measuring, with all panels attached properly apart from the back panel, where the remaining nuts were just fastened with screws for now

Machine pictures

[zinc gust]

Your acceleration is to high for input shaper to compensate

[frosty pagoda]

Gotcha, any advice on getting the picobilical sleeved up? Once that’s done I can just zip tie them together

[frosty pagoda]

Not right in which way? The way they run to the toolhead, different tooth count, or tension, or is there no way to tell😅

[frosty pagoda]

I’m sorry, can’t read this😅

Okay so the toolhead is causing some vibrations I see, I’m gonna go make sure everything is tight and properly mounted, thanks

[frosty pagoda]

Went to go and double check the toolhead, even with all the screws tightened up, there’s a lot of wobble on it, it might need some kind of brace somehow

For toolhead sleeving, I found some stuff called MDPC-X, any clue if that could hold up to the movement? If not, any other suggestions?

I apologize already for all the questions I’ll probably be asking, I wanna understand this

Is my Y accel limited by how my Y input shaper looks? Like, do the vibrations and the extra spike because of the toolhead limit me, I’d imagine a V0 should handle those accels normally

[frosty pagoda]

I’m trying to both keep good graphs, and to get some good speed going

[hasty siren]

dragonburner for me was a bit wobbly with top inserted heatset insert on bambu nozzle mount (force and heat kept backing the inserts out slowly as I cranked down on the orbiter mounting screws)

what helped a lot for me was using rigid standoffs behind the orbiter to the toolhead board and using a toolhead mounting board that attached ridgidly to the back of the carriage on both sides. Without that, the entire toolhead is basically dependent on the 2 front screws and slightly on the top back screw into the toolhead. Since they're pretty close together vertically and mounting point wise, it's like being attached to the front face of the carriage with 2 screws, and carriage can flex torsionally pretty easily when umbilical pulls on it. Back mount that also connects to the orbiter is like adding the 2 rear legs of a chair and adds rigidity

[frosty pagoda]

Ahh thank you! I was looking into on of those, since I have the metal carriage, you basically mean one like this, right? https://injured.lol/mn29xmc0.png

Ofcourse not that exact one, but one that fits the metal carriage instead of the back of the rapid burner, and the picobilical

[hasty siren]

deffo's look at reths recs too probably the toolhead wobble is a part of what he's saying, and check xy/motor/ front idler spacing, washers that are not spec caused a lot of noise for me too

[frosty pagoda]

Ah gotcha, this is all from a cheap formbot kit, but replaced a lot of parts

[hasty siren]

back of the carriage itself should have screws no? so back of the orbiter ->2 brass standoffs -> pcb holder plate -> back of x carriage

[frosty pagoda]

Yeah that's what I mean sorry, extremely tired so I didn't wanna bother to take a proper picture 🤣 let me load up cad

[hasty siren]

I think of it like a chair lol, the 2 toolhead holes are the front legs (on the R of your pic) and the orbiter + standoffs are the seat area, pcb mount and 2 x carriage screws are the back legs

if you only bolt down the 2 front legs, you can still tilt the chair forwards and back from the flex of everything, if you bolt down 4 legs and add rear legs, its less flexy

[frosty pagoda]

https://injured.lol/jcm9py1s.png
Basically just what's in the stock cad, then slap the picobilical pcb on that

[hasty siren]

yeah

[frosty pagoda]

Yeah I see now

Thank you, that is exactly what I was looking at, but forgot to follow through

Is there a specific one you used? or is that more extruder-specific

[hasty siren]

still in the process of opening cad do to do one myself, db repo has the universal one to mount orbiter + umb/pico/ebb36 board on the back

[frosty pagoda]

https://www.printables.com/model/479584-voron-v0-picobillical-dragon-burner-mount/comments

[hasty siren]

Imo it places the toolhead too far up for me so need to drop it down closer to the rail carriage

[frosty pagoda]

Ahh gotcha

[hasty siren]

was going to try and modify driftrotors one to have the orbiter mounting holes aligned with DB but haven't gotten around to it yet

that's the most rigid one I've seen/used, just haven't adapted it to DB yet

[frosty pagoda]

Ah yea I just realized that one's for the DB, not the RB

[hasty siren]

lmk if you do end up doing that and making one, would love to not have to cad it myself

[frosty pagoda]

I need some coffee, apparently,

So that should help a bunch with toolhead vibrations, getting that more rigid

Then there's just the belts and the one slide i couldn't really decipher (Sorry Reth 😅 )

[hasty siren]

It did with mine, a lot of the toolhead wobble/forwards "nodding" came from flex of x carriage/toolhead. Speficialy the orbiter on top of db, it was a bit flexy when hot and I could pull/flex it forwards and back a bit by hand

[frosty pagoda]

Yeah agree, I also noticed that with mine

[hasty siren]

not sure how cnc x changes that but might help clean it up slighly to identify other issues better

also, not sure what reth recommends (listen to him) but I've had sucess with scv 5 for IS graphs and upped the accel/hz till I hit low 1e5 range on the top left, something like 100-125hz

[frosty pagoda]

I'll definitely get the pain umbilical and "load distributor" made and printed, for the rest, I'll keep this in mind and see what others have to say aswell

[frosty pagoda]

Testing a mount right now for the Rapidburner and orbiter v2

Then I'll re-run IS

[frosty pagoda]

TLDR

Friend made a design for the rapid burner and orbiter v2, which also screws into the carriage,

Printed the bottom brace, rerunning IS, then gonna derack the gantry, because I definitely screwed up belts

So it seems that my toolhead vibrations got a tiny bit better on Y, and X is a whole lot cleaner

-edit, it seems that when comparing todays result, after installing the umbilical strain relief, and the rear brace, that they line up better? But the max accel went down, and the vibrations on smooth_ei went from 0.0%, to 0.2%

I'll try deracking the gantry, as far as I've understood, it's just

remove belts

make sure it sits fine

re-route belts

Make sure theyre the EXACT same tooth count

Tighten them a tiny bit, switching between A/B steppers

[zinc gust]

https://youtu.be/cOn6u9kXvy0?si=W5O6SQd2fOsyvW2a

[frosty pagoda]

Thanks, any input on the old vs new shapers? (old X attached)

Seems it introduced more vibration with smooth_ei, though i also have no clue what i'm looking at, nor what to look for

Wow that belt comparison is much better, definitely just needs a good deracking

[zinc gust]

The 125hz is something with toolhead, could be a loose screw

Or needs filament in extruder

[frosty pagoda]

Filament is in extruder,

I'm gonna print a newer version of the brace and strain relief combo, then gonna take off the toolhead to try and derack the gantry, tightening all screws in the process, thanks so much for the input

[hasty siren]

That’s significantly better than before

The scale on the left is 3.5 & 1 for 1e4

Before it was much higher and less clean

[frosty pagoda]

Yeah I'm printing an improved version of the strain relief mount

Or atleast, hopefully improved

[hasty siren]

Try doing accel per per hz up to 100-125 or so, set scv to 5, run it and see if you hit low 1e5 scale

[frosty pagoda]

Will do, I'm gonna try de-racking first, after this mount is printed

[hasty siren]

Try it without detaching

Might have ok results

Deracking

[frosty pagoda]

Will do, I'll derack after then, i think this printer desperately needs it looking at the gantry sometimes

[hasty siren]

Monolith discord recommends upping accel per hz for stiffer printers till it hits 1e5 to have more usable is results

I’ve found that to be useful as well

[frosty pagoda]

Gotcha, will try 5 SCV 75, 100, then 125

[hasty siren]

I’d jump straight to 125

That’s my guess

You want scale on the left to be 1 ish or sub 5

And top left to say 1e5

So 5scv, accel per hz 125

[frosty pagoda]

Alright, gonna do a 75 once, just to see the comparison between current strain mount, and "improved" version

Thanks for all the help, really appreciate it, never messed with anything like this stuff before

[hasty siren]

Make sure the mount is flush with carriage before putting in those bolts, I had to file down the standoffs or it’d try to push it forward

And also toolhead tightening order for me is front screws, standoff to backplate, backplate to carriage, back middle carriage screw, retighten front screws

But try is first

To see the diff

[frosty pagoda]

Ahhh I see, they might not be fully flush then

They're alu standoffs though 😅 That'll be fun to file down

[hasty siren]

And then fix toolhead after, will sleep so can look at graphs

[hasty siren]

Can just test with printed spacers down the line to see if there’s benefit. Or push toolhead board back and have the carriage mounts go inwards at the bolts/spacer there too

Just don’t go below head of a bhsc screw or it rubs on printed motor mounts, idk about cnc

[frosty pagoda]

Wdym by below the head of a bhsc ?

