Spike in y shaper at 80hz to 100+ hz
Been trying to work this shit out for weeks on my open V-core 3.1 400
1. belts are not riding up or down on any of the idlers
2. using FMMM with BRS lever m1
3. aligned my y-axis linear rails with provided alignment tool while loosening all bolts on rails and on xy joiner carriage
4. Custom vz printhead with short goliath hotend
5. Tensioned to 150hz at 150mm from front idler
6. Let me know if you need any other information
47 Replies
3.5 1e6 suggests your belts are tight as fuck or you changed
accel_per_hz config value or adxl is shot
run those 2 commands with the printer idle to check adxl:
ACCELEROMETER_QUERY
MEASURE_AXES_NOISEmy accel_per_hz is 75
ACCELEROMETER_QUERY - accelerometer values (x, y, z): 1184.329507, 1184.329507, -75996.673402
Axes noise for xy-axis accelerometer: 578.485700 (x), 560.394648 (y), 511.786969 (z)
accelerometer is lis2dw
there's your problem :-)
the accel_per_hz?
I have seen other recommend a value of 100
your Z acceleration is 75996.673402, unless you live on Jupiter it should be around 9800 š
OMG
i didn't realise that
š¤¦š»āāļø
faulty accelerometer?
some people also tested that lis2dw in general have issues with IS graphs, something about sampling rate
definitely
hmm let me switch back to my usb adxl345 for a quick sanity check
USB ADXL on nozzle
accelerometer values (x, y, z): -148.041188, -74.020594, 9269.755526
Axes noise for xy-axis accelerometer: 857.439630 (x), 613.311064 (y), 529.914581 (z)
the noise values are still a tad high
yeah, but much better otherwise. do you have fan running somewhere?
disabled on my toolhead, only fans on my pi and mcu at the back
you can try disabling them to see if they are the reason for high noise, but i'd otherwise send it, to see the difference. as a side note, the way you route the usb cable can make big difference in cleanliness of the reading
the fans are not spinning if the motors are not active though. pi fan only spins up when above 40 deg which it doesn't at idle
@Prav have changed to adxl345
These are my new readings
accelerometer values (x, y, z): 0.000000, 296.082377, 10955.165621
Axes noise for xy-axis accelerometer: 14.298753 (x), 18.073656 (y), 28.270914 (z)
y is till the same
Detaching the wire loom from vz-printhead holder
80hz is almost gone
much better, amplitude in check, still some funny business at 100hz, but probably usable.
if you were to keep those results, for y axis set the mzv at 45hz instead of 28hz from the graph.
Well I have to figure out how to reattach my wire loom before being able to start printing again
reattached the wireloom securely on the the printhead... all hell broke loose
by moving the gantry rapidly back and forth manually a few times, the gantry seems to seize up. I am still trying to find the point of binding, have tried aligning y-linear rails multiple times.
Loosen the y rails and then move the x gantry back and forth a few times by hand to get the alignment correct and then retighten the y rails
Already did that, I have a feeling the binding is coming from my BRS m1 mod
Hmmm maybe. I don't have that mod so I'm not sure how to help with that š
I am beginning to figure out what's wrong
I am able to seize up my carriage repeatedly by rapidly moving my toolhead up and down in the diagonal (45deg) direction. Initially thought that it was due to the belt binding somewhere but realized I am still able to move the toolhead in the x direction but not the y direction.
I suspect that it's the bearings in mgn12 blocks on the y linear rails seizing up/jamming. Gonna try lubing it with white lithium grease without dismantling. If it helps but still causes jamming randomly, I will have to fully dismantle and clean the bearings in the block. Suspect it may be dog hairs getting in the block haha
At this point, I really just want to give up... Cleaned the rails, took out the bearings, cleaned up. Lubricated it with white lithium grease. End up still seizing up when I move y axis rapidly
Toolhead able to move left and right
Does it move smoothly without belts?
I have not tried removing the belt, will do that next
I am so tired from all the experimentation today
Check to see if it moves smoothly without the belts. That will narrow the issue down to the belt system
And it seems to be always the same spot
Disconnected belts from the back, loosened front, still seize/jam when I move it rapidly
Detached gantry, both carriage seem to roll fine
The problem is with the y axis rails or carriages then I'm afraid. The carriages should move smoothly with or without a load. Since you have disconnected the belts there is no tension from the belt system binding the carriage so the issue must be the rails imo. A few options I can think of.
1. The carriages might be missing some ball bearings.
2. The rails are damaged in the area where the seize occurs.
2. The x gantry is not aligned when attached. When you have the x gantry on the rails and you bring the extrusion all the way to the front do both ends touch the front y stop printed part?
1. The right carriage is missing 1 ball bearing and each side of the carriage
2. I am leaning towards this, it's ocurring at sinilar spots
3. Yes, aligned to the front, xy plate is squared
I think it's normal for the carriage to have a free spot on each side but not 100% sure.
Looking back at your previous message it could also be the grease you are using seizing up the ball bearings. I use sewing machine lubricant personally with the RR rails. I would try that first before buying completely new rails
Before changing to lithium grease, I was using sewing machine oil, same issue hence ripping everything apart and cleaned it
Ahhh ok, damn that sucks I was hoping it would be a simpler fix.
I can't really think of anything else other than completely changing out the rails and carriages for the y
Me too š
I just did one final check. took out the right linear rail because that seems to bind up more. Place it on a table and move the carriage with my hand, unfortunately at the spots that always bind up with the gantry, it feels rough and occasionally causes the carriage to slip off my finger because it got momentarily stuck
Sounds like that rail is causing all your problems. I'd suggest replacing both rails and carriages so both rails are the same. You might be able to get away with replacing just the one rail and carriage though
I will just replace both at this point, sigh...
I'm glad you were able to find the issue at least. I was hoping it wasn't the rails but it is what it is
Hopefully that's the cause of the bad IS. If not, back to square one.
if Y rails move smoothly without X beam attached, the X might be bowed, and in result trying to twist the Y carriages, so they bind in result, or you might have Y rails attached like that:
so good news, swapped out my y-linear rails for new ones, not perfect but a lot better
But facing issue now on X graph
I have isolated the problem, the peak at 75hz seems to originate from the red span of belt resonating while the 85hz peak seems to occur when the blue span resonate. @Prav I have tensioned to 100hz according to your guide on my v-core 400
belt resonance graph shows two peaks as well, though only upper showed the peak at around 85hz
