Belts are whacky

[pseudo mortar]

Never seen a chart this crazy before.
Voron 0.2 don't remember what toolhead been a while.
EBB36 ADXL345

Belts are 112hz +/-1hz measured using spectroid I was surprised they were so close last time I checked was 9 months ago. I (think) I set up TMC autotune this afternoon. Just trying to get the machine dialed in.

Hopefully I got enough information here if not then yell at me and I'll scramble and get it.

I updated klipper and now I got this if I try and run it again as well now.

[frank crow]

Need input shaper also please. And, can I get a side shot of the toolhead and umbilical path please.

[frank crow]

@lavish leaf you know what this error is? Do they need to downgrade klipper?

[lavish leaf]

I'd search for that error message here. I can't recall exactly

[pseudo mortar]

Ok I'll give it a shot.

[pseudo mortar]

I think its working fine on my trident. I was working on that then noticed the X ends are rubbing on my insulation so I am working on fixing that which turned into working on the v0 before I print different panel clips for it lol,

I would have to check what version that is running.

Trident is running 13.0.213 and this one is 0.13.0-435.

[pseudo mortar]

The umbilical just barely rubs on the roof when the toolhead is centered.

I checked all the screws on the A/B drives and shook it down and there's just a hair of play on the toolhead up and down but its good other than that. I think its got newer bearings and belts but I cant remember when that was. Few hundred hours ago maybe.

Got issues again trying to run input shaper so I guess thats where I need to start.

Luckily I have no idea what that means.

I did not update the display or the pi MCU could that be causing issues? No change after updating.

[nimble oracle]

Use klippers native version

See if that works

[pseudo mortar]

Ok dumb question what does that mean?

[nimble oracle]

To use klippers input shaping commands instead of shaketune

[frank crow]

https://www.klipper3d.org/Measuring_Resonances.html#measuring-the-resonances_1

[pseudo mortar]

Ah ok I see what you mean.

Testing.

Works native. @nimble oracle

Just from running that I think there is something wrong somewhere. I get twice the accelerations on Y than X I gotta look at the charts yet.

Y does not look like its going to have very nice peaks by that.

I would like to get shaketune working obviously as step one to get further information.

Oh I think I found a mechanical issue taking things apart to delve deeper.

[frank crow]

that may do it. The graph is really wierd. I have not seen that sort of double hump on y

[pseudo mortar]

Yeah it is strange how they are so similar. It is also terrifying that an expert isn't sure about it LOL!

I love putting a printer back together and wondering where the last screw came from. Especialy when the length is nothing close to any of the screws I touched.

Better.

[frank crow]

you have two graphs on top of each other

you need to delete the old data before redoing the graphs

[pseudo mortar]

Oh yeah I was just starting to wonder whats going on.

[frank crow]

I would double check your screws holding the y-axis together. That secondary peak is weird because there is not much z in it

[pseudo mortar]

Ok I will check again.

Would slight racking cause that you figure?

Nothing seems to be loose.

Slightly less with the tophat hinged up.

It is only seeming to be testing to 99hz why is it picking up that third small peak at 130

[frank crow]

Where is your adxl?

[pseudo mortar]

Using the built in EBB36.

I could try a different place with a different one I got a couple stand alone ones if you think that would help.

Think those fan wires would cause any wierd stuff?

I set the limit to 135 HZ here.

Slight toolhead wobble on the x rail. It should be a z1 rail.

Well what the heck.

Now Y is nice and X sucks LOL

I should probably post this.

I am out of ideas get back to me when you can.

Thanks!

https://github.com/Frix-x/klippain-shaketune/issues/259

[frank crow]

why are yuo using that adxl

that is not the ebb36 adxl

[pseudo mortar]

No it is not. That is one test using an external one to see if it was different.

Everything before that picture was the ebb36.

Sorry I wasnt clear there about what I was doing.

I am downgrading klipper right now to what my trident is running fine to try and get shaketune working.

Shaketune with the external adxl:

Belt chart pending:

This looks anything but exelent to me 🤷‍♂️

Better than before anyway.

That loose idler bearing was definately part of a problem. Maybe the ebb36 adxl but I want to switch back and see what changes.

[frank crow]

Your belts are not even. Something is going on in belt path a

A = right side front idler

[pseudo mortar]

What should I look for? Binding?

[pseudo mortar]

Also, what points you to that idler?

[frank crow]

The shape of the belts should be the same.
What could have happened is that you tightened on belt all the way then tightened the other

This can lead to different belt lengths even though the share the same tension

[pseudo mortar]

Ok. Tommorow I'll start over with the tension and make sure to get them equal. I do remember cutting the belts exactly the same length when I put this set on because I've been down that road and it is a pain to get them wrong.