[hasty siren]

The 2 screws in the rear are m3 bhsc screws, clearance between that and rear motor mounts in cad is basically just the edge of the screw head on the bottom of it

So if printed parts extend below that it’ll rub the motor mounts

Did ya get graphs yet?

[frosty pagoda]

Nope, was printing the improved bracket

[frosty pagoda]

Ahhh, I’ll double check clearance

[hasty siren]

Ahh

If ya wanted to run is, I can have a Quick Look, otherwise I’m going to head to bed and will have a look tomorrow

You should see stronger more isolated single peak, higher accel and good vibration recs, gets rid of noise for stiffer setups afaik

[frosty pagoda]

Ahh, I'd say just head to bed for now, I won't be filing any standoffs tonight, but will definitely have some graphs you can look at tomorrow if you're feeling it

[hasty siren]

Might be worth running it current setup before changing a whole lot

Might give you clean enough shaker recs

[frosty pagoda]

Definitely, gonna try with just the new strain relief, then try 75 to compare to old ones, then 125

[hasty siren]

Then you can change and reference difference

[frosty pagoda]

Should have that once once im certain that this print wont warp, beds still at 77c

I should definitely also make a filter for this thing

[hasty siren]

Also be sure to validate with reth cube settings

Should see a difference (removes umbilical noise from is results) and gives you cleaner print

Idk if it’s overcompensating or just too much noise to generate clean IS but too low frequency on my v0 would always have ringing/vfa’s. not so much with low 1e5 graphs

Gl!

[frosty pagoda]

Will need, thanks!

[hasty siren]

Already massive improvement, x is ie4 4.8ish -> 3.5 and cleaner single peak so more rigid and less umbilical effect

Y as well but toolhead wobble is more apparent but actual value dropped

[frosty pagoda]

Yeah it's great, got worried when the first belt comparison suggested poor mechanical health... like damn, I spent so long on this frame and these rails...

[hasty siren]

Bit over 0.2 to under 0.2, just more apparent due to lower scale on left

At 125 x axis point

[frosty pagoda]

Yeah, I really appreciate the help i got from both you and Reth, helped tremendously

[hasty siren]

Went through all this recently with my v0 with a ton of is testing and setups, should be a good starting point and low 1e4 indicates high rigidity so once clean it should run nice

[frosty pagoda]

with the improved version, the X graph is slightly cleaner, and Y looks basically identical apart from the estimated damping ratio being a bit higher, gonna try 125

@hasty siren For whenever you're up, I'm also heading off but these are my graphs @ 125hz/accel, 5 SCV.

Y Didn't seem to enjoy it, X did

[hasty siren]

Y needs to go higher

1e5 up it in increments of 25 or jump straight to 150

[frosty pagoda]

I'll give 150 a try tomorrow, printer was sounding scary at parts of 125 🤣

[hasty siren]

Go play around with it, maybe detach and sort toolhead whatnots

Definitely look for loose things on y, 125hz toolhead issues

[frosty pagoda]

and the Y couldn't be bad belts?

Will do

[hasty siren]

Most likely binding with lower peak

And toolhead wobble somewhere with higher noise

[frosty pagoda]

Definitely toolhead wobble, It's still not that secure

[hasty siren]

Could be exrruder or if you’re certain everything right none be then up accel hz

[frosty pagoda]

Extruder is definitely tightened

[hasty siren]

It’ll be loud and shake it, rigidity needs more shaking

[hasty siren]

Orbiter gears themselves or shaft

[frosty pagoda]

Ah, i mean it's tightened to the main body of my toolhead

[hasty siren]

G2e and orb have toolhead noise issues can search armchair or monolith or ask around

[frosty pagoda]

Any better extruder?

[hasty siren]

That peak might be a bit high though, might be toolhead nodding

[frosty pagoda]

I'm down to replace it, this thing has been through some things I wouldn't wish upon anyone

[hasty siren]

Doesn’t affect prints with high enough accel hz

To minimize noise like on x

1e5 goal

[frosty pagoda]

Just wondering, is something like a cnc sherpa mini better for it, in terms of noise?

[hasty siren]

But yeah you have nodding still

[frosty pagoda]

Definitely, not sure how to minimize it

Could it be my lightweight gantry?

[hasty siren]

Doesn’t translate to printe in my experience

Maybe

But spacer things and proper spacing for toolhead board thing

So it doesn’t push extruder forward

[frosty pagoda]

Definitely, I'll get some pictures of it tomorrow and see what needs filing

[hasty siren]

Had similar 125hz when extruder was pushed forward from back board

Measure gap and print some standoffs to test

IIRC it’s 17.8mm or something

[frosty pagoda]

Gonna write this down so I don't forget it LOL

[hasty siren]

75-125 might be umbilical as well

Zip tie and zip tie to toolhead thing might help

Reth probably knows more about that

Gl

Graphs look promising though, probs just need to shake it more

18k Mzv x with 0.1%x, need to validate with print when done

[zinc gust]

Should Only need 75, 100 max. Monolith does okay at 125

[frosty pagoda]

Will try 100 tomorrow, and also look at printing spacers for better fitment

[frosty pagoda]

Gotcha, I have plans for tidying up the umbilical,

About tool head binding, I made sure everything was tightened down, the only thing I can imagine is the new brace + strain relief piece pushing on the extruder like Cabbage suggested

[zinc gust]

Binding is separate from tool head. Something is not moving freely. Check belt path. Make sure there is no rubbing

[frosty pagoda]

Gotcha, thank you!

[hasty siren]

yeah, some of the metal monolith needs a bit higher, around 150ish. Nicks is cnc gantry so might need higher, but needs test prints to verify

But yeah this is probably binding or umbilical shaking the toolhead board

[frosty pagoda]

Going to check belts in a bit, I could imagine some cables from the rapidburner are hitting them, that thing has sub par cable routing imo

[hasty siren]

y axis at 80ish hz is umbilical I think

the green line for y axiz movement on accelerometer is toolhead nodding front and rear

blue z from 80-125ish I'm guessing is something binding or loose, 25hz as well

besides what reth said about checking for binding, check xy and motor mount bearing gap, and whatnots

also try zip tying umbilical to brace in the back

[hasty siren]

but this y vs the new y

makes me think a lot of the 100noise might be from extruder pushed forward

low end still shows maybe binding on both

[frosty pagoda]

Definitely will measure, and print some standoffs, might do that first just to clean it up a bit, then look into all the daunting tasks😅

One issue with this orbiter, is that it’s been through some stuff on the left side, namely being filed basically all the way down through some of the metal

[hasty siren]

could just shake it at noisy frequency and touch things till the noise stops to find whats causing it

[frosty pagoda]

I remember 87hz being especially bad

[hasty siren]

gotta find out whats shaking it or just go through the whole motion system

put it in the center like in IS and shake/touch things. Might be panels even but probably not

[frosty pagoda]

IS immediately puts it in that position

Not sure if I have to change a config somewhere, but after homing, I do the axis shaper or whatever it’s called, and it puts itself up there

[hasty siren]

if you move it to 60 60 before is

does that home before running or just run in the center?

[frosty pagoda]

It homes before IS, sits at 60,60, then to start it, it goes to that top right corner

[hasty siren]

check your resonance tester in config

[frosty pagoda]

Ah, the probe points are at 100,100, 20

Seems I didn't change the default, damn

I'm gonna guess that all the shapers i've provided haven't been all that useful then

[hasty siren]

shouldn't be that different, I like to set it center just because it makes belts even path wise

idk if that makes a difference

[frosty pagoda]

Yeah for sure, I'm re-running them right now, and noting loud frequencies

75hz/accel
6scv
toolhead centered this time

[hasty siren]

up accel per hz

scv5

[frosty pagoda]

100hz, 5scv

[hasty siren]

this is relevant

I had similar noise and high peak 70-100ish from umbilical wires buzzing against each other and not ziptied down above the board

cnc gantry has less tolerance for error than plastic gantry, no flex so you have to make sure all your tolerances and rail mounting is on point

check bearing gaps against manual for undersized washers, make sure rails are completely parallel and free running front to back both slowly moving and fast moving

also saw your xy joints are shcs on 2 of the corners, make sure it doesn't bottom out in the carriages or isnt too short if you dont have fhcs

last bit, if you're 100% sure that everything is tight on toolhead for 125hz, try flexing it forward and back and see if anything has play or gives a bit @ xy joints and across the beam/carriage. The ultralight beam on larger printers have less torsional rigidity so that might account for some of the nodding. I haven't put mine on and run tests so just guessing for that part

[hasty siren]

that being said, if you increase accel/hz to 125 or till you hit aroud 1-3 1e5 scale, a lot of that noise might decrease and you might be able to get good prints still

[frosty pagoda]

I'll definitely mess around with a lot of this stuff later, and double / triple / quad check everything

[frosty pagoda]

Rails should be parallel, theyre pretty smooth and I used those installations helpers

[frosty pagoda]

They should be the right length iirc, will check now

[frosty pagoda]

Confirmed, same length

Will be checking bearing gaps now

Now would be the perfect time to find those rail alignment things..

