Bed compensation is not working

Hello! I have noticed the after I start the print job, bed compensation is not activated. I am using a S3D slicer, I have copied and pasted start and end scripts from ratrig page. Is this something missing in start script o something is wrong in the printer's config files?
25 Replies
armearme.
armearme.OP•2mo ago
@miklschmidt maybe it has something to do with the fact that I am using contact beacon mode? 😄
miklschmidt
miklschmidt•2mo ago
bed compensation is not activated
How do you figure?
armearme.
armearme.OP•2mo ago
Z axis screws does not move while printing a layer. But I actually have another more important question at the time: communication timeout during homing - is this something physical with the cable itself? or might be something else?
miklschmidt
miklschmidt•2mo ago
Put a piece of tape on them, and make a little flag, they move 🙂
communication timeout during homing - is this something physical with the cable itself? or might be something else?
Depends on the actual error, timeout is usually a host overload / resource issue.
armearme.
armearme.OP•2mo ago
what do you mean "actual error". Where can I see it?
miklschmidt
miklschmidt•2mo ago
If you couldn't see it how did you know you got a communication timeout? 🙂 Just screenshot the error, or upload a debug zip after it's happened.
armearme.
armearme.OP•2mo ago
here it is
armearme.
armearme.OP•2mo ago
and as for the bed mesh, something is clearly not working, first layer is far from even. Some areas are too close, other are not.
miklschmidt