[pseudo mortar]

My gantry was slightly racked.

I hope my 2.4 works to print some tools for belt tensioning because the other two are in pieces lol

I cant remember if it does. The bottom panel being removed is suspicious.

Its been a few months.

that was the last prints name looks like it was successful 😅

[pseudo mortar]

Same thing.

Nicer peaks though

Thats at 119 hz both belts. The B drive tensioner is very slightly further tight if that makes sense.

I wonder if my belts are not even length like I thought.

[frank crow]

It looks like you reduced the 125 hz

Give input shaper a run

[pseudo mortar]

Too late just ripped it appart to check belt lengths. Exactly the same.

Idlers are all equally smooth.

Is there anything should do on reassembly to get things working better?

[frank crow]

Check squareness of the frame and gantry.
Make small adjustments when you tighten the belts. Start with A-belt

[pseudo mortar]

Y rails are perfectly aligned.

[frank crow]

When you run belts the next time so long as there is not a huge 125hz spike go ahead and run input shaper

[pseudo mortar]

I did some messing around and the belts look more even now. I tried hard to get the teeth meshing in the carriage properly.

Belts are 120 +/-1HZ

[pseudo mortar]

Hang on toolhead was not centered for that one let me test in the middle of the axis.

I have no idea how to use the cross belt chart or what it means.

Note I still am using the adxl bolted to the X carriage.

I really like the numbers on that Y chart. Be nice to dial in the X.

[frank crow]

Use the ebb36

Increase accel_per_kz to 100, and reduce scv to 5

[pseudo mortar]

My bowden tube was rubbing on the top panel I did not put it in the right spot when I put it back together apparently.

huh

[frank crow]

Are you using a metal x-carriage?

[pseudo mortar]

Stock 0.2 carriage.

Unless the dragon burner uses a different one.

[pseudo mortar]

The machine is pretty much all stock. Just got a dragonburner and some different panel clips and lazycam idlers.

[pseudo mortar]

Don't feel the need to reply during Christmas.

[frank crow]

Have you checked for cracks in the x-carriage

[pseudo mortar]

I have not, no.

[frank crow]

Might want to do that. The cracks will be around screw holes.

[pseudo mortar]

Ok thanks.

I got some ASA ordered. Polymaker had a good sale so I got 10kg coming instead of the one I meant too lol.

[pseudo mortar]

Doesn't look like it although I see how it would split the layers apart with those nuts.

[pseudo mortar]

Will fan wires touching the belts right beside the carriage cause anything wierd?

I am going to print a cube and then run tests again in the chance something needs to settle down.

[pseudo mortar]

Smoothed it out some.

[frank crow]

Go ahead and try a test print setting x and y to ei

[pseudo mortar]

My slicer profile is 10mm/sec2 so we will see how this goes.

I would like to get it to 20k again. Thats the goal.

[nimble oracle]

make yourself very aware of the name of this channel please #1308016330326413393 . Its "basic", so our goal is to have everyone print without VFAs and ringing.
for anything beyond u have to get a serial.

[pseudo mortar]

Thanks for the reminder and I apologize I was getting too far ahead. I'll have to see about getting my cereals back.

[pseudo mortar]

I'll turn down the acceleration to within the max recommended and do a quick PA tune and see how it looks.

[pseudo mortar]

Print settings are Voron part spec. Filament is sparta abs+ 110 and 270 degrees. Acceleration was set to 7420mm/sec2 and speed 200mm/sec SCV was 5. I tuned PA prior and it is at 0.019 PA and ST is 0.02. There is slight ringing and some VFA but I printed perimeters inside first and it could be infill so I'll try outside perimeter first. I'll change some settings and see what happens.

That top layer though 🤤

Oh thats why it took so long. My layer height is 0.1mm.

[frank crow]

I am not seeing ghosting per se.
I think what we are seeing is VFA

can you run vibrations please

[pseudo mortar]

Yeah looking at the newest cube I would agree.

[frank crow]

Stay in the green zone, and avoid the peaks in angular speed energy profile

If you can print it I think 105-120 for extra perimeters should be good

[pseudo mortar]

Ok I'll try that.

If we can get this machine printing nicely then its good enough for now. I'm getting tired of it and the trident is unhappy too. Thats for another day.

[pseudo mortar]

Still got VFA at 120. If anything it is more visible.

[frank crow]

can you post picture please

try 150

[pseudo mortar]

Testing.

[pseudo mortar]

Best yet but it is still there.

[pseudo mortar]

What do I need to check to elimate the VFA? Once I get my plastic I might rebuild the whole gantry even though it was done just over a couple hundred hours ago.