Found them, phew

TLDR:

rails weren’t lined up length wise, they are now, so gantry looks good, no racking from what I can tell

Rails are smoother in the middle for some reason, when tightening I do from middle , and then slowly going out

Maybe I should redo belt tension, should I just completely untighten the screws on the motor mount?

[frosty pagoda]

Going through toolhead now

Hotend is 100% mounted securely

Brace on the back is secure

Extruder is tight

Definitely gonna need printed standoffs on the orbiter because of that filed down side

[frosty pagoda]

confirmed belts have same tooth count

[hasty siren]

Are you using a backer rail or printed nut bar

[frosty pagoda]

Using a metal thing with holes in it, which sits inside the extrusion

[hasty siren]

Siboor nut backer I had twisted the rails

That’s probably why it’s binding and smooth in different sections

[frosty pagoda]

Ah cool lmfao

New rails? Or new nut bar things

[hasty siren]

#1278041667311767717 message

The rail backer things weren’t perfectly straight when drilled

[frosty pagoda]

Oh jeez

[hasty siren]

So when I tightened down it twisted the rail slightly

And caused a bit of waves that had binding in parts of the rail and not others

But tbh I got not bad prints still with this because the rails had a bit of slop

Definitely much better when I moved to printed nut backers, is this siboor kit/siboor rail backers?

[frosty pagoda]

Formbot

So still el cheapo

[hasty siren]

Should clarify, these were flat before I installed, tightening them down caused the bends and the rail to also be uneven after, uneven hole spacing

[frosty pagoda]

Whhhh

Ahhh

I see

[hasty siren]

Yeah idk if it’ll be the same for formbot, try retightening the rail not from the center first

Have 2 end screws set to printed tool height

Loosen everything else and tighten one end to the other, sliding the tool to each screw before tightening down

Also check gantry is squared, touches both sides of motor mounts when back and both front idlers when forward

Had my z top horizontal beam slide back from x hitting it when tensioning etc. was a trapezoid with one corner further back so one side was longer from idler to motor mounts

[frosty pagoda]

It does, thankfully

[hasty siren]

Yeah should be good then

Try tightening rail down the side from one end to the other with the last 2 screws locked for vertical alignment first

[frosty pagoda]

Oh my god I’m an idiot.

Video incoming

[hasty siren]

Yeah can just drop the motor, put belt loop through hole, and slide belt over pulley + remount

No need to take apart motor mounts

Make sure the y rails slide smooth first though without belt

That’s the main thing from is graphs

And if you’re taking x gantry off id try putting fhcs screws on the upper bolts and the socket head ones in the lower In case the socket don’t push down enough on the joint

[frosty pagoda]

👌

[hasty siren]

This doesn’t tilt like this when everything is tightened down right?

Watch it front to rear

[hasty siren]

no binding?

make sure to go fast and slow and also with a bit of pressure on top of x while sliding as well both fast and slow after

[frosty pagoda]

Should I tighten my belts first?

[hasty siren]

no do it loose

make sure its even and freely moving, like only lubricant resistance and no notchiness in areas up and down the rail

I mean do it without belts attached to x first haha sorry

[frosty pagoda]

Yeah I took em off haha

Bit of resistance near the top (motors)

[hasty siren]

like free sliding no notching fast slow speeds

do you have a 200mm extrusion?

check the gap at the front and at the rear near the motors if so

[frosty pagoda]

The ones that the front towers screw into?

[hasty siren]

https://media.discordapp.net/attachments/1278041667311767717/1278500227153920051/656BDB91-6937-44C3-9337-6BC16CD6EB20.jpg?ex=679c0e9f&is=679abd1f&hm=303807cdf38e495f5c060c8bcfa8440d7f41d9db1f0162f7bcb4af9a62f55c75&=&format=webp&width=2168&height=2168

this is the opposite if its tight at rear and loose at front

but if you have a free extrusion you can just see if it fits between the y's above the rail, should slide in tight

if not then can measure or look for the sides moving in and out

as you move it forwards and back

https://media.discordapp.net/attachments/1278041667311767717/1278500226881032264/454C3720-B57D-4FB7-91D3-64734CF39DDD.jpg?ex=679c0e9f&is=679abd1f&hm=8d00e93563d7a58a4acf554a7327b971e16dde2d5e503ac8c7e101af5d5b4f2a&=&format=webp&width=1626&height=2168

https://cdn.discordapp.com/attachments/1278041667311767717/1278500226541424711/E246F199-D7B1-40F3-8C2D-9BB4C38EA8F8.jpg?ex=679c0e9f&is=679abd1f&hm=8ef687eda3f18aaef52f07bfd223be6f1ec93d33cfbd935f0f18dcb5a6f06657&

this is what I mean, I usually just use a spare beam to measure the gap and make sure it's squared if its binding at front or rear

[hasty siren]

oh not there

[hasty siren]

here, on top of the z extrusions

In the first pic, the extrusion slides in tightly between the y gantry right in front of the z extrusions, but can't go in in the front part of the y gantry in pic 2

[hasty siren]

because my white frame wasn't square so the sides came together at the front

yeah that looks fine

I wouldn't worry about it too much

maybe like half a deg off but nbd it looks snug both front and rear

reth has a gantry racking video without belts and with belts I think, can watch that to check for racking

or just push x back against motor mounts with loose screws, tighten it so its even and no gap either side, check front idler gap to make sure it contacts both sides

then slide it along y slow and fast to check for notchiness or resistance/binding

[frosty pagoda]

Gantry was fine until I attempted belts

[hasty siren]

after try putting a bit of force on top of the x rail carriage

[frosty pagoda]

Prob is fine now that no belts again

[hasty siren]

like a toolhead and do the same

yeah you're doing the belts like 1/8 turn each side right? alternating sides?

not just cranking down on one then cranking down on the second to match it?

[frosty pagoda]

I was doing one notch on the knobs at a time

[hasty siren]

bearing gap between everything ok?

It's like 19mm xy joints? and like something for motor mounts

[frosty pagoda]

Xy has 19mm yeah

[hasty siren]

if everything above checks out then belt it and see where you get binding

tophat also adds rigidity against side flex inwards and outwards so try with tophat secured if it sticks without it

[hasty siren]

last one might just be the last screw

if you loosen the last 2, put the guides on and loosly tighten to set in place

and then go back with guide and hold it close to the screw while torqueing down from second to last and then last screw

but yeah ends aren't too much of an issue I'd worry about if adjusting once doesn't change anything

[hasty siren]

your xy joints hit the front idler no?

or are they hitting the rail stops

[frosty pagoda]

Rail stops it seems

[frosty pagoda]

Video refuses to load 😂

[hasty siren]

Not sure if cnc mounts are supposed to be pressed against

[frosty pagoda]

Didn’t immediately seem like it

[hasty siren]

But v0 normal you want 1mm gap rails to back extrusion

And it’ll overrun the rails slightly on the front

To press against the front idler joint holder things

[frosty pagoda]

In the back I’ve got 3mm to the back extrusion it seems

[hasty siren]

Because that’s the mounting point that needs to be the same distance

If one rail is further forward by a bit then the end stop is further forward

So it may still knock against it even if one idler is further forward if that makes sense

Like skews it as it goes forward because one belt path is longer

[hasty siren]

Might be rail length shorter than 120mm

My siboor ones were a couple mms short I think

[frosty pagoda]

Weight savings🔥🔥

[hasty siren]

Yeah that’s solid no racking

Can bring gantry to the front against idlers and tighten rail stops against carriage

[frosty pagoda]

Thank god

[frosty pagoda]

Just did, seems nice (ish) now

[hasty siren]

also that back gap might be the reason why it binds slightly

[frosty pagoda]

I’ll move it to around 1mm

[hasty siren]

if its around endcap length

I'd just leave it

gantry looks solid and slides smooth

[frosty pagoda]

Ahh gotcha

[frosty pagoda]

For once

A printer I built, not having 70 issues, unheard of

[hasty siren]

cnc gantry has not a lot of margin for misalignment

[frosty pagoda]

I did notice that

The hard way

[hasty siren]

if you really wanted to try filing the gap then maybe

https://www.printables.com/model/368008-voron-02-sensor-less-y-axis-bumpers

but tbh I'd leave it if the gap isnt close to endcap length

balls will stay in and itll be fine

do the belts and tighten it

as you do one side there might be a bit of a gap on one side, tighten the other to straighten it

or if it's small enough increments itll just be no gap, shining flash from bottom up and looking at gap from top down for light to leak through is usually easiest to tell

if that makes sense

[frosty pagoda]

Yeah I get you, gonna take a few minutes for a break, my entire floor is covered in my once-working printer now LOL

Really appreciate all the thorough help

[hasty siren]

been through all this myself with the siboor white frame lol

[frosty pagoda]

Wouldn’t wish it on my worst enemy

[hasty siren]

wasn't fun and not a lot of info other than reths videos at the time

[frosty pagoda]

2am shenanigans

Everytime I drop something I feel it’ll wake up the entire family

[hasty siren]

but yeah when you do the belts and whatnots

double check front idler spacing and gap for racking

belt can pull idlers back uneven ammounts but cnc should be fine with that

[frosty pagoda]

Will do, probably have to loosen up the idlers first to put the belts on in the first place

I don’t have any slack, so it’s gonna be a lot of pulling

[hasty siren]

and if you get resistance mid movement,

repeat it at that point and look for belts walking up to pulley edge

[hasty siren]

ya know how to right?

