Deep Resonance
I have recently built my vcore 3.1 500 and I am having an issue with very pronounced ghosting on the X axis. Deep enough to feel. The Y axis is mostly fine.
Tightening the belts past (what I think) is too tight does not fix it. And at this point loosening the belts does not impact it. I have just done a check and did not find any loose nuts or bolts. I do not believe the belt is rubbing against the printed parts.
I do not yet have an ADXL345.
I need to test lower accelerations. I thought I had limited it to 2000, but looking at the dashboard mid print, it does exceed that limit. It seems putting
max_accel:2000 in my config did not actually limit it.
63 Replies
you can do the shaper calibration without a adxl345
https://www.klipper3d.org/Resonance_Compensation.html
You can, I did and forgot to note that.
Manual resonance compensation did reduce nearly all the ghosting in the Y axis but not X
currently in printer.cfg
this cant be right
this is from my VC500
# Input shaper
[input_shaper]
shaper_type_x = zv
shaper_freq_x = 67.2
shaper_type_y = zv
shaper_freq_y = 38.2
also, your format is wrong
shaper_type: mzv
your frequencies are way too highI could remove them and do another test print
Ok, removing the input shaper lines does not change the results
bc you still dont have the right ones
Let me double check the math from the resonance test
I definitely did the math wrong
Is incorrect. It should be
Thus, I should add this to my config
you dont need that much decimals
2 are enough
yeah, figured. that's just what I copied from the calculator
also, x should be around 65hz normaly
y at arround 35 hz
I could try a manual 65hz
I do plan on getting a adxl345 in the future
just buy one, 2 bucks
buy a pack of 4 for 5 bucks, you will need them
It's the soldering I can't do yet
"yet" because I do plan on it, but dont want learning soldiering to be the blocking element to getting passible prints
believe me, soldering pins is first day solder class, you can learn that in 10 minutes
just try it, if you fail, try again, ...
Hm, good point
But soldering for the adxl345 is like the first or second printer upgrade I plan on doing after it's "done"
Or, upgrades I consider non essential
the adxl is the cheapest possible update, but it has the highest imapct on print quality
just get some
the printer is not complete without one
changing the input shper values to 32/35 and 65/35 did not change the ghosting
of course not, you need the values for your machine
thae values i mentioned are typical values for you as a orientation
the 32/35 values were from the test I did with the corrected math
yes, but at elast the x vcalue cant be right
again, typical value for x is around 65hz
for y 35
Hm, so I was going to say if all 3 gave identical results...
but in trying to take a good picture, it is more clear that the results are not identical.
Thus I should re print the manual compensation test now
Running the resonance compensation print again gives the x and y numbers of 33.95 and 33.86
but the results are not considerably better
edit: tomorrow i will be purchasing a soldering iron and adxl345
compatible-crimson•2y ago
@justdecker any luck?
Running the accelerometer tuning now
Soldering was in fact considerably easier to learn than expected
Results were as followed, running a short test print now
is this the information i was looking for
Playing with belt tightness, I dont know what any of these mean and I cannot divine a correlation
They're just for making the two belts the same tightness by making their graphs match. Use the X and Y resonance tests to determine if they're tight or loose
the ones that look like this? (though these are both pre adjustment, from the same time as the 1st graph posted today)
it looks already good, the green part is ok, thats how it should look, but before you continue you should try to get rid of the red part
before the red part isnt gone it doesnt make a lot of sense to start with the input shaper graphs
ok adjusted back to the starting point, except the colors swapped because I got the order wrong,
how do i actually swash that peak? does it imply the relative tension is off?
the peak at 125hz means that something is vibrating like hell
at this frequnecy it is mostly located at the eva toolhead somewhere or the yxy joiners
the belts are already super equaly tensioned, thats good, you are almost there
just that one resonance its what holding you back from having a great tuned input shaper
joiners?
the XY joiners are the two assemblies at both end of your gantry, its where the X and Y axis are joined, thats why we call them XY joiners
got it, I'll check those for loose parts and then run the test again
also check for forgotten micro shims
it couldn't be the bridging artifacts, could it?
thats not ideal, but i doubt thats the issue here
i bet something on the toolhead
please share a photo your your toolhead setup
Yeah, I didn't find any "loose" nuts but a few that had room to be tightened
I can also zoom in or focus on specific sections or angles if you'd prefer
first suggestion is to finish your cable management first
make sure the cables cant move
so that is one thing I was wondering, should I expect the adxl345 to be always on the toolhead? or just for the calibration
this isnt important, thats up to you, but with cable management i mean all the other cables
probe, hotend, thermistor, .....
Alright, that is one aspect I"ve been struggling with
the toolhead has multiple zip tie pockets, use them
oh is that what those smaller holes are for
you have them on both sides
I do not know how to get these wires any tighter
something is vibrating, try to track it down by only measuring the affected frequency
try
TEST_RESONANCES FREQ_START=100 FREQ_END=150
its faster and you can probably hear from where its comingit's the 4028 fan
screws not tight or what?
running multiple 4028s without that issue
thats my assumption but I cant get the screws tight enough to stop the sound. but the sound does stop when I totally remove it and hold the fan alone (except the wire)
Yeah if I hold the fan with wires slack the whole time
if the fan can move then a screw is not tight
this is with all the screws removed
Or, the fan is as removed as it can be without removing it from the wire loom for this test
test if your fan shourd is tightned
the big one in the back from the toolhead, where the 4028 sits on
The shroud is on as tight as I can make it
the buzzing appears as soon as the fan is connected with more than 1 screw
I dont think this is fixable with current equipment. I either live with the resonance and maybe a new version of the fan shroud will fix it, or I remove the fan entirely
e: unless I can print a spacer out of TPU?
the design if the shroud is totally ok
also tpu spacers wont fix anything
something is vibrating on the printer, i know its difficult to find, but its possible
it takes some time
I believe that the design is fine, but because the sound stops when I isolate just the fan makes me think it is something about either the fan or the interface of the fan to the shroud
It could be my specific shroud
wait
why now
edit: this is the fan attached, running the test as normal
you removed the fan anbd reattahced it, right?
then you have probably tightend the screws better then before
I must have. And I did try pulling the nuts a little closer. I do still hear the buzz at those frequencies but I guess that's not the actual problem, just a potential symptom
But also: thank you for putting up with me being wrong and eating my words all day