[frank crow]

you can also make sure no belts are riding on the flanges of the pulley and bearings

[pseudo mortar]

Ok I'm pretty confident that is not an issue but I'll see. I'll probably rebuild it or at least take it all appart to find whatever is causing the problem.

[pseudo mortar]

Maybe I shouldn't see what input shaper does on the 2.4 before I print the 0.2 gantry parts lol

It prints the best out of my printers albeit slow accelerations. Some very faint VFA. Gonna see what its vibrations look like then probably ignore anything else.

[pseudo mortar]

I got some pretty dang nice parts printed for the gantry now on my 2.4. I love the V0. 10 minutes and the gantry is disassembled.

Bearings all move super smooth I think they were new last rebuild and I also made sure to pack them with extra grease.

[pseudo mortar]

I found an issue. I think one of the stacks was missing a shim. I am missing one on reassembly. Now I could have easily lost it but I don't think I did.

[pseudo mortar]

Nope found the missing shim hiding in my bit tray.

I got it all back together gonna do a heat soak for a while and then print a few things then do some tests. It just feels smoother moving the toolhead around by hand now and I don't know why. It was never rough before. I never found a smoking gun.

[pseudo mortar]

I made sure the belts ride good on all of the idlers and in the middle of the drive pulleys.

[pseudo mortar]

I think I found a problem.

[frank crow]

???

[pseudo mortar]

Getting info

So I don't really know how much of an effect that difference of the TBL on Y being define and not on X but thats not right to me lol

[frank crow]

I do not know either. Comment out the TBL and see what happens

[pseudo mortar]

The default value is 2 so the blank one shoiuld have assumed as much but we will see.

[frank crow]

I generally do not set those individual values

[pseudo mortar]

Yeah I am not sure why I did. It was weeks ago too so I should know but I don't 😅

I coppied it form my 2.4 which also has the X and Y differing and it has no VFA.

[frank crow]

Did we already take the belts completely off and reset

[pseudo mortar]

So what I did was replace every part in the gantry and check bearings and reasemble. I then printed a few boats and let it heat cycle a few times. Apparently the belts really settled because I had them set to 120 and now they are a few under 110 so I'll even them out to 110 and run another test. Is there any other information that would be helpful to you?

[frank crow]

No

It looks like belts loosened.
What belts are you using?

[pseudo mortar]

Gates GT2.

I got them to equal at 113hz so I will print something and then verify their stability and then run a test.

[frank crow]

That should not be a big issue

[pseudo mortar]

Ok I thought not.

Belts are whacky

[pseudo mortar]

I tightened B a little and am running another.

[frank crow]

if your prints are coming out good that is all that matters

[pseudo mortar]

Yeah I need to run input shaper and print a black cube and see if the VFA is gone. If not I'll have to decide what to do next.

I think that may be beyond basic resonance help?

[frank crow]

Depends on prints

[pseudo mortar]

Ok. Let me know if I am getting out of bounds I don't want to be a bother.

[pseudo mortar]

What shaper would you suggest?

[frank crow]

@fair rivet are we seeing harmonics here?

[fair rivet]

Yes. That's what all the parabolas on the vibrations heatmap are. There's a couple of basic resonances and their harmonics. The peaks happen at certain angles where they constructively interfere, e.g. peak 2 is because at 0, 90, 180 and 270 degrees the strongest resonances add together.

Peak 3 is where the 1st harmonic of peak 1 and the main resonance of peak 2 add together, and so on.

[frank crow]

I was talking about the I put shaper graphs. Is that what you are talking about

[fair rivet]

That's the vibrations graph, but on the shaper graphs you see those diagonal lines are all harmonics

[frank crow]

I am at a loss on this thread. You got any ideas

[fair rivet]

The intermittent lines you see in Y are distortion... something is vibrating hard enough to collide with something else and create extra harmonics

That should be very audible

[frank crow]

@pseudo mortar peak 1 in both x and y are the “real’ peaks. Try excite at vibrations on the other ones and see what gives

[fair rivet]

Peak 3 on Y is probably something loose in the toolhead. Endstop lever, fan rotor, some wire or connectors, loose screw, something of the sort.

Possibly XY joints or a rail mounting screw

[pseudo mortar]

I know the gantry is all good and tight because I put it together with new printed parts just this morning and ran a few prints and re checked. I was going to look at the toolhead next.

[fair rivet]

At that frequency, it should be pretty noisy and easy to track down by ear

[pseudo mortar]

There is one noise that I know is my electronics fan buzzing a bit on the extrusion it is mounted to.

Surley something on the bottom of the printer wouldn't pick up on the toolhead that much?

[fair rivet]

No, and it should appear as a more or less vertical line in those plots. I don't see it.

Maybe a faint purple line at 140 Hz in the X graph

[pseudo mortar]

Its not the fan running it is the mount the fan is on shaking on the extrusion.