[frosty pagoda]

I usually just loosen front idkers

Idlers ^

[hasty siren]

put then in x carriage flush

then back steppers out and loosen screws if really needed on the motors

and idlers to stretch it to the front

[hasty siren]

yeah

[frosty pagoda]

Will give this a try once I have the courage😂 (like 5 minutes)

Then gonna find a video on proper belt tensioning so I dont screw it up yet again

[hasty siren]

haha gl, should be fine

reth has 2 videos I think

[frosty pagoda]

Perfect, I’ll watch those

[hasty siren]

https://www.youtube.com/watch?v=1_ZWJWKanO4

https://www.youtube.com/watch?v=yevkZB6JC4o

[frosty pagoda]

Almost back at square one now

[hasty siren]

but you'll be fine if you check for gap real quick every twist

[frosty pagoda]

Will do, is there a chance of racking the gantry by improperly tightening?

[hasty siren]

if gap appears left or right then tighten opposite a bit till flush

yeah that's the last reth video

mechanical vs belt racking

[frosty pagoda]

Ahh I see

Man this v0 has come a long way

Getting the AC bed installed soon aswell

[hasty siren]

also move the gantry to the front and back every so often

to redistribute the belt tension evenly

[frosty pagoda]

Ahh yeah I dee

See

So tighten bit by bit, check gap

If gap on left, tighten right,

Move gantry up and down

[hasty siren]

yeah i don't remember if its gap right tighten left

or gap right tighten right

but you'll know because its the one you just tightened that creates the gap, fix it by tightening the other one haha

[frosty pagoda]

Equal and opposite reactions 😂

Should edpm belts be tensioned more than normal gates gt2 6mm? Think I read about it somewhere

[hasty siren]

and stop one side slightly below tension you want

tighten other side to catch up in tension

because both will go up if you do one side

[frosty pagoda]

Ah yeah I remember trying to make sense of that

[hasty siren]

yeah think so

haven't changed to epdm yet so can't really say what works best

[frosty pagoda]

I already have edpm, it’s a pain

[hasty siren]

might give you more vfa from small pulley

so start with 120

or 110 or whatever the rec was

[frosty pagoda]

110 iirc

[hasty siren]

best with tension meter thing

[frosty pagoda]

Yeah I only have the biqu one which seems pretty useless

Should’ve gotten that pfmakes

[hasty siren]

on mine, if i push both joints against gantry then frequency on right becomes higher, belt tension is even

without pressing it becomes lower because that side of the gantry pulls forwards slightly when plucking

[frosty pagoda]

So should I hold it against the gantry when plucking?

[hasty siren]

might just be me needing to look over my motion system

[hasty siren]

i try not to, that's how I put the top extrusion back a bit

I'd do by sound till I'm a bit under or on 110 for both belts with even sound

then use the biqiu thing to make sure its even on both sides

and move the gantry around a bit to redistribute belt tension before checking again, centered carriage and pushed back against motor mounts after moving around

[frosty pagoda]

Ahhh I see

Will definitely read over this while doing it

[hasty siren]

https://wiki.bambulab.com/en/x1/maintenance/belt-tension

lol

i also move the x carriage left right then center before doing this

but yeah watch reth video

[frosty pagoda]

Ahh gotcha

Will do

[hasty siren]

he's made a couple, there's one with 2.4 as well

same principle

good luck haha

also if you dont want to sand down spacers to the orbiter, you can just add spacers between x carriage and toolhead board mount

[frosty pagoda]

Will need it, definitely feel more prepared for all this

[hasty siren]

like a printed m3 spacer

[frosty pagoda]

Ohhh yeah true

That orbiter is uh

Well

It’s seen better days

[hasty siren]

still don't really know what part is sanded down

[frosty pagoda]

One sec

[hasty siren]

so have no idea whats best

[frosty pagoda]

Replacing the stepper or entire extruder LOL

[hasty siren]

oh

any 10t stepper should fit

did you get one with your kit?

[frosty pagoda]

I did not, I also ruined the thread in the body itself on that side

[hasty siren]

yeah if you do change it out and are planning to go HT with all metal and epdm belts

have a look at moons 10t stepper, runs much cooler

[frosty pagoda]

Yeah this thing is toasty at .8 run currentt

[hasty siren]

should still be fine to just measure the gap and print a thick spacer version for it fyi

or even just print one so its easy to sand down and match the other one

good luck with your build

[frosty pagoda]

Will need haha, putting in belts now

[hasty siren]

feel free to tag me if you run into any issues still haha but belts + toolhead spacer should get you 90% of the way there

[frosty pagoda]

Hopefully👌 really appreciate it, hoping to be back running IS stuff by tomorrow

[hasty siren]

x looked promising, y should be as well once its sorted

try the is thing, scv 5 and up the accel/hz till you hit low 1e5 and see if you get clean graphs with low noise, low vibration, and high accels

if its under 1e5 then up it till it hits it or if it starts having a lot more vibrations on the IS graph going up, revisit binding and nodding and whatnots

but yeah don't be afraid to shake it, anything that shakes loose wasn't going to last anyways

[frosty pagoda]

So up the accel / hz until I’m at 1e5?

[hasty siren]

yeah till its 1e5 top left and the scale on the left is around 1-3 at the highest peak

y needs higher than x from what I remember

rigid needs higher to get the shake going and get usable IS

higher belt tension does as well

[frosty pagoda]

Belts are on, tightening carriage now, then gonna start tensioning

[frosty pagoda]

my back feels like i skipped all of my 20's, and 30's

enough of that for today, gonna watch the vide on belt racking tomorrow

[frosty pagoda]

Trying to tension them evenly, yet I get some belt racking from what I can tell, gonna measure with the meter

-3.94 left side

-3.89 right side

The biqu tension meter says they’re at -3.74, and -3.78, yet the right side is wayyy racked

[hasty siren]

go off gantry gap after moving around

dont overtighten and use IS belt tension to fine tune

move toolhead around between runs

[frosty pagoda]

Yeah pretty confused, tightening them the same amount, in the back near the motor mounts they sit flush from what i can tell, no sound, no light, and in the front, it's not flush, the right side hits after

[hasty siren]

you can measure distance pulley to pulley

but most likely racked still, use IS belt tension to tension belts even

[frosty pagoda]

Yeah good idea, will do in a while, been messing with this for an hour or so by now

[hasty siren]

Yeah usually use belt tension meter and gap to ballpark it then IS to make final adjustments

but belt tension changes with temps, and stretch as they break in so might just need settling in after a bit of printing

[frosty pagoda]

Yeah that's fine, belts have just always been the bane of my existence

[frosty pagoda]

Alright, they're tensioned (ish), gonna install toolhead, everything else, and run the belt comparison

[frosty pagoda]

Also, think I fixed a lot of the flex

The CNC carriage calls for m2x4 screws, but they seemed to not bottom out all the way, tried some m2x6, way more rigid from what I can tell

[frosty pagoda]

Running belt comparison

Yeah that's uh, something

Does that mean Belt A needs a lot more tightening?

[hasty siren]

a 1/8th turn probs

[frosty pagoda]

And A is the right belt iirc, gonna try that

should i turn down the B belt while turning up the A belt?

doesn't seem like the A belt moved at all with this LOL

[hasty siren]

Moved up a bit

B might need to go lower if it gets too tight when cranking A

But drop both a bit below target and do the process again inching up both sides incrementally

[frosty pagoda]

I loosened belts to retry, even with 1 turn on the knob i can see some light coming through on the right side, unless i push it flush

It might still be racked I guess

Might take belts off, and double check it

[frosty pagoda]

Okay well, it’s not racked

So that’s always something

[hasty siren]

yep as long as it doesnt move and bump

but that doesn't matter too much, moreso to make sure its roughly even before IS

if you undershoot belt tensions a bit and use IS to tighten the loose one it usually ends up right at the range you want

make sure you move the gantry around full range of motion before running again or checking belt tension frequency

[frosty pagoda]

Knock on wood, I think I got it close ish

[hollow halo]

Late to the party here, but I just tensioned the belts according to the graphs

i used the 110hz as a starting and then adjusted them until the peaks matched

also, your belts look different than gates. Which belts do you have?

