Unusual Resonance Profile

[blissful rock]

• V2.4, 350 mm
• Beacon ADXL
• Pin Mod
• Beefy Front Idlers
• Clicky-Clack Fridge Door
• AB: LDO 42STH48-2804AH, TMC5160T Pro, 24V, Double Shear
• Z: LDO 42STH48-2004AC, TMC2209, 24V
• Belt Tension tuned using BIQU tension tool

I'm wondering what could be causing those secondary resonance spikes. I didn't see these when I originally built the printer. However, I've touched almost everything since then, so that doesn't tell me very much. Nothing seems to be obviously mechanically wrong with the machine, and the additional spike seems surprisingly "clean".

[hidden sapphire]

👋

Please ping me or reply to this message so we know u are ready

[blissful rock]

I'm here

[hidden sapphire]

Holy moly, why 2.8A?

[blissful rock]

Oh, I was just testing how far they can go.

[hidden sapphire]

Yea no, dont do that

Change it back to 1.2 ASAP

[blissful rock]

This is within spec for the motors and drivers, though.

But sure, I can change it

[hidden sapphire]

u have plastic motor mounts mate 😅

[blissful rock]

Those are PC-PBT-GF. They'll hold up to like 160°C

[hidden sapphire]

My entire Trident is made out of pcpbtgf, i know.

Still it will melt it

Rule of thumb with any motor is to use the lowest possible current without losing steps at your desired speed

[blissful rock]

Yeah, I was testing how high I can go with accelerations

And it kept making a difference so I kept increasing the current to see if there are diminishing returns at some point

But there weren't 😄

It just kept going higher and higher

[hidden sapphire]

The biqu tool is only usable for comparing, not setting tension.
Get yourself the pf make tension tool. #replimat-🇩🇪-🇪🇺 sells hardware kits, thats the only valid tool

[blissful rock]

How come? Is their formula incorrect?

[hidden sapphire]

How fast did u print outer walls in terms of accel of the cube?

How does your input_shaper section look like

[hidden sapphire]

Its a wire gauge tool repurposed. Its not great and doesnt work as they sell it to

[hidden sapphire]

Install and configure tmc autotune please

[proven quarry]

@blissful rock https://github.com/andrewmcgr/klipper_tmc_autotune

[hidden sapphire]

A big limiting factor are the chains. Im pretty sure thats what we see in the graphs

[blissful rock]

Perimeters were at 120 mm/s, so pretty slowly. Accelerations were around 5000 mm/s² IIRC

[hidden sapphire]

Do u have another adxl besides beacon that can be used?

[hidden sapphire]

5k is above what u can print without ringing, hence why we see ringing on the cube 🙂

[blissful rock]

I do, but I don't think I can connect it anymore. The LDO power board it plugs into seems to be dead.

[blissful rock]

Talking about this

[blissful rock]

Yeah, I've tried it before. Unfortunately it caused quite a bit of skipping, even at the "performance" target.

I didn't expect that, but maybe I'm just running it too close to the limit.

[hidden sapphire]

Thats very very unusual.

It shouldnt cause skipping

[blissful rock]

I'll try it again

At least during the synthetic tests it happened quite a bit

[hidden sapphire]

Also please dont forget your input_shaper section, so we can verify u have the correct settings set

[blissful rock]

Yep, getting that now

[input_shaper]
shaper_type_x = 3hump_ei
shaper_freq_x = 93.2
shaper_type_y = mzv
shaper_freq_y = 38.2

[hidden sapphire]

With the VORON config default of 1.2A u should be able to reach ~10k.
How fast are u printer other features?

[blissful rock]

I've printed infill at 10k

But everything else is around those 5k

[hidden sapphire]

These are not correct unfortunately or at least not the latest.
Please update them with the values of the graphs u posted and reprint the cube at 4k outer accel with the settings metioned in the guidelines
Do not forget to add the damping_ratio to the IS section,

[hidden sapphire]

Yea for that u dont need 2.8A.
1.2 should be fine

[blissful rock]

Travel moves are at 40k right now though.

And they've been as high as 75k

When I was testing

[blissful rock]

These AB motors specifically seem to be pretty current hungry, but I'll see what happens.

[hidden sapphire]

I mean its your printer but to me that doesnt make sense 😅
Infill 10k but travel 40?

[blissful rock]

What do you mean by that?