Induced by printer vibrations if that makes sense

[fair rivet]

Ah. Well, try taping it down and running again. If the graph changes, you know what to do.

[pseudo mortar]

Ok something sounds loose exiting the third peaks frequency.

[fair rivet]

Yup

[pseudo mortar]

Kind of like a nut loose in an extrusion. Its hard to pinpoint.

[fair rivet]

That would do it

[pseudo mortar]

It sounds like it is coming from everywhere and nowhere lol

[fair rivet]

Aluminium is pretty resonant and carries sound everywhere

[pseudo mortar]

Aha!

I found that little rascal.

Still got the sound even without that little nut hiding on top the end stop nut.

[fair rivet]

Something else then

Still, good to catch these things

[pseudo mortar]

Yep. I'll find it yet.

Yep. I just noticed a loose z screw mount loose too.

I bet it is half my X rail screws are unpopulated but if I look in the holes there are nuts.

[pseudo mortar]

Same noise at the second peak too.

I fixed a few minor loose things with no success. Its 11:30 and I work tomorrow so I am going to call it a night and investigate it further later on. Thanks for the pointers so far!

[pseudo mortar]

I notice that there is a 123 hz peak on the belt and both input shaper charts which is interesting.

I gotta find whatever sounds loose.

I am doing some digging around trying to learn more about printers resonance I find it very interesting. When I get interesting in something I like to know how it works down to the minutia. I will figure it out eventually. @frank crow Your videos are very useful and I appreciate the time all of you guys have put into helping the dummies out.

[pseudo mortar]

This is real frustrating. Been looking for a couple hours now.

Can rails make noises? I have checked everything. Still does it with the toolhead removed.

It sounds just like a nut or something in an extrusion.

[pseudo mortar]

It has to be my rails. My brand new 200 hours ago rails. Oh well. I'll try a clean and lube and inspect em close.

[pseudo mortar]

One of the Y rails looks like it was sliding part time instead of rollling.

[pseudo mortar]

I think I got rid of the vibration.

Only took 5 hours of work.

Hopefully charts look good. Gotta do some reassembly.

It is gone!

All I did was take half of the printer apart no big deal.

I think it was the rail nut bars something rattling in there on one of the Y ones.

I also cleaned and re greased the rails and all that.

My deck panel vibrates some too I need to shim it with something.

[frank crow]

Do you know what was doing it?

[pseudo mortar]

It was that most likely.

I had to take appart part of the frame to replace them.

[pseudo mortar]

The deck plate was also resonating I noticed. I just shimmed it so it won't move with some pieces of an ellis PA pattern between the extrusions and panel LOL

[pseudo mortar]

There are other things I am noticing vibrating. I think the URBI is dancing a bit at the higher end. I'll check back in when I get nothing sounding loose.

Thats where it is at now. Little better looking. I will play with the belts and see if I can improve it with some tightening or loosening of one of the drives.

This looks a little odd to me.

I don't know how to interpret that.

Ah I forgot to check my belt tensions after printing something after reassembling. They are a little over 100 Hz so I'll bring em back up and try again.

[pseudo mortar]

I found some powge idlers I could try as well in my parts stash although I think it printed without VFA forever with whatever formbot sent me years ago.

[frank crow]

Looks like something in the tool head

[pseudo mortar]

Well the URBI is flopping back and forth is the latest thing I found at certain frequencies and the Bowden tube runs through it so I think that could be in turn shaking the toolhead some.

Once I cure all audible vibrations I will look into the toolhead.

[pseudo mortar]

Is that due to the x+y+z is higher than the actual axis on the Y chart and that last peak has a different axis resonating? I want to understand this.

I think I have got the noises fixed other than the spool I have loaded seems to resonate at times

Something with the rails or nut bars was loose, the nevermore was loose, the deck plate was loose, and the URBI plate was resonating.

[frank crow]

it is the 125hz in in y

[pseudo mortar]

I just measured the VFA and it is 2mm pitch.

I printed a boat and if it wasn't for the VFA it would probably be the nicest benchy I have ever printed.

[frank crow]

Try speeding up the ext perimeter speed.

[pseudo mortar]

Ok I'll try that next. I should take apart the toolhead at some point and look for anything loose too. I just loosened the pulleys and made sure that the belts are as centered as possible. The one was a little close to the edge.

[pseudo mortar]

Could that be an X rail issue you think?

[frank crow]

Wait are you not a helper.

I would look along the a-belt path

Potentially xy joints

Can you post new pictures of the xy joints, front, left, and right of toolhead please

[pseudo mortar]

Yep. I am wondering if it isn't two issues. Toolhead vibration and a belt.