[frosty pagoda]

Gates, they're EDPM though

Also, got sidetracked today as I wanted to install my new graphite AC heated bed, part of it broke off, so that's sick

[hollow halo]

man you ordered all of the fancy parts

[frosty pagoda]

Yeah but that isn't worth anything if they're breaking 🤣

im gonna JBWELD it so no graphite particles are getting out, then still use it

[frosty pagoda]

really wish i'd made more progress this week with the shaper, since that's the actual goal of this, i get some jbweld on monday so i can seal the bed again, then gonna get that installed, and run some IS

[errant berry]

Any update for us?

[frosty pagoda]

Sadly not, got the RTV silicone, and JBweld for the bed, but they didnt have any wagos in store, and not available online with fastish shipping, tomorrow im going further into the city now to get hold of the wagos

[errant berry]

Sounds good. Do u have an ETA for us when we can expect progress here?

[errant berry]

So i dont have to bother u in a few days 😉

[frosty pagoda]

I mean, I'm hoping to get it done by tomorrow when the building shops open, need to get some stuff anyway

[frosty pagoda]

Grabbing wago’s now, and realized the RTV silicone I have, is in fact, not rtv silicone, so I’ll have to order that

[frosty pagoda]

Okay, i got everything.

Bed has been filed down, and applied JBweld to it, now I'm waiting for that to dry, then attaching heater, and thermal fuse tomorrow night ish, then wiring it up and hoping to turn it on

[errant berry]

Awesome, looking forward for the results 👍

[frosty pagoda]

you and me both, definitely seems sturdier than before i ripped it apart, so im hoping that my IS looks a lot better

[frosty pagoda]

Waiting for it to dry, so close to running it

It’s wired up now, gonna double check it when I wake up, then hopefully run it

[errant berry]

its been almost 2 weeks, where are we at? any update for us?

[frosty pagoda]

Sadly no update apart from the bed heating, unexpectedly got laid off from my job, so the printer has been low priority for a while, super sorry for the idle thread

[errant berry]

no worries, life and family goes first.

[frosty pagoda]

Yeah, friend is down to learn about printers though, so well likely have a look at atleast getting it moving by tonight

Knock on wood

[frosty pagoda]

Bad news

Went to test it now that i finally have some time, and a new job secured, and i can't get the ADXL to get recognized when using the ribbon cable from the picobilical

I'm suspecting the ribbon is somehow toast, gonna have to find a replacement

[errant berry]

How are things going? do u have any update for us?

[frosty pagoda]

have yet to get the ribbon cable working, nor find a replacement im 100% sure will work

[errant berry]

there are also plenty of USB C ADXL's u can buy 🙂

just a few bucks

[frosty pagoda]

Oh yeah, i forgot those existed, i'll see if i can find a mount for the rapidburner, then get one of those

good thinking lol

[frosty pagoda]

Arrives tomorrow, trying to print parts now, had a few moments of panic but apparently the belts are still getting settled lol

Printed the mount for the ADXL345 in surprising quality considering i haven't touched the printer in a while, and it wasn't set up well, when i was messing with it lol

[frosty pagoda]

ADXL delivered, gonna get it installed and set up now

[frosty pagoda]

Wondering, how much does the path of the USB cable for it matter? I can either route it outside the enclosure, then around and into the pi, or i can take it apart and try to squeeze it in with my bowden tube, any input on whether there's a noticeable difference in either?

[frosty pagoda]

If you've got a min to come with some input^

I wanna make sure i get good readings on this now that i can finally do it

[errant berry]

I have a minute 🙂

[frosty pagoda]

perfect, just trying to think of how to route this cable and if it'd have influence to just put it around the printer, all the way to the back

[errant berry]

You want the lowest possible stress on the cable and ideally nicely routed from toolhead to bed to the host

[frosty pagoda]

So something like the path of the bowden tube would be ideal

[errant berry]

If u share a picture of how its currently connected we might be able to draw you a line in regards of a ideal routing

[frosty pagoda]

well currently its just laid on top of the printer while i figure it out, ill grab a picture of my wire routing though

I’m thinking of trying the blue line approach

[errant berry]

@zinc gust take out your pen, its time to draw

Usb cable routing

[frosty pagoda]

Here’s a clean slate😂

[zinc gust]

That is a stiff usb

Need a more flexible lightweight usb cable

Is that an ldo toolhead board? If so does it have the adxl with the ribbon cable?

[frosty pagoda]

It does, i somehow managed to break some of the contacts on the picobilicals ADXL, and wasn't able to flatten them out again

Would a right angle braided cable be of any help?

[frosty pagoda]

I have the ADXL hooked up as close to the bowden as i can, with the right angle cable, and got the firmware updated for it, tomorrow I'll try doing some runs and post it in here, first day at my new job though, so we'll see how much energy i have lol

[frosty pagoda]

75 Accel/hz
5 SCV

Belt comparison

[zinc gust]

this looks like a toolhead issue. It can be a loose screw in the toolhead / x-carriage or it could be the adxl mounting. can't know for certain. But when you excite in y somethign is loose.
One thing you can do is setup your cellphone to record slow motion when you run the y-input shaper. This may be able to show you what is moving

@smoky lintel do you have a preferred setup for how to record this movement

[frosty pagoda]

Will try to record in slowmo, might be able to lie my phone on the handles for it

[smoky lintel]

I think the 50Hz and the 125Hz are so far apart that it won't be an issue here at 10-20k accels.

[frosty pagoda]

Slowmo caught nothing moving, that I could tell

[frosty pagoda]

Not really sure how to go about catching the stuff going on, on the Y axis, i can't see or hear anything out of the ordinary

[zinc gust]

Can you set both x and y at EI and print a cube.
lets take a look to see if the 125hz is actually an issue

[frosty pagoda]

Will do 👍 Should i turn down my accel since it recommends like 5400? or should i just keep it at where its at, (10,000 iirc)

[frosty pagoda]

Should I also go with the recommended accel for the El shaper (113 on y, 94.2 on x) or should i just put them both at 125?

[zinc gust]

go with the recommened values in the table

[frosty pagoda]

Also for accel? (Sorry for all the questions lol

[zinc gust]

actually use smoothei for y with an accel of 5500
for x use smooth ei also

[frosty pagoda]

Alright, will do, just any primitive cube, or would you prefer a voron cube?

[zinc gust]

voron cube = 3 bottom/perimeter; 0 top/infill

[frosty pagoda]

Will get printing, thanks for the help

that was quick

Option 'shaper_freq_x' is not valid in section 'input_shaper'

Once the underlying issue is corrected, use the "RESTART"
command to reload the config and restart the host software.
Printer is halted```

From what I can read both in the klipper, and Kalico configuration references, the shaper_freq_x is set under input shaper

#shaper hz 
shaper_freq_x: 73.2
shaper_freq_y: 48.6
#shaper type
shaper_type_x: SMOOTH_EI
shaper_type_y: SMOOTH_EI
#damping ratio
#damping_ratio_x: 0.1
#damping_ratio_y: 0.1```

okay, changed it to smoother_freq_x/y and that works, i should read bleeding edge documentation at times lol

[zinc gust]

needs to be all lower case

this is in reference to smooth_ei

[frosty pagoda]

Ah, does it not apply if I use uppercase?

[frosty pagoda]

I’ll attach some pictures too instead of the video

[errant berry]

ringing free cube

looks good

[frosty pagoda]

Yeah, that's with Smooth_ei, and 5k accel

Would like to bump up the speeds, and smooth out the graph considerably😅

[errant berry]

try other shapers even if they have vibration left and see if the vibration is ***really *** visible on the cube.

u also do not need to print infill

waste of filament

just follow the "how to post here" cube settings

[frosty pagoda]

Oh, that'll cut down on print time significantly

[frosty pagoda]

Have just found out reading isn't my strong suit, thanks LOL

0 Top, so it's just open at the top layer? (just wanna make sure im not misunderstanding)

[errant berry]

correct

[frosty pagoda]

Gotcha, thank you! I'll play around with them

[frosty pagoda]

Yeah pretty visible, atleast comparatively

[errant berry]

hm was worth a try

[frosty pagoda]

Agree, still not really sure what’s causing that huge peak on Y

[errant berry]

125hz is toolhead related

so i would start disassembling it and building it again, making sure everything sits correct, no parts cracked etc.