[hidden sapphire]

They are, and they have a lot of VFAs most of the time at lower speeds. Thats why 2504 are better unless u REALLY hunt speeds 🙂 speaking vzbot speeds

[blissful rock]

I've been following the advice of people here to choose them. They said at 24V these would be better.

[hidden sapphire]

U said u printed infill at 10k #1399548886867120299 message and travel at 40k #1399548886867120299 message. that setup doesnt make much sense to me personally

[blissful rock]

Yes, I'm asking why.

What would you expect instead?

[hidden sapphire]

Infill 40k

[blissful rock]

I think I had PA issues, which made me go down a bit.

[hidden sapphire]

24 is better, maybe i missed your question in #landofschnitzel-🇩🇪 or u did not ask there, otherwise i would have advised to not get 2804 🙂

[blissful rock]

I talked about it in #voron_2_discussion I think

But I was going to reevaluate the accelerations anyway, since I was trying to find out what the IS artifacts are about first

[hidden sapphire]

Travel moves wont save u much time. Infill does. So i would always prioritize infill accel over travel. Push the hotend to its flow limit

[hidden sapphire]

Well next time u know where to ask 😉

[blissful rock]

I think my hotend might be the limiting factor right now for print moves.

Revo HF

I think I'm capped at ~25 mm/s³ with that.

So I was speeding up travel since it isn't limited by flow

[hidden sapphire]

Anyways, as i mentioned, u printed the outer walls too fast, thats why we see ringing. Lower it, update your config with the fresh values including damping and reprint it

Its not much ringing so our work here shouldnt be too much 🙂

[blissful rock]

I didn't realize SAVE_CONFIG doesn't update all of the input_shaper values.

I didn't even know about the damping parameter

[hidden sapphire]

With shaketune u have to enter them manually

And i would recommend to always double check in the klipper log if they are being actually active or wrong ones are applied

[blissful rock]

Makes sense

[hidden sapphire]

In terms of the vibration graph - with or without autotune doesnt matter for this information: u want to avoid the white areas as thats where u will get the most vfas.
Green is good

So before printing check the sliced preview in your slicer to verify u are not printing in any of those ranges

[blissful rock]

I see, okay

[hidden sapphire]

The accels u get recommended are what we would consider normal for a 350 with chains and SB

So we might be able to get a few hundred more but not much

SB is heavy
Chains add weight and vibration

[blissful rock]

Alright, so there's nothing fundamentally wrong with my movement system

I guess I'm surprised I didn't see these artifacts previously.

I'll see if I can find some old graphs

[hidden sapphire]

If the new cube comes out without ringing, IS does it job🙂

[blissful rock]

Here are some very early ones I found

[hidden sapphire]

Printed part > graph

At the end of the day the print is what matters. Thats also why the cube is a mandatory part of the guidelines

[blissful rock]

It looked way more consistent there, so I expected something to be wrong.

[blissful rock]

Yeah, makes sense

[hidden sapphire]

Y is basically the same.
ZV has 1.2% vibration, thats not what we want. We want <=0%, so on the old graph its MZV - which is 4k

[blissful rock]

Yeah, I meant just looking at the measured graph, it looks so flat compared to my recent measurements.

I'm wondering why it looked so smooth then.

[hidden sapphire]

Belt graph doesnt look great, but i doubt improving it will help drastically. Beacon is not ideal for belt measurements

[blissful rock]

I'll get the other tension tool and see how that goes

[hidden sapphire]

@hot forge do u mind having a look at the belt graph #1399548886867120299 message please.

The cube was printed at 5k outer wall, which exceeds IS by 1k, probably the reason it has slight ringing- i requested a new cube with fresh IS values already as the current config ones are not updated.
350, chains and SB - personally i consider the recommended accels as normal and doubt we could improve them drastically. Maybe a few hundred give or take - whats your opinion?

[hot forge]

@blissful rock what toolhead board are you using?

[blissful rock]

LDO's split board

Part of the kit

[hidden sapphire]

Its hartk's passiv breakout board thing

[hot forge]

Is it a NHSB?

[hidden sapphire]

No mcu

[hot forge]

what are you using for adxl?

[blissful rock]

Beacon

[hot forge]

can you use the adxl that came with the kit?

[blissful rock]

Maybe, I have to give it a try, but I think the PI power delivery board it plugs into is dead.