[pseudo mortar]

Right now I need to rebuild the toolhead and I have been busy the past few days. I'm hoping to get something done soon again.

[frank crow]

Don;t worry we charge by the hour

[pseudo mortar]

Do you finance?

[frank crow]

We will need to run a credit report, we charge 43% interest, and have a 200 usd sign up fee

[pseudo mortar]

I'll call my bank.

[frank crow]

let me know

[pseudo mortar]

They sounded concerned and kept saying a word like spam or scam.

I don't know why they were worried about canned meat 🤷‍♂️

I'll just sell some organs.

Or do you take kidneys?

[frank crow]

nope. Cash or gold

[pseudo mortar]

Shoot. I was hoping to eliminate the middle man.

[last forge]

@frank crow is probably a liver and onions kind of fella..

[frank crow]

Sir this is a private thread.

[pseudo mortar]

Just what are you implying?

[last forge]

it is a fact that your liver can regenerate... just throwing that out there

[frank crow]

Can I get a debt collector (i.e. moderator) in here. Looks like we got a person who ain't paying their bills. And, we have another that looks to be causing some trouble

[pseudo mortar]

Please I can pay just give me some time!

[frank crow]

https://tenor.com/view/time-is-ticking-hurry-hustle-late-niecy-gif-12424862

[pseudo mortar]

I almost got the new toolhead and extruder done just gotta print a EBB36 holder and shoot I gotta mirror it I just realized.

[frank crow]

cool

[pseudo mortar]

Tiny bit of rail slop

I am not 100% sure if that is the z1 rail or it ended up on the Y

I am about 87.78799% sure.

The only difference is some Chinese engraving on one rail and that should in theory be the preloaded one.

[frank crow]

what x-carriage are you using

[pseudo mortar]

Stock 0.2

I just fired it up and the electronics fan made a horid noise and self clearanced itself or something 🤣

[frank crow]

I shall say a prayer

[pseudo mortar]

Took a peak off!

I will do some test prints tommorow. Gotta fix some heater wires apparently.

[frank crow]

Y=umbilical
X=toolhead

[pseudo mortar]

Could you explain why you would draw such a conclusion for me sir?

[frank crow]

let me think again

So for x, the hook at prepeak 1 and peak 1 is generally indicative of a loose screw. In addition, post peak around 125hz are associated with toolhead.
In regards to y, I thought that peak 2 was z, it is not is y with a little z. This is similar to what we see with canbus cablles, or tap's magnets not fully engaging with the screws.

In addition for your belts, A belt is looking like it mimics the toolhead, and your b belt is really low

What I think you have happening here is that A belt is tigher than B

Try loosening A belt and see if B starts to take shape

[pseudo mortar]

Thanks for the explanation I'll work on that tonight.

[pseudo mortar]

https://tenor.com/view/look-at-this-graph-nickelback-gif-12683147

Belts are at about 125 hz each here.

Gotta do some work yet. I need a real tension tool. I am just using spectroid or exiting at a frequency and tensioning til belts resonate.

[frank crow]

@pseudo mortar

[pseudo mortar]

Im screwed no matter what I do. I have not a unicorn burner.

I need to get an actual tension tool. PF tension tool i remember now watching that.

With my bigger printers I tension them exactly the same way as you described. Its a lot more tricky on a V0. I don't think I have any belt induced racking.

I may buy some new belts and pulleys and rails to rule some things out.

I worked on my truck for a lot longer than expected at work so I only had time for a quick tension and test this evening.

Hopefully tommorow is a bit different.

[frank crow]

V0's suck. One thing that will likely break on the v0 is the plastic x-carriage

[pseudo mortar]

Brand and new.

Course nobody sells the PF tension tool in canada yet. I know I can make one but that sounds like a bit of a pain.

55 bucks to ship a 15 dollar parts kit from west3D. I suppose I am making one.

[lusty prism]

What do you mean by that they duck? I built my v0 recently and the mention made me worry lol

[pseudo mortar]

Suck to tension stock compared to a trident or v2 with screws for tension.

[lusty prism]

Ahh sorry my bad then. I agree lol

[pseudo mortar]

My V0 is my favorite printer. I am just having some issues with it currently.

[frank crow]

@pseudo mortar I agree the belts of a v0 suck in getting them equal.
Have you tried the tulip mod?

[pseudo mortar]

Not yet I have been thinking about it.

[frank crow]

When I had a v0 I think I went through 2-3 sets of belts before I figured out how to make them equal

[pseudo mortar]

What did you learn?

The teeth counts are the same but the danged carriage could be off by a hair it's tough to get them in there good and equal. They seem equal but a half a mm might be enough to throw things off with such a small machine.