[frosty pagoda]

I could try re printing parts of it, just to have spares, that thing has been through a few attempts of getting it to fit

taking it slow and steady with the 2hump_ei for now

[frosty pagoda]

Considering switching to Yavoth instead, heard a lot of people talk about it

[frosty pagoda]

Waiting for a 5v delta fan, then switching to yavoth, should be way more rigid, also got some mdpc-x sleeving for the umbilical (warped a bit, but didn't seem to stay all the way through, so i guess itd be cosmetic)

[frosty pagoda]

Installing Yavoth now

[frosty pagoda]

Gonna put some sleeving on my toolhead board wires, then run some IS tomorrow👍

[frosty pagoda]

Big setback in the form of my umbilical cable factory crimps not being good quality, so the wires start slipping out, gonna get that fixed tomorrow when I get my microfit crimps, printer has been down for a few days because of it

[hollow halo]

canbus ftw

[frosty pagoda]

Looked into a NH36 or something like it, but the operating temp made me leave it be, 65c max operating temp for the NH36

Also update, got my FEP cables for the wires, no crimps for the 18awg, ordered some for friday, excited to have it running

[hollow halo]

Does your chamber get above 65C? Mine peaks at that temp and I have a reasonably sealed printer

[frosty pagoda]

Yeah, I hit 70c with not-great seal, its gonna be even more sealed soon once i can print new clips

[frosty pagoda]

Okay, after a really long time (again lol, sorry) the accelerometer is back on, about to run the shaketune suite yet again

Gonna send some pictures before i do so, of the wiring

USB cable for the ADXL is zip tied to the Bowden tube

[frosty pagoda]

Shaper graphs with 5SCV, and 75ApH

I don't even know how, or where to start trying to decipher this i'm ngl

[frosty pagoda]

I've measured my belts a few times with the PFMakes tension meter, and they both consistently sit around 2.1-2.2 on the meter, which afaik is normal for EPDM belts, not sure what's going on with them

[hollow halo]

could be belt path issues

do you see anywhere around the pulleys where there is belt dust?

also the obligitory make sure all screws are tight

[frosty pagoda]

will have a look around later today, all screws are tight though

[frosty pagoda]

Finally got around to it, noticing my panels vibrating a LOT, especially the door panel on the colony clacker, should I run IS / Belt comparison with panels on, or off?

These are my belts, significantly better atleast, though not perfect

The door when running IS💀

Also don’t think these clip on side panels are helping much, but the door is the main culprit, atleast sound wise

And that’s just on X^ Y is a lot worse, as im actively finding out

[errant berry]

do u have a seal between the panel and extrusions?

or foam

[errant berry]

just to see how it looks, could u attach the ADXL to your nozzle and measure again please?

[frosty pagoda]

There is foam between the colony clacker and the actual printer frame, nothing between the polycarbonate panel, and the colony clacker frame though, driftrotor suggested me some silicone sponge cord for that

[frosty pagoda]

Not sure how I’d go about attaching it to my nozzle, any ideas?

[errant berry]

yes u want something between panel and extrusions of the door

[errant berry]

by using the hole 🙂

[frosty pagoda]

Yeah, I ordered some stuff, hopefully here before the weekend haha, didn’t think that far

[errant berry]

its made for a M6 screw

so just screw it into place where the nozzle would be

[frosty pagoda]

Oh hahaha I see, so just grab a m6, and plop that in with the adxl between, instead of a nozzle?

I’ll see if I have any m6 screws, either way it’ll have to wait atleast until tomorrow so I can go get some

[errant berry]

BTT ships u washers and a screw with the ADXL 🙂 see https://biqu.equipment/products/adxl-345-accelerometer-board-for-36-stepper-motors?_pos=1&_sid=4999e48ea&_ss=r

[frosty pagoda]

I’ll see if I can miraculously find those, thank you, will update tomorrow as it’s late here👍

[errant berry]

sounds good 🙂 see u soon

[frosty pagoda]

While im doing the nozzle ADXL, and I don’t have the silicone sponge cords, should I take the colony clacker door off, to prevent it potentially skewing anything?

[frosty pagoda]

Also, I have a mellow nozzle ADXL bracket, which seems like i can just put that around my nozzle

https://injured.lol/g5azsu3q.png

[errant berry]

No dont take it off

[frosty pagoda]

Gotcha, would that mellow nozzle ADXL board be fine to use, instead of the BTT ADXL?

[errant berry]

Yes

[frosty pagoda]

Damn, can't get the device to go into bootloader mode to flash the firmware on it, that's weird

[frosty pagoda]

MCU Protocol error

This is frequently caused by running an older version of the firmware on the MCU(s). Fix by recompiling and flashing the firmware.

Your Klipper version is: v0.12.0-568-g23a5b65a

MCU(s) which should be updated:

mellow_adxl345: Current version v0.11.0-175-gcba119db-dirty-20230422_101631-firefly
umb: Current version v0.12.0-526-g3fcbc0b4

Up-to-date MCU(s):

mcu: Current version v0.12.0-568-g23a5b65a
rpi: Current version v0.12.0-568-g23a5b65a

Once the underlying issue is corrected, use the "RESTART" command to reload the config and restart the host software.

mcu 'mellow_adxl345': Command format mismatch: query_adxl345 oid=%c rest_ticks=%u vs query_adxl345 oid=%c clock=%u rest_ticks=%u```

Getting this, but can't seem to get to re flash it, since I can't get it into bootloader (somehow)

[frosty pagoda]

Got the nozzle adxl mounted (thanks for the help spitzbirne), and now my Y somehow looks amazing Hahaha

[errant berry]

@zinc gust which graph do we trust more?
the one above via nozzle ADXL

or this #1333602373335126067 message with toolhead board?

[errant berry]

welcome 🙂

[zinc gust]

I trust the toolhead board more. But the shape of this nozzle one is not that bad.
However, for diagnosing mechanical issues we need the toolhead board

[frosty pagoda]

The toolhead board one hasn’t worked for a while, so I slapped a USB ADXL on the front of the printer

I can just install that one and swap between them as needed no worries

I’ll do another nozzle ADXL run, then install the BTT ADXL on the toolhead and get some graphs there

This door sounds like a motorbike in a tunnel lol, glad I found out that I need to dampen it

[frosty pagoda]

Also, either the nozzle ADXL isn't very repeatable, or the fact that i put my top panel on with the 8 screws really messes with my IS

Or, i guess the cable / zip tie could be rubbing up against it

Tomorrow I'll re-mount the ADXL that sits on the toolhead and get some runs to figure out if / what mechanical issues are present, thanks again yall

[zinc gust]

The nozzle is like a pendulum.

[frosty pagoda]

Had a run with my ADXL mounted on the front of the toolhead

This is the inside cable run, and I changed out the door for a colony clacker recently ish, which im waiting for a fix for, with the door vibrating

Belt response:

[hasty siren]

You’re getting a lot of movement on accelerometer for 2 axis on the y graph, it’s going front to back (z blue line) but also up and down (x red line)

Have you tried taking a slow mo shot on your phone to see if you can find anywhere where the Notting is coming from potentially? Be sure to include any wires near accelerometer that could vibrate too

[frosty pagoda]

My Y axis is having a few issues, one being my door vibrating like hell, and also pretty sure a zip tie is touching my top panel

(Door in question)

[frosty pagoda]

Will try this when im home, is there any indication on whereabouts it could be? I know people say toolhead is usually like 125hz, don’t know if there are other markers

[hasty siren]

IIRC toolhead is 125, umbilical is around 100ish? Normal oeak around 75 ish?

https://tenor.com/view/egentlig-ark-rock-drake-ark-survival-cat-gif-26653544

See if your toolhead does this

[frosty pagoda]

Okay I’ll put that on my checklist

• Does toolhead nod like excited cat?

• use Hold_Freq macro

• use zip tie on toolhead 👍

• take door off initially

It’ll be a few hours since im at work, anything else to put on this important list?

[wet fox]

Use the Hold_freq macro to find the vibrations

[frosty pagoda]

Goof idea, will add that to my list

[hasty siren]

Yes sometimes it’s things boxing cause not tight, sometimes it’s the whole thing nodding

You should be able to see if you record a slow mo on your phone while running IS, to rule out door you could also just run it with the door off

[frosty pagoda]

Will do

[frosty pagoda]

Yeah im thinking to take the door off for a bit when testing

[frosty pagoda]

With this, do you mean i should remove the zip tie holding the USB cable and the bowden tube, and instead just ziptie the umbilical to the toolhead board?

[zinc gust]

umbilical to toolhead board

[frosty pagoda]

Perfect, should I keep the ziptie on the bowden and USB cable?

