vc400 CoreXY Belt Tension
Currently trying to get my belts even
92 Replies
Okay current result looks pretty good
What changed between the two graphs?
I tightend the lower belt a tad, I also reduced the top belt length by 1 tooth based on CrashTestCharlies advice
that was prior to these two graphs
they are pretty close to the same tension but for some reason you are not seeing a bigger first spike.
what size printer?
400
1e6/2.5, overall maybe a bit loose?
maybe but if you get to tight the numbers come down as well
home the printer so the toolhead is centered and then go into real time analysis and pluck the belts on each side of the toolhead as close to the middle as you can
I pinch it between my thumb and index finger and pull the belt out a centermeter or two
you should get two spikes on the real time graph, we are looking roughly what hz range the first spike is
https://discord.com/channels/582187371529764864/1223286652949106748/1272625387871928494
How do I enter real time mode?
you know how to get into the configurator on the left hand side?
Yeah
from inside that menu you can go into real time analaysis
I see it now
then just start the analsys. I am on my mobile and my printer is off otherwise I would graph you a screenshot
How am I supposed to see the result :pepehands: the graph reverts so fast
Be faster 🙂
I had a laptop so it was easy.
Do you have a way to pluck and be near a computer
My computer is about 5 feet away lol
thats how I did it. Do you have a laptop or a tablet you could get closer to the printer?
I have not tried it on a phone before. It may work but be hard to see
at this point we are just trying to see if you are in the range of correct tension. ~100hz if the print head is centered and you are plucking between the toolhead and the y jointer plates on a 400mm machine. It would give us an idea if you are to tight or to loose
Yeah I can try on my laptop
you can try one of the audio tuning apps on your phone but I never had any luck with those. This is reading it right off the ADXL so should be way more accurate
This is alternating between top and bottom, top plucked first each time
its hard to tell when you alternate. Can you do one and then the other?
sure
it looks like one is 55 and one 65 roughly?
Top
so about 55hz
and it should be at 67?
according to that spread sheet?
I went with 9lbs so measuing between the toolhead and y jointer on a 400mm machine that is 100hz
Bot
I do the plucking test while tightening them and get them equal. As you can see in my video I drag the cursor over so I have a vertical line on where I want to tune to and then pluck and turn at the same time.
once they are equal and where I want them I tighten up the set screw on the tensionor block and then rerun belt tension
I end up with my belts like 5hz different but they matched on the belt tension graph so I called it a win 🙂
this plucking is to just get you in the ballpark quickly to where you need to be and then you can fine tune. As you pluck and adjust the tension you will see a 1/8 turn is a pretty big change. You will also learn quickly how adjusting one belt adjusts the other one as well. Not a 1 to 1 but its a game up tighten up one and loosen the other. Just play with it and get them both close to a 100 and try your graph again. I suspect they were to loose to get the first spike much as that is the y motion of the entire gantry that the belt tension graph picks up. the 2nd spike is the x movement and since x is lighter it shows up as a higher frequency
Okay that is helpful
So I should try to get them both to 100hz?
I got them both to around 71hz just now
Try to get close to 100hz
This is with both around 100hz
Solution
How does this look, dropped both to around 92hz
Other than some loose wires in the 125 range, nailed it
The part cooling wire is probably the problem, I am not a fan of the wiring for it
Yeah, I rotated so the wires come out the right side, but it's still awkward
Looking good. Tie strap up everything tight and you should be good to go
@chicken So I haven't touched the belts since last night and ran the test one more time before installing the hybrid belts and the graphs are off again 😦
the stretch a little I suspect. I have not retested mine since I started printing and they feel looser by just touching them. You could try to tighten it just a hair or leave it and print with it. They are still close and clean for decent printing.
what does the shaper graphs look like before you install the hybrid. You want those clean most importantly
Okay that makes sense. Haven't run input shaper yet, I will check right now
14600 not bad!
Y is much lower
ZV at 5100
I took the top off and got way better results for the tension, the umbilical was vibrating against it
yah it can do that. I dont know what the range is for mzv on the y of a 400 but I think you are in the ballpark
I had to stop when I realized my cable wire from the toolhead wasn't staying in place, and I was getting wacky inconsistencies. Then I ran out of zip ties.
my 500 was around 3300 if I remember correctly
the recommendation is to use MZV even if it recommends ZV
I got 11000\3500 with a Jane umbilical
Jank
I was like what did Jane do to you 🙂
its clean though which is the most important thing. The accel only is set for the outside perimeter. You can crank it up for everything else
What even is max for something like infill?
I am just using prusa defaults. I think its like 10k for perimeters and 15k for infill. I am sure those are conservative though
Weren't we supposed to have the official profiles by now?
Okay sweet, on Orca slicer there are Official profiles
So based on the input shaper results I should be good to start on the hybrid belts?
There are profiles in orca and prusa but they are conservative it sounds like
I think you are good to go. Just don’t touch the coreXY belts going forward as you cannot really tune them with the hybrid on
Is this okay or do they need to be identical?
It’s more important to have both sides the same length but you might want that top one in a little further so it does not stick out as far as
okay I think I took the wrong approach then, Ill start with putting the end in first and working backwards
By both sides, I mean the left and right hybrid belts. Like the coreXY you want them the same length. Off to bed. I’ll be around tomorrow if you need any help.
@chicken around what hz should I shoot for with the y belts?
Graph with hybrid belts around 93hz
I think she is ready boys
:chefkiss:
hey sorry ended up asleep by the time you posted. The belt tension graph you can ignore once hybrid is on. Its all about the shaper graphs and yours look good. Hows the print quality 🙂
make sure to remember to put the mzv values in the printer.cfg and your y max accel as your exterior perimeter accel
No worries! Looks like I need to tune the slicer profile a bit
Cooling profile
FIFY
Fan sounds above 60%.
Yeah it's at 100 lol
Should 60 be the max?
That's all tuning, but that fan is a monster. I'm going to cap mine at 60% until results say I need more. It's also funny how quiet it gets <75%
Are you still using the blob prime?
It cakes the nozzle every time
I did my first layer test with shittiest pla I had, no problem with the blob, but gamma master has a coating
Thinking I may need to slow printer down to tune. Too many variables printing at the default speeds
Then you get to switch on performance mode.
Keep getting layer shift
Those corners are ugly too. Have you done the orca calibrations? Start with temp tower, pressure advance, then max vol flow, then babystep a first layer?
I did a temp tower and PA but need to dial in the flow
whats the other direction look like. It looks like layer shift but curious what the other side looks like. Layer shift is most likely the gear on the stepper loose. Can you double check one of the set screws is tight against the flat spot and the other is tight as well.
Turned speeds way down and was able to get a finished benchy
I noticed on the back left stepper the belt rides up and down the gear
There might be a misplaced shim
Not a benchy. That's a bell with all that ringing
Unfortunately that is normal to have some movement as long as everything is lined up. It should be so small that it will not affect print quality. The vast majority of open belts are spiral cut out of a large wide closed belt. Even high quality belts like Gates are done this way. Because of this the belts have a natural biases and will move up and down a little. If you look at a long section of the writing on the belts you can even see the writing slightly skewed.
Okay good to know! I won't stress about it then.
I also fixed my ringing issue, I was a ding dong and put the max accel values in the config instead of the frequency. Printing like a champ now
Awesome. Nice work