[frank crow]

Make belts equal to start, and try your hardest to get a good pre-tension on the belts before actual tension, and then tension belts without toolhead on to proper tension and make sure everything is aligned

Cutting the belts should be the last thing

In addition I learned that umbilical matters, and there is nothing really carrying the weight of the arch

Arches are bad for umbilicals umkay. Waterfalls are better

But v0 does not allow for a waterfall

[pseudo mortar]

Would adding a spring wire help any do you think?

[frank crow]

no, you want to remove springs from the system

[pseudo mortar]

Ok.

[frank crow]

a flat spring steel may help. Similar to N3MI's umbilcial mod for stealthchanger, but don;t have the spring stell extend the entire way though the umbilical

[pseudo mortar]

Is z1 preload on x and y a bad idea?

X sure Y I am not sure if that would drag extra.

[frank crow]

Z1 is fine

want to stay away from z0 especially for the x-axis

z0 tends to have more rotation. For y's and z's it is not as much of an issue, but for x it can contribute to rotation

[pseudo mortar]

Sounds good.

I'll get new rails and belts soon.

The bearings hardly have any time on them I do have an extra set. They are BOM fushi.

[pseudo mortar]

I needed to print a spool bushing and it looks like it is printing slightly better.

[pseudo mortar]

I have parts on the way. 3 medium preload rails and belts.

[pseudo mortar]

I should have the parts tonight. How are EPDM belts different than the standard belt in terms of how I should tension them?

[frank crow]

I am sorry to inform you that your introductory input shaper help line service has expired. Please feel free to subscribe to our extended help.

[frank crow]

Looks like you are interested in extended support. We offer three tiers. Our suggestion is that most people find the service they need in the silver tier. Thus tier is 1000 a year which can be paid monthly for 100 a month

[pseudo mortar]

I can give you 1000 of this

[frank crow]

Sorry usd, gold, or silver only. By order of management

[pseudo mortar]

Almost back together with new rails and belts. With the new rails it is a lot more ridged and the toolhead has zero wiggle. I cut the belts very precisely to the same length. I spent way too long checking and double checking before I snipped them.

I did find one belt path issue and it was foam I laid on the URBI to stop it from rattling from earlier on was contacting the belts. I trimmed it off and it should be good now.

[pseudo mortar]

AHA!

Much better.

Belts are at 121 hz and I have not done any printing to let anything settle.

[frank crow]

go ahead and set both x and y to EI and print a test cube please

[pseudo mortar]

I tried mzv and got an improved cube. Still some very faint ringing and some VFA only on the back side of the cube which is strange to me. I am going to do some printing with it and get maybe 10 hours on it to get the belts settled and then re tension and try again.

What do you think the double X peaks could be?

[frank crow]

Loose screw. Let me check the vault

[pseudo mortar]

Also, when I tensioned the belts I went back and forth 1/12 turn at a time and then a hair at a time until they were very close to even.

[pseudo mortar]

Almost seeing the same situation on the belts chart.

[frank crow]

it has been so long, what printer is this

[pseudo mortar]

V0

Stock gantry dragon burner and a sherpa micro with an EBB36 bracket. Now with new medium preload rails on X and Y as my x rail had a little play and new EPDM belts.

[frank crow]

The issue is closest to loose screw or belt tension

belts look like belt tension, x looks like loose screw

[pseudo mortar]

Ok thanks I will do some tinkering and I will be back with hopefully good results. I want to print some new panel clips and I'll use this machine. It is printing good enough for that now.

[frank crow]

The name of the game is the prints.
If the prints look good, don't worry about the graphs

[pseudo mortar]

Yep. My 2.4 graphs are what I would call good enough and imperfect but it prints real nice so I haven't touched it in a long time. My trident now prints very very nice and it has some umbilical showing on the charts but I am scared to change anything. This one has a hair of ringing left that I want to figure out.

[pseudo mortar]

Think it could be a fan?

[frank crow]

like a bad fan. No we would see a vertical line on the spectrogram.

If it is a loose fan? I do not know

[pseudo mortar]

Loose fan is what I mean yes.

I am not running any fans during testing.

[frank crow]

the test can stillpick up a broken or about to break fan. This will show up as a faint vertical line in the spectorgram

even if the fan is off

[pseudo mortar]

Ok interesting I'll keep that in mind.

I can't get over how much more ridged and smooth this thing is with the new rails.

[frank crow]

I was talking to Frig at portland vice. he stated that he replaces the rails once a year.

Depending on how much you print that might be a thing to do

[pseudo mortar]

Not that much! I imagine pif machines have thousands of hours a year on them not a couple hundred. All I really print is parts for my printers as I tinker on them and tuning prints to try to get them as good as they can be. The odd part for a project or for somebody else.