[zinc gust]

Rule of thumb, if the wires are being run together they should be zip tied together, and not zip together where they diverge

[frosty pagoda]

Ahh I see, that’s good to know

I might dismount the door now, then run another shaper test, then check for toolhead wobble later

[hasty siren]

can do all in one go

just be sure to record it while IS hits y

[frosty pagoda]

I don’t think im seeing the head ‘nod’ like that, just the entire head moving back and forth?

This was the X from that run

This was the Y

Also* The door was off during this run

[hasty siren]

This is from the y run when it hits 100hz on the console?

[frosty pagoda]

Ah no, i didn't think to time it

This was straight from the start, let me re-do it

[hasty siren]

Could just do a mini sweep from like 50-125

I’m putting my bet in beforehand though

My moneys on this at 95hz

If it is then maybe try changing it so it hangs downwards instead of overhanging upwards

[frosty pagoda]

I can't see any movement at those high hz of like 75-110hz ish

will swap the adxl position, then try to get another run on it though

I also feel like my shaky hands might contribute to me not seeing any movement in slowmo, ill see if i can prop my phone up on something

[hasty siren]

It’ll be easier if you shoot it from the front instead of the panels, or if you take the panel/tophat off

[frosty pagoda]

What would showcase it the best?

This is the Y axis with the ADXL mounted facing down

[hasty siren]

Side shot usually max resolution

[hasty siren]

Yeah looking at that it’s not unlikely that it’s wiggling around

[frosty pagoda]

Gotcha, can try that again with something to stabilize the camera

[hasty siren]

There’s more z in the second peak

For the downwards facing one

But it’s still wiggling a bit

[frosty pagoda]

Would i have to change any axis in my config, since its turned 180degrees?

[hasty siren]

Nope

[frosty pagoda]

currently its set as "axes_map: -x,-y,-z" from the btt config

gotcha

[hasty siren]

Is there any way you could

Like drill a hole in the middle

And just attach it like a nozzle?

[frosty pagoda]

https://injured.lol/ybb17v6w.png

like this one?

Also, that one already has a cutout so you could use it like a nozzle, but this one has a better shape, i think Reth said

but he also said, that to find mechanical issues, i should be using the toolhead mounted one

[frosty pagoda]

Oh in that thing

And then use the m6 screw to secure it?

[hasty siren]

Yeah basically seems like centre mount would be better cause it’s a pendulum otherwise

Yeah

M6 or something

[frosty pagoda]

yeah its m6, haven't managed to find the included one yet though

[hasty siren]

Do you have an m6?

Wait you have a nozzle one too

Does it fit?

That one looks smaller like it’d fit flat against it more no?

[frosty pagoda]

Probably not outside of the one thats included, which i likely have somewhere

https://injured.lol/7ikjp8cx.png

This is the other ADXL i have mounted currently

[hasty siren]

Yeah does the bolt holes match up to the front plastic mount?

Width should be the same

[frosty pagoda]

I can unmount it and check

[hasty siren]

Or alternatively

[hasty siren]

Flip the plastic mount

Attach the accelerometer

[frosty pagoda]

Flip it and drill a hole?

[hasty siren]

Make sure you don’t short the backside

(No hole sorry couldn’t erase that)

And tape it

[frosty pagoda]

Pretty confused, do you mean have one ADXL on each side of the ADXL mount?

[hasty siren]

Uhh one second

[frosty pagoda]

Like, revert the toolhead adxl mount to upright, then rotate the adxl?

sorry man my head is cooked after work today, need everything explained like im a 3 year old trying to put a square in the circular hole

[hasty siren]

Haha yeah to keep most of the weight over the 2 toolhead screws really

Idk if it’ll help much

This is first peak, z is blue in the graph and markings of movement

Like if the board is resonating, it’ll flex back and forward while the toolhead moves with it, so front to backwards position (orange = board angle) means the board will go more horizontal and it’ll clock backwards movement as more along the board vs through (x is along board, z is through board) I’m thinking this could be the second peak that shows more red x axis movement along the board vs movement going through

[frosty pagoda]

It doesn’t clear sadly, did try to mount it that way originally yo keep it over the mount

[hasty siren]

Does it change?

[frosty pagoda]

I can’t mount it with the adxl mainly facing forwards, as it doesn’t clear the screws for the adxl mount

So either I flip the adxl mount entirely, and still have it stick out, but at the bottom, or it stays

[hasty siren]

What’s the mount bracket

Can you send the stl?

And you don’t have an m6 bolt right?

[frosty pagoda]

https://github.com/chirpy2605/voron/blob/main/V0/Yavoth/STLs/ADXL_Mount_Front.stl

[frosty pagoda]

Not currently, haven’t been able to find it

[hasty siren]

Do you have a washer big enough to clamp it?

[frosty pagoda]

I can probably get an M6 bolt somewhere, maybe even during the weekend

[frosty pagoda]

took my panels off just to see if that made any difference

I haven't had any success getting slowmo videos of anything moving at 100hz on Y

[frosty pagoda]

Took my tophat off and panels off

Y seems a lot better now, so I guess the fact that the Y isn't locked into place has a lot of effect? maybe? I don't even know anymore

[hasty siren]

Huh interesting how’s your umbilical routes and is the tophat pushing it down

I was gonna redo the adxl mount for you but forgot

Will do tomorrow if you remind me

[frosty pagoda]

I'll get some pictures in a second, i'm running some more IS now

I've tightened all nuts with some spare screws I had, so nothing is loose anymore, I've taken off those top Y beams, and tightened everything up there, all bearing stacks, ab drives, xy joints, etc, all tightened but bearings still spin freely, now I'm running yet another IS on a concrete tile in the kitchen

[hasty siren]

Does the nozzle adxl fit on the front?

[frosty pagoda]

Have not been able to check, haven't dared to touch the toolhead since i put it on, had to take off the silicone sock to even fit it around the nozzle

Visually looks more or less the same

[hasty siren]

Isn’t it 2 screws out of the alu clamp?

[frosty pagoda]

Yeah it is, it's been moving almost constantly since I started working on it today

Will unscrew it between this, and the next run, once i find something to change

Most recent X+Y graphs after this, will go check every screw now to make sure I didn't miss anything, and also get you a picture of the umbilical routing

[hasty siren]

So I’ve just seen the nozzle graphs

Did you try a print with those settings?

[frosty pagoda]

I haven't, can dismount the nozzle adxl and put the silicone sock on tonight, then send a small print

This is the umbilical routing btw

You reckon I should remove the nozzle adxl for now? (Bit of a larger operation

[hasty siren]

Does the Bowden tube hit the tophat?

I’d zip tie the accelerometer to the umbilical instead of the ptfe

I did not have good results typing the ptfe tube in on mine

[frosty pagoda]

Yup

Well, not right now considering I absolutely nuked the top hat and all the panels of the printer

But usually, yes

[hasty siren]

Could see the Bowden transferring the vibrations down

Try zip tying it to the umbilical

And leaving Bowden undone or only like 1 zip tie on the rear half of you wanted to

[frosty pagoda]

Will try zip tying the usb to the umbilical

[frosty pagoda]

My Y didn't seem to like that, at all, way better results following the PTFE tube

[frosty pagoda]

One last thing before i head off for the night, how much variation is normal in something like belt responses? I'm seeing what to me seems like quite a bit of variation around 110hz ish between runs

Thanks for the help as always, got one more IS run tonight that i'll attach, if anyone has any ideas ill try to get through those

Pictures of all angles I could think of, in the current setup

[frosty pagoda]

Somehow it changed a whole bunch on Y since last night, and now, X also has some higher amplitude stuff

[frosty pagoda]

This machine is starting to feel cursed, maybe i should un-tighten all the extrusions, and use a square to make sure they're squared? idk if this could be a frame issue or what, but I'm very confused about all of this

[hasty siren]

It’s the axis movements

It’s wiggling in 2 axis

Hold off on that let me see if I can redo the mount a bit

[frosty pagoda]

sure thing, thank you

Will see if I can find an my bolt somewhere and cut it to a more reasonable size

[hasty siren]

Does a nozzle clamp it down?

[frosty pagoda]

yeah usually it’s be sandwiched between the nozzle, and where the nozzle threads into the hotend

Might have some heat inserts but they’re probably way way too long

I also don’t see how my accels halfed on both my axes overnight😂 so cursed

[hasty siren]

Also perhaps consider taping it to the toolhead till then

Cause I’ve done that before lol

[frosty pagoda]

yeah I just might ngl

I checked my belts and they’re at 2, and 2.1 on the pfmakes belt tension meter

[hasty siren]

Why don’t you just unscrew this and put it on the front?

[frosty pagoda]

Unscrew the nozzle adxl and mount that on the front?