I'll have to get back to you on that one

[hot forge]

we need a toolhead adxl or the ldo adxl to see if the beacon adxl is good enough for our purposes

[hidden sapphire]

If the LDO thing is dead, there is https://biqu.equipment/products/adxl-345-accelerometer-board-for-36-stepper-motors?_pos=1&_psq=Adxl&_ss=e&_v=1.0 for a few bucks.
Can mount on the SB front+default adxl SB spot as well as on the nozzle if needed.

Question is if u want to spend the money on it or save it for a USB or CAN board - thats up to u to decide

If u decide to get the above, make sure to get the ADXL345 version

[blissful rock]

I was likely going to replace the chains with USB umbilical if I decide to actually pull the trigger on an alternative toolhead. But not as of right now.

I'll see if the LDO ADXL can be saved first.

The chip itself should be perfectly fine, I just have to find a way to connect it to the Pi.

Thank you for both of your help so far! I'll update this with more results when I have them.

[blissful rock]

Also, is it worth switching to 2504s? If so, what's a good place to buy those in EU? Specifically the long ones for double shear.

[hidden sapphire]

If u are happy there is no need to switch

https://www.omc-stepperonline.com/de/nema-17-hochtemperatur-schrittmotor-55ncm-77-93oz-in-55mm-rundwelle-isolationsklasse-h-180-c-17hs19-2504s-h-v1 they are the number 1 motors u will get recommended on the monolith server. They run pretty damn good

With 24 and 48 volt

I dont have the omc ones, but still 2504. this is my graph

#landofschnitzel-🇩🇪 message

Iirc thats 48k for everything

#voronuser_mods message

Around 85k was the limit

[blissful rock]

Huh, those seem surprisingly cheap. I haven't tried round shaft motors yet. Any issues with slipping?

[hidden sapphire]

https://www.cr3d.de/produkt/2gt-pulleys-geklemmt/ those are the best ones to get with roundshaft

[blissful rock]

Ah, that makes sense. I was wondering how well grub screws would hold up.

[hidden sapphire]

Totally fine. They dont sell 8mm ID ones so i have to use normal pulleys on my motors. No issues at all

[hot forge]

i found that https://provok3d.com/product/nozzle-adxl-board/?v=0b3b97fa6688 is a good one. It is small and can be put almost anywhere

[blissful rock]

Do 8mm shafts give you more torque?

[hidden sapphire]

But again, if u are happy, dont buy them

[hidden sapphire]

Nope

[blissful rock]

Why do you use them?

[hidden sapphire]

Unfortunately importing US things into germany, especially during the current situation, is pretty expensive

[blissful rock]

Yeah, import costs have only gotten worse

[hidden sapphire]

Because i prefer to not use things everyone uses. And im running a lot of tension compared to most 😉 also double shear on top of 8mm

[hot forge]

didn't know this was german.
Damn man you bringing all your friends in here 🙂

[blissful rock]

It seems that after a bunch of troubleshooting, the X rail carriage itself has rattled those screws on the side a bit loose over time. I've re-tightened them, and now it looks a lot closer to what I was expecting!

Now I'll check Y more closely as well, and see how smooth I can get it

[blissful rock]

I've gotten the new belt tension tool, but I'm not particularly happy with it's performance. It seems kind of inaccurate.

And I can't for the life of me figure out why my Y is behaving so poorly

[hidden sapphire]

Can u elaborate that more please

[blissful rock]

I've been trying to get more repeatable readings out of it. It's gotten better now that I've used it a few times. I was a bit disappointed how much the reading depends on how fast you let go, etc.

[hidden sapphire]

Did u apply grease to the tool?

[blissful rock]

Yes

The movement is smooth

But I didn't need to put any tension on the spring. It was exactly at 1.9 already.

[hidden sapphire]

Yup that’s what happens when you buy the correct hardware 😉

[hidden sapphire]

Once u tensioned both belts, run a G28, QGL, G28. Confirm tension is still consistent
If that is the case, run IS and belt again please

[blissful rock]

Tension was on the high end before this, but not by much. Now it's 2.7 on Z, and 2.1-2.2 on XY in a heated chamber. I'm going to run the tests again now.

[blissful rock]

Are these usable on a D-shaft as well?

[hidden sapphire]

Not recommended, no

Usable yes of course, why wouldnt they?

[blissful rock]

Because I figure the fact that even clamping force is applied and a part of the surface is missing, it would potentially actively misalign because of that.

[hidden sapphire]

But that would still make them usable, wouldnt it?

Normal motor pulleys on roundshaft arent centeref as well

[blissful rock]

Yes, sure. But I wasn't certain if there might be other issues I'm not aware of. I figured it's better to ask 😄

I haven't used that kind of pulley before, so who knows what kinds of issues I might run into

[hidden sapphire]

Most dont even know they exist 🙂

[blissful rock]

I bought two of them for potential future upgrades, but if there is no reason to put them on my D-shafts, I'll keep them in storage until then rather than putting them on early.

[blissful rock]

Well, they are damn expensive, that's for sure 😂

10 bucks each, or whatever it was

But I get that it's probably rare to need them that accurate.

[hidden sapphire]

Considering the difficulty to machine them, i wouldnt say they are “damn expensive”

[blissful rock]

Yeah, I meant that it's quite the jump compared to the standard ones. It might still be worth it.

[hidden sapphire]

Also the classic “angebot und nachfrage”🤷‍♂️

[blissful rock]

Yes

[blissful rock]

That's what I was getting at here

Tests are currently running, btw

https://tenor.com/view/waiting-spongebob-lonely-sad-raining-gif-5248000

With a heated chamber, btw

Since that's how they'll run in practice

But no running fans on the toolhead of course

Weird how it keeps looking different every time I run it 😂

Now X is worse again out of nowhere

Y still has that odd secondary peak

I've looked at it, I think the first Y peak is the toolhead shaking pretty violently back and forth, and the second one seems to be more of a distributed resonance frequency, I can't isolate a specific piece that vibrates more than others.

The chains don't seem to resonate much at all, surprisingly

Inconsistent measurements tells me something might be wandering, like a loose grub screw, maybe

I'll check all of those

[hot forge]

How is it printing?

[blissful rock]

This is at 300 mm/s and 3500 mm/s² based on those new calibrations. I probably still have to dial PA in a bit. I'm still hoping to get higher accels though, because it's obvious that that's by far the most limiting factor for print speed at the moment.

It never really gets to 300 mm/s at these accels.

[hidden sapphire]

3500 is in the low to mid range for a 350 V2.4

nothing drastically we can do about that

chains + heavy SB as i said

[blissful rock]

As I've said, this printer had way higher top accels in the past. I'm not sure why I should just accept it being lower now.

[hidden sapphire]

cube looks good ringing in terms of ringing. clean print

[blissful rock]

And the chains don't seem to have much impact on the resonances, at least. The motors have plenty of torque to move them.

From what I can tell, messy resonances are the only thing holding it back right now, so I'm trying to debug those.

[hidden sapphire]

yea of course the double peak on Y is not great and has a negative impact on accel. maybe u are lucky using the EXCITATE freq macro of shaketune to help find it

[blissful rock]

Yeah, I'll keep digging. Maybe I'll figure it out.

[hidden sapphire]

i think u are mixing things up here. careful, it has nothing to do with torque

[blissful rock]

Maybe you can explain?

You're talking about inertia due to the chains, I assume?

[hidden sapphire]

chains add weight, chains vibrate and chains change the COM as they are attached to the toolhead. these factors impact measurements

motors have enough torque to completely rip the chains apart

[blissful rock]

Oh you mean they change resonances too, yeah.

For sure

I'm not sure if they are part of those peaks though, because I've not seen them vibrate much during testing. I figured they'd flop pretty violently if they hit their resonant frequency.

[hidden sapphire]

using the macro i mentioned above + slowmo video capture can help

[blissful rock]

Yeah, I've tried the macro to find the previous issue with the carriage. Maybe I can isolate something else too.

[hidden sapphire]

its been 9 days. any update for us? 🙂

[blissful rock]

I've decided to replace the chains with umbilical to see if that makes a notable difference. I'm working on that right now.

Because I couldn't figure out which part exactly was causing that resonance spike

I've isolated one part that causes a lot of noise, though, maybe it'll help others in the future

It's the beefy front idlers

They use 2 screws to tension the belt rather than the one screw used by the stock idlers. What happened in my case was, that one of the screws was like a quarter turn too loose, so it didn't actually take up any tension. The screw was rattling against the idler face plate.

The carriage basically got lodged against the layer lines of the housing instead of being taken up by the screw.

Seems like a fundamental design flaw, tbh

It's not like the screws were totally unbalanced, it was really just a quarter turn off, probably due to vibrations over time.

[hidden sapphire]

glad u found something. thanks for the update.
do u have an ETA for us when we can expect progress here again so we have something to work? 🙂

[blissful rock]

I'll finish the conversion later today, hopefully

So I can get some new data then

[hidden sapphire]

awesome 🙂

[blissful rock]

It seems that changing to umbilical had basically no effect at all. If anything, it's a slightly worse shaper result.

[hidden sapphire]

send us please updated pictures of your printer

especially so we can see the umbilical routing

[blissful rock]

I'll add a spring steel wire to it when it arrives, but it already stays up on its own enough to not get in the way

[hidden sapphire]

dont route the umbilical like that

[blissful rock]

Like what?

[hidden sapphire]

https://youtu.be/IzYNO515-M0?si=nA4cL-7zNE1tzACP

Reth has A LOT of great videos, would highly recommend to watch a few

[blissful rock]

I was thinking about other ways to route it because I figured oscillation of the umbilical would be worse than chains, so I'll try some options.

Problem is that this is by far the shortest umbilical from what I can tell, because it doesn't need to stretch across the diagonal in three dimensions

This was by far the easiest way to route it using existing cables

So that's what I did first

I'm not a fan of having multiple different cable paths like through the back, but I'll see what works

I was also thinking about using a horizontal chain, which would probably be a lot more predictable than umbilical movement

I also don't quite understand how mounting it on the back would make it vibrate any less if it's the same length

Maybe you can explain the effects

[hidden sapphire]

Maybe 😉

It helps with how resonance/vibration is being transferred as well as how much the umbilical impacts IS measurements.

[blissful rock]

I have one of those retractable thingies for my VR headset cables, I'll jerry-rig one in there and see if that does anything, just out of curiosity 😄

[hidden sapphire]

nah they are not good

[blissful rock]

You're right, that just made everything worse 😂

It just spreads out the vibrations over a wider spectrum

[blissful rock]

So I ziptied it to the top to see how it changes measurements, and it's somehow gotten even worse accelerations 😅

I'm not sure what's going on, but I'm clearly not moving in the right direction.

[hot forge]

https://youtu.be/IzYNO515-M0?si=ts00dgqE1P0pQRND Take a look at this video

[blissful rock]

I did, that's why I moved it to this.

[hot forge]

in your picture above, I am not seeing a galvanic bowden holder, nor am I seeing the umbilical and bowden connected together

[blissful rock]

Oh, I missed that they all use the guide thingy

Bowden and cable loom aren't connected because they don't come from the same place. The bowden is fed in through a spherical bearing from the top, because there is no space to feed the printer from the back at the moment.

Do you think the looser bowden is a major factor?

I guess I should try without it connected to see what it does to the graph

Although it's been a constant this whole time, so I'd be surprised if it's responsible for the results getting worse.

I'm going to print the arm thing and report back with that.

If it's just for rigidity, I can add an additional PTFE tube to the umbilical.

[hot forge]

The Bowden thingy is to carry the weight of the Bowden and umbilical

[blissful rock]

I have some spring steel in this umbilical, if that makes any difference

[hidden sapphire]

remove it. its known to be a 50/50 thing

[blissful rock]

Alright

I only added that after the first umbilical measurements, so those are without.

[hot forge]

Ideally you want your umbilical and Bowden to have same path

Not sure what two different paths will do

[blissful rock]

Hm, I'll have to see if I can find a way to get the bowden back there without introducing a kink.

[hidden sapphire]

most route it out the exhaust 🙂

[blissful rock]

Yeah, that is unfortunately the worst spot for me, because the shelf the printer is on makes it very difficult to get filament to the back. It's quite the detour rather than going in straight from the top.

[blissful rock]

I've removed the spring steel wire, added the arm thing to the back and did another run of measurements. It got even worse on both axes now.

I'm going to try without any bowden tube to make sure it doesn't mess with the results now, then try adding a PTFE tube to the umbilical.

Here is the updated picture

The bowden tube being removed didn't make a huge difference, but it was a small overall improvement.

Now I'll test with a PTFE tube on the umbilical

[hot forge]

Are these measurements from beacon or toolhead

[blissful rock]

Beacon, but I've compared it to the LDO kit accelerometer before and didn't see a major difference in results.

I am not sure if ~130Hz is really the toolhead, because the major toolhead resonance seems to fall right onto that large spike when I look at it. It shakes pretty violently. Or are you saying there is some other secondary thing rattling around in there?

I'm testing umbilical with PTFE now. Afterwards, I can whip out the LDO accelerometer again if you want, to compare.

[blissful rock]

The umbilical PTFE tube didn't seem to make a difference in accelerations.

The second hump on Y shifted closer to the big spike, though.

[blissful rock]

The last spike in this one seems to be an outlier. I didn't see it on other graphs before or after.

[hot forge]

Please use toolhead adxl for troubleshooting purposes

[blissful rock]

Yeah, I'm running one from the nozzle right now.

[hot forge]

Not the nozzle. Either the toolhead board, or Ldo’s on the side of steathchanger please

[blissful rock]

Oh, didn't you say nozzle was best a couple of days ago?

[hot forge]

If I did it was a mistake in regards to troubleshooting

[blissful rock]

Toolhead doesn't have an ADXL, so I'll do it with that SB side mount.

Here's the nozzle results anyway in case those are interesting at all.

Since I've already done them

[hot forge]

The other ones were from beacon?

[blissful rock]

Yes

[hot forge]

Beacon no bueno

But use ldo adxl, let’s see what it says

[blissful rock]

Huh, really? Is the carriage that different from the toolhead for measurements?

[hot forge]

Yes

[blissful rock]

Interesting

I never noticed a big difference before

[hot forge]

Each position gives different readings because they are experiencing different vibrations

Or same vibrations in a different way

Does your toolhead have an lis2dw?

[blissful rock]

I don't think so. It's just the passive LDO SB board

No chip on there afaik

[hot forge]

Ok

[blissful rock]

Alright, running it with the side mount now

I suppose having the accelerometer further away from the carriage would show stronger toolhead vibrations, since everything is rotating around the rail carriage.

Here is the side-mounted ADXL

[hot forge]

your fine

I think we were chasing ghost caused by the beacon and/or x-carriage

this is similar to what I have found for carto

[blissful rock]

These accelerations still seem pretty low to me. I don't understand what's causing that.

[hot forge]

what is your belt tension at

[blissful rock]

Let me check again to make sure it didn't change

2.0

(hot)

[hot forge]

wait till it cools down and recheck. I suspect you are two low

[blissful rock]

Oh, I always tuned it hot. Was it supposed to be that high when cold?

[hot forge]

yes

[blissful rock]

HUH

Must have totally missed that

Okay, will do!

But yeah then it's definitely low

I have double shear too so I could probably even go a little above stock at this point

If necessary

I'll run the belt comparison again as well.

It does seem like something is off here. Although I've measured the same tension on both sides.

[blissful rock]

Wait, that's not from today 😂

I ran the wrong command

[hot forge]

which adxl?

[blissful rock]

The side-mounted one

[hot forge]

start looking at the toolhead

125hz is generally associated with toolhead

[blissful rock]

But isn't the issue at ~50?

That's the belt frequency showing as asymmetric

[hot forge]

50 is about where you want your belts

is this hot or cold

[blissful rock]

It was still hot when I ran this, but I'm waiting for it to cool down at the moment.

[blissful rock]

I assumed the issue here is that y1 and y2 are far apart, when comparing the belts

[blissful rock]

I've tuned the belts to 2.1 (cold) now

It looks better at least

Still has that peak

But the amplitude is a lot lower

Now it looks like this

Accelerations are finally going back up a bit

These are done while cold though, of course. So they might not be directly comparable.

[hot forge]

What are you measuring on the pf meter?

[blissful rock]

This was at 2.1

[hot forge]

Something is dampening the x, or creating that 125hz.

Did you check to make sure belts are run correctly

[blissful rock]

You mean if they are misaligned with an idler or something like that?

Rubbing on something?

[hot forge]

Double check belt routing and make sure the belts are not rubbing (double check xy joints) and they belts have the correct belt path

[blissful rock]

I will, but it would be very surprising to me

Just ran this too, just in case you see something interesting.

[hot forge]

Can you set both shapers to ei and print a voron cube: 3 bottom/perimeter, 0 top/infill.

Use a speed from the green zone for int/ext perimeter, and set ext/int accel to 3k please. Then post pictures of the four vertical sides

[blissful rock]

Will do

[blissful rock]

Here's the config for reference, and I'm printing the perimeters at 3000 mm/s² and 300 mm/s:

[input_shaper]
shaper_type_x = ei
shaper_freq_x = 55.2
damping_ratio_x = 0.053
shaper_type_y = ei
shaper_freq_y = 42.4
damping_ratio_y = 0.065

I've jerry-rigged the real bowden tube to go to the back now, but I'll need to find a better way to do it later. But at least this way it'll be comparable to the measurements.

So I don't re-introduce another variable by going in from the top again.

[blissful rock]

Seems like the slicer picked 120 mm/s instead, because 300 mm/s can't be reached on a short stretch like this at 3000 mm/s². But that's still in the green zone.

But I can't avoid them entirely, right? It has to accelerate through the bad zones at some point.

If I don't want to stay below 50 mm/s that is

[hot forge]

Based upon your cube I think you need to adjust PA first. A bad PA can create some artifacts which look like resonance

[blissful rock]

Yeah, I guess I need to redo that now that I've changed other settings as well.

[hidden sapphire]

Its been 14 days. Any update for us? Do u want us to mark this as SOLVED?

[blissful rock]

I've actually converted to a XOL toolhead now, because I got too annoyed by it 😂 Initial measurements look decent, but I'm still experimenting. I suppose we can close this, since I've essentially replaced everything that moves at this point.

That also means that I have a toolboard ADXL now, which has given me some pretty smooth graphs on first attempt

I can send those later if you are interested

Maybe that's a good resolution

Or you have some final tips for me

[hidden sapphire]

Yea u can

[blissful rock]

These were the first ones I did with the new toolhead, but I need to redo them later today, since I've tinkered with it a bit.

Y looks a lot cleaner now. X is surprisingly messy still I feel

But overall, much smoother measurements

I'm not sure what the second spike on X might be yet. Maybe it's the fans rattling or something like that.

Either way, I'm clearly limited by Y still.

Despite the much cleaner spike

Here is a picture from assembly

[hot forge]

X is fine for now. If you worried about the second peak in x don't be.
Set input shaper for ei for x and y, making sure to include damping ratio
then run vibrations profile

[blissful rock]

My bad, I didn't get a notification about this reply!

I'm currently finishing my new panel clips, but after that I can do a proper new set of measurements

That might eliminate some potential rattling

[blissful rock]

What's the reason for choosing ei specifically?

Compared to what the calibration recommends

Is it more predictable or something like that?

[hidden sapphire]

0% vibration

[blissful rock]

Ah

Is that generally the best choice, or are we just doing it for diagnostic purposes?

[hot forge]

So you can use mzv so long as your vibr% is less than 5, however you should do a print to verify. I like ei because I know it will be able to report 0% vibr.
This is the reason.

[blissful rock]

Yeah, that's good to know, thanks!

[blissful rock]

How much flex around the X rail would you consider normal? Mine does deflect by a couple of millimeters when pushing against the toolhead. Seems like a lot at first glance

I assume high acceleration will produce higher forces than my finger will.

[hidden sapphire]

Ideally none

[blissful rock]

Hm, maybe my MGN12 LDO rail isn't great, then

If close-to-none is achievable, generally

Do you have any suggestions for rails I could give a shot?

I assume I'd need a high preload rail

[hidden sapphire]

In my opinion LDO rails are bad

[blissful rock]

Would explain a lot I suppose

[hidden sapphire]

On Y and Z they are fine-ish

But if u had a good rail once, u know whats good and bad. Afterwards u can never life with a bad one

[blissful rock]

Yeah I assume on Z it doesn't matter that much, since you deal with some variance due to QGL anyway. And on Y you have two so it's more rigid.

What's a good rail?

I might swap my X at least

[hidden sapphire]

Depends a LOT on what u define as “good” as well as what your wallet says

[blissful rock]

Since that affects Y measurements a lot, I'm guessing

[blissful rock]

Of course, I'm relying on your recommendations for what's a good value 😄

I know that I could probably spend infinite money on it if I really wanted to

[hidden sapphire]

Ok so we are talking price/perfomance factor

Then berserker from #alchemy3d-🇩🇪-🇪🇺 are awesome

Rock solid and working grease ports

[blissful rock]

Oh that's nice

No disassembly is a big plus

This one? https://alchemy3d.de/products/berserker-viking-skis-by-west3d-linear-rail-mgn12h?_pos=2&_sid=8a784996d&_ss=r

[hidden sapphire]

Next step would be IKO https://alchemy3d.de/products/iko-linear-rail-lwl12-carriage-mlg12

[blissful rock]

Right, that's quite a step-up haha

[hidden sapphire]

After that we are talking schneeberger 😉

The carriage alone costs as much as the whole IKO

If u are lucky

[blissful rock]

How much better are these, in comparison?

To the berserker one

Does it make a huge difference, or is this gonna be close to diminishing returns?

[hidden sapphire]

IKO vs schneeberger
Very little. Schneeberger offers lifetime grease, so never ever regrease them. They are very much top of the line

[blissful rock]

Yeah, I figured that must be really close to optimal.

[hidden sapphire]

Berserker vs IKO
Noticeably but not by a lot.
There is a reason IKO is such a known company in this category- same for schneeberger

[blissful rock]

Yeah, makes sense

I mean 40 bucks seems like a no-brainer upgrade, tbh.

[hidden sapphire]

According to the product image the carriage also has lifetime no-maintenance, but double check that with #alchemy3d-🇩🇪-🇪🇺

[blissful rock]

But 170 or whatever it is for the IKO ones is quite a bit more

[hidden sapphire]

Yes. Make sure u use the mentioned grease

[blissful rock]

I do love "no maintenance" 😂

But I have no idea of whether the IKO rails are over 4x better I suppose, at that price

Damn, they don't work as weapons. Shame.

[hidden sapphire]

I know a few who can tell u that😉

@mental roost @fallow creek
Berserker vs IKO from #alchemy3d-🇩🇪-🇪🇺 : how much difference do u feel and see (regarding input shaper)? Worth it?

[mental roost]

I haven't put them in yet😅

[blissful rock]

Relatable

[fallow creek]

I Do Not have graphs to compare 🤣

[hidden sapphire]

And if u move the carriage, do u feel a difference? In terms of resistance due to higher preload

[hidden sapphire]

Samma
Da brauche ich euch einmal und dann so was

Has Migo the new IKO installed?

[mental roost]

you underestimate my laziness xD

[fallow creek]

Yeah much smoother, high Level rigid

[blissful rock]

You're really making me consider the IKO ones 😂

[blissful rock]

What are you comparing to? The Berserker ones?

[hidden sapphire]

It will hurt once, then u know it will be Everything but not the rail

[blissful rock]

Yeah, that's usually what I go for

Buy the best I can afford, assuming it was an issue to begin with

[hidden sapphire]

As reth said, graphs look good

[blissful rock]

Yeah. I'm just a bit sad that the Y accelerations have barely improved overall

And the toolhead is definitely the thing that vibrates the most during tests, so I figured making it more rigid would help

[hidden sapphire]

Do the prints have ringing while printing near the IS limit?

[blissful rock]

I'm doing another test later today to check it. I've had a lot of issues with the BFI screws rattling loose and screwing with belt tension.

[hidden sapphire]

Ok and prints in the past were good or no?

[blissful rock]

Alright I would say, yeah.

No major defects, but I'd still prefer it a bit faster 😄

[hidden sapphire]

Post a fresh print, if its good, i would say we are good to go here.
As the name of the channel, #1308016330326413393 , suggests we focus on basic help.
We do not have the capabilities to help everyone get to their desired goal 🙂

Our goal, hence the naming, is to have everyone with a good base to go on from there to whichever direction they want

[blissful rock]

Yeah, of course. My bad.

I'll see how it comes out now

And thanks for the additional tips so far!

[blissful rock]

This was done at 300 mm/s and 3700 mm/s²

ei shaper on both axes

Based on the results

[hot forge]

there is no resonance on these cubes.

[blissful rock]

Yeah, I only see those corners being a little bulgy

[hot forge]

what you see are some, PA issues, and speed issues. They are very slight

[blissful rock]

That's probably PA

[hot forge]

the buldgy corners are PA

[blissful rock]

Yeah

I retuned it, but maybe I gotta do it in smaller steps

So I suppose that resolved this thread, then

I will add the last shaper graphs for reference

[hidden sapphire]

as Reth said, no ringing 🙂
so u are good to go.

enjoy your printer