These ones had maybe 300 on em. I remember doing something that would hurt them and I can't remember what. I just remember thinking hopefully they survive.

They were youmtong. 0 preload on Y and z1 was supposed to be on X but I could have got it mixed up with a Y rail. I have a youmetong X rail on my 2.4 and all around medium preload on my trident and no issues on either of those with well over 300 hours. They are also are bigger rails so there could be something there. I want to get some preloaded rails on Y for my 2.4 now after seeing how much stiffer it makes the gantries of the other two.

I could have used the wrong grease or something on these I am not sure. I know the first set of formbot rails on the V0 I smoked with a white lithum grease full of PTFE. They were full time sliders by the time I replaced them.

[pseudo mortar]

Trying to get 10 hours on it that's what I'm getting.

[frank crow]

That is vfa. Likely due to motor speed

Do you have slow down for layer time on in the filament section of your printer

[pseudo mortar]

Yeah I never thought about that.

Oh I don't know if I sent this either.

[nimble oracle]

how fast was this printed? the middle parts

in mm/s

[pseudo mortar]

I should check I printed that very sparse so maybe the layer time was cutting the speed down to the rough area.

Yeah lots of it was printed around 100mm/sec or less

[nimble oracle]

print it faster

ideally >=150

[pseudo mortar]

You bet.

[nimble oracle]

before starting the print, analyse the gcode in the slicer to verify its acutally printed in the desired speed

[pseudo mortar]

I set the minimum speed to 200 in cooling I'll see how that goes.

[frank crow]

Generally we just turn off min layer time and auto cooling

[pseudo mortar]

I've always run it on my V0 and I don't have much of a reason other than PLA and PETG suck. I'll learn to tune around a set speed.

[pseudo mortar]

If I was smart I would have turned it off instead of doing this

[pseudo mortar]

Left is 200mm/sec locked

Lots better but still some there.

[nimble oracle]

a V0 is not easy with VFAs

thats unfortunately how it is

[pseudo mortar]

Due to the bearing idlers on the X ends I presume?

[nimble oracle]

everything basically plays into this.

[pseudo mortar]

Smaller diameters too probably.

[pseudo mortar]

How much does belt tension play in? Is there a sweet spot?

[nimble oracle]

i cant tell u how much, but a certain amount

#1259699805693808681 improves some aspects but its not a "solve everything"

without double shear u are limited to 110hz

u can play with motor run_current, see if it changes anything

[pseudo mortar]

Ok thanks I will give those a try.

[pseudo mortar]

Yeah I was thinking about trying that.

Are there any specific motors that can help reduce it too?

[nimble oracle]

not that im aware of

[pseudo mortar]

I am still running the okeada ones from formbot 5 years ago.

[nimble oracle]

hit or miss

[pseudo mortar]

Belts are at 97 and 93 after 9 hours of printing. I will re tension and test again and see what I get for the time being.

[pseudo mortar]

Course I had to tighten the belts one tooth on the cariage to get it over 107hz lol

[pseudo mortar]

Good thing I had to take off the toolhead. I discovered my revo nozzle to be not quite seated and had some wiggle.

[pseudo mortar]

I was going to print some hollow cubes to play with some numbers like motor current to try and reduce the VFA. What does it mean when it is visible on the top surface?

[frank crow]

That can be em, first layer issues translating to surface generally because you have no infill.

Not enough top layers

[pseudo mortar]

Ok I wasn't sure if it was toolhead wierdness or not.

[frank crow]

This is explained in Ellis print guide. Look under em I believe

[pseudo mortar]

I slightly improved the VFA modulating run_current.

It is almost acceptable I'll try a few other things.

[pseudo mortar]

I just made an interesting discovery. The VFA is more prevalent on the right and rear of these cubes I have been printing. I am making some rail alightnment tools to hot tighten the trident and while watching the first layer go down I realized that the right and rear is where the PTFE tube contacts the tophat.

I should print a diagonal one to see if it is a specific drive too I don't know why I haven't tried that yet.

[frank crow]

Lets see what mr diagonal says

[pseudo mortar]

I'll run that test when I get home. I'm having a 5 O'Clock coffee at work before I leave after fixing my truck in the shop. It's fun when the boss is sick and says do whatever and the other one is a days drive away.

[pseudo mortar]

Kept it in ori notation of printing on the bed there. 200mm/sec.

Kinda the same everywhere the back left side is not as bad as the other 3

I can feel the waves with my finger.

[frank crow]

can you rerun vibrations from 0-300 mm/s please

[pseudo mortar]

You betcha

[pseudo mortar]

Enjoy

[frank crow]

can you drop pwm to 22e3 and rerun

[pseudo mortar]

I am entirely unfamiliar with what that means.

More so how to do that.