[hasty siren]

Yep

[frosty pagoda]

Havent done it yet bc I didn’t wanna take the toolhead off again

Will do that in a few though

[hasty siren]

Do you have to take the toolhead off?

Is t it just 2 screws for the adxl to the aluminum nozzle clamp?

[frosty pagoda]

If I wanna take the entire thing off, I have to get to a screw that’s around the nozzle which is barely covered by the hotend fan duct

• I have to put the silicone sock back on which I might aswell do when I take the adxl off

[frosty pagoda]

I don’t think it would fit on the same mount, as there’s this small rp2040 chip on the back of it, hindering it from sitting flush

[frosty pagoda]

This is the closest I could get it sitting

The BTT board is basically flat on the back side

Maybe I can drill some holes so the mounting screws to the toolhead slot into place, allowing the BTT board to sit on the mount itself, instead of sticking out?

Like this

[hasty siren]

Yeah that’d be optimal

One sec have a bit of time in a bit to load cad

[frosty pagoda]

Okay I’ll just drill out the mount a bit probably

[hasty siren]

Put some electrical tape over screws

In case of short

Wait actually it needs to be parallel/flat

[frosty pagoda]

The screws are completely countersunk now, will add tape

There we go, sits flush, and used kapton tape to cover screws

[hasty siren]

oh yeah that looks much more solid

was gonna open cad

[frosty pagoda]

Agree, running belt comparisons now, just to see if it done goofed again, then gonna do an axis map

[hasty siren]

want me to modify or do you think that'll do?

[frosty pagoda]

I think that'll do, i did some DIY-countersunk screw modifications, in the form of my dads drill

[hasty siren]

if that doenst flop around and you're still getting 2 axis

I'd be looking at toolhead

cnc x carriage?

[frosty pagoda]

Yup

CNC-everything i could get my hands on

[hasty siren]

orbiter was a bit floppy for me on my toolheads as well so perhaps trying it with something like stock sb mini

then changing to things one at a time would rule out what compnent if its still messy

you get what the red blue and green lines are right?

[frosty pagoda]

Theyre the axis that something is vibrating on, right?, so if i get a bunch of red on a Y axis, something is causing that on the x plane

belt comparison is looking interesting

[hasty siren]

yeah it's movement in 2 planes on that

blue us z through the board, x is short side I think and y is long side

optimally you'd want 2 of the flat at the bottom and only movement in 1

[frosty pagoda]

Yeah

[frosty pagoda]

So since this is a Y graph, I’d want to solve the X and Z movement?

X graph has a lower amplitude on the higher frequencies like 70hz+, but its mainly Y that is acting up, when running that

And Y is having a really bad day at work

[hasty siren]

Yeah the y needs some working

Shake it at the second peak

See if you can identify where it’s shaking on the x

Axis

[frosty pagoda]

That's a nice way to say it LOL, gonna give it a go after dinner, thanks

[frosty pagoda]

https://injured.lol/pau4rw1n.png

Okay it did not enjoy that command

[hasty siren]

Huh

I wouldn’t worry about x for now, y tracing down the extra movement is probs more important

Cause if it’s toolhead nodding then you’ll have to solve that or have extra noise everywhere else

[frosty pagoda]

Yeah it doesn't seem to like doing that on any axis whatsoever, just shuts down my entire printer saying internal error on command

been trying to find it in my klippy log for a bit now, been unsuccessful

[hasty siren]

Interesting

What’s your electronics?

[frosty pagoda]

Yeah, I've never seen that before, trying to find something about it

[hasty siren]

Pico and pi4?

[frosty pagoda]

LDO Leviathan, RPI 4b 8gb

[hasty siren]

Ah can’t help with leviathan

Everything up to date?

[frosty pagoda]

Should be, will double check though, let me go plug in the printer again

[hasty siren]

Don’t have both adxl plugged in at the same time?

[frosty pagoda]

Nope, only one plugged in at a time, other one has also been uninstalled by now

https://injured.lol/sygt7bcb.png

All is up to date

Will probably ask in the klipper_discussion channel to see if anyone has any clue

[hasty siren]

What if you tell it to excite

At a range

[frosty pagoda]

Will try

https://injured.lol/wvk3039z.png
the accel/hz should just take on a default from my config file

input 75 as the ApH and it still just completely shut down

[hasty siren]

Ohh mainsail macro has a range

Maybe you can’t select accel chip

[frosty pagoda]

I've tried without

[hasty siren]

If you don’t select create a graph?

[frosty pagoda]

Shouldn't I get it aswell then, since I'm on mainsail?

[hasty siren]

I’m guessing you changed times as well?

[hasty siren]

Oh wut

[frosty pagoda]

Yup, I've tried leaving most of them at defaults, apart from axis

[frosty pagoda]

https://injured.lol/5ev7caqh.png 😅

[hasty siren]

Capital Y

?

[frosty pagoda]

will try

though the example is in lowercase

https://injured.lol/25lushtq.png

[wet fox]

https://gist.github.com/kmobs/f6def5db272ca5c1b81727482f53bed8

[hasty siren]

Yep throwing stuff at the wall haha I’m not an electronics guy

[hasty siren]

Ohh this is useful

[frosty pagoda]

You and me both

[frosty pagoda]

Will give this a try, thanks!

[wet fox]

found many problems with this 🙂

[frosty pagoda]

With how it’s going, im sure I’ll find a lot aswell

IS has been nothing but problem on problem on problem so far😂

[hasty siren]

Going to leave but

Try and get the single axis peak or find where it’s coming from

Usually it’s nodding or

Wires resonating

Or ptfe resonating, goes fuzzy on recording at that frequency

Also the belt peaks and whatnots instead of a clean 2 peaks could potentially be bearing stacks

[frosty pagoda]

Bearings seem to run smooth when I move the toolhead around manually

Can try loosening them though

[hasty siren]

A lot of the fushi bearings and washers are like .05 or more over/undersized so across a stack I’ve seen -1mm or so

[hasty siren]

Check the stack tolerance

Will have it in the manual

Like the gap

[frosty pagoda]

The v0 manual?

[hasty siren]

I’d leave the belts if they’re tensioned right and not racked

[hasty siren]

Yeah

From memory the xy joints are 19mm?

[frosty pagoda]

Yeah I already re-tensioned and deracked them once today

[hasty siren]

Motor mounts 18?

[frosty pagoda]

Will double check with the manual, thank you again

[hasty siren]

And idlers like 17?

[hasty siren]

Gl

Also the nozzle graphs looked fairly clean

Definitely run a print based on those settings

Maybe bump up accel per hz to make the top left number 1e5 and the highest peak at low 1e5

Might remove noise and reflect in print

[frosty pagoda]

Does this "clash" with ShakeTune? Now I'm getting this popping up https://injured.lol/gd2vnwbo.png

[wet fox]

shouldn't

[frosty pagoda]

I'm also an idiot, that's why ,i put it into resonance_tester.py, not resonance_holder.py lol

Got it running, no change in sound when I press on things though

34hz specifically is pretty bad

Sounds like it’s coming from the back ish?

Doesn’t sound like it’s on the toolhead

[hasty siren]

Press on the motor mounts tensioner

[frosty pagoda]

The big tensioner knobs?

[hasty siren]

The red tensioner motor forks

But the lower peak id leave

Chase the higher one

[frosty pagoda]

Ah, that is the lower peak

I'll give it a try in a bit, its in the living room right now and its really obnoxious lol

[hasty siren]

Be sure to record a side view of the toolhead when you chase the second one

That’s the off axis peak

[frosty pagoda]

Yup will do, so set it to like 90z on Y, then take a slowmo recording of it?

[hasty siren]

Yep would do touching first then when you think you found it record

[frosty pagoda]

Gotcha, will go ahead with that later, thanks

I’m still confused on where that low hz peak came from honestly

That thing willed itself into existence

[hasty siren]

Could be a lot of things, mounting noise etc, find the off axis one might clean up the others

[frosty pagoda]

Yeah, not worried about it, just perplexed haha

[frosty pagoda]

Okay, on 94hz I felt like a bit of it disappeared (not all of it) when pressing down hard on my mainboard, will get a video

Or I could just be tripping, it’s hard to tell

Not seeing any toolhead wobble at 94 either

Not hearing any change when pressing on the XY joints, stepper mounts, toolhead, front idlers, or the motor cover

[hasty siren]

record one of the toolhead

94.1 record the adxl from the side

slow mo

[frosty pagoda]

Will do

94hz

[hasty siren]

fuzzy cable and idler?

have another usb cable?

oh wait is that just the focus?

[frosty pagoda]

I do, not any 90 degree though

Yea idk wtf is going on with the focus lmfao

[hasty siren]

didja touch the idler and extruder parts