[frank crow]

add this under each motor section in autotune
pwm_freq_target:22e3

[pseudo mortar]

10 4

[frank crow]

no 22e3 🙂

[pseudo mortar]

But 104 is a better number 😤

[frank crow]

you can try that, but I am not sure what the effect will be

[pseudo mortar]

So my guess is you are targetting the spikes on the motor frequency chart with that number?

[frank crow]

I am attempting to see if we can't calm down the angular speed energy profile spikes about 100

[pseudo mortar]

Oh yeah ok interesting.

Will the trident heatsoaking with fans running beside it mess with the vibrations data?

[frank crow]

not really

[pseudo mortar]

Ok good

[frank crow]

now increase run_current to 0.8 if your motors can handle it

[pseudo mortar]

They can thats what I ran up until I started printing these boxes and found 0.65 to help a little.

[frank crow]

also what do you belts currently look like

[frank crow]

your belts are not even, but I do not think that is the issue

somethiong is going on with the A drive.
@fair rivet what were you thoughts again as to why the A-drive is not being picked up. Or, was it @alpine hazel

@pseudo mortar based upon what I am seeing can you redo the print with an ext/int perimeter speed of 150 please

[pseudo mortar]

Ok diagonal or square

[frank crow]

you can print them at the same time just rotate one 45 degrees. Make sure to mark them so you know which is which

[pseudo mortar]

will do

super slicer is basic CAD I made hollow boxes with text so I don't loose track lol

[frank crow]

you might want to add +x, +y and 45, 135 on the bottom so you know the orientation if you are going to label them

[pseudo mortar]

woah settle down thats too much work I have that screenshot I can go off of for the rest of orientation

[pseudo mortar]

More worser

Locked at 150mm/sec

[frank crow]

can you measure the spacing between the VFA's please

[pseudo mortar]

2mm

Peak to peak

[frank crow]

Belts

Are you using gates parts

[pseudo mortar]

gates EPDM belts and whatever pulleys came in the formbot kit. I have some powge ones but they have the incorrect teeth profile with sharp edges whatever that is called I remember not.

fushi bearings.

[frank crow]

Check your form bot kit. However, you may want to purchase gates parts

Pulleys and idlers

[pseudo mortar]

Yeah thats probably a good plan.

[frank crow]

According to Ellis 2.0 is belts, and since you have gates belts the next thing I would check are pulleys and idlers

[pseudo mortar]

Sounds logical. I was thinking about doing tulip but I'll see. Might be a while before I have an update. Thanks for the help so far I appreciate it.

[frank crow]

Okie dokie artichokie

[pseudo mortar]

🤣

[pseudo mortar]

It prints so dang nice if you take a picture at the right angle it hurts.

[frank crow]

This will be for most anything.

The led lights in printers are extremely harsh. Take it outside to look at

[pseudo mortar]

Yeah I know but the lines are consistent and the top layer I have dialed in pretty well. Just wavy walls. I'll improve it some how.

[frank crow]

Try changing speed

Ever so faint vfa

[pseudo mortar]

Not faint with any other angle.

[frank crow]

lol

This is the nature of the beast

[pseudo mortar]

I can feel the waves. Maybe I should just not look at them at any other angle

[frank crow]

Stop printing straight walls 😉

[pseudo mortar]

It is worse on circles even with super fine STL files

[frank crow]

Have you tried 105 for ext/int perimeters

Back to cleaning

[pseudo mortar]

Not yet.

[pseudo mortar]

What are the odds it's a couple of bad flange bearings? I felt all of them and they were all super smooth they aren't very old fushi bearings that I packed with grease. I think I might have another set I could try.

Nope they are F695s.

Shoot. B drive belt is riding up against the pulley flange. I checked that a few times and somehow never caught it.

[barren grail]

if it's not too agressive it shouldn't affect it like your pics imo, you'd see dust and whatnots

uart_pin: gpio9
tx_pin: gpio8
uart_address: 0
interpolate: True
run_current: 0.976
home_current:0.7
sense_resistor: 0.110
stealthchop_threshold: 0          
diag_pin: ^gpio4 

[autotune_tmc stepper_x]
motor: ldo-35sth52-1504ah
tbl:2
toff:3
sg4_thrs: 110
pwm_freq_target:39e3```

try this for autotune settings, change the sensorless homing etc,

[pseudo mortar]

Yeah it seemed like it was just making contact one direction and coming off the other.

[pseudo mortar]

Thanks I'll try that.

[pseudo mortar]

Behold: @frank crow

I'm about to make a technological advancement.

Once I get a printer working well enough to make the parts that is.

[frank crow]

lol

[nimble oracle]

23 days have passed. Whats the current state here? Do u need further help? if so how can we help?
if not, can we mark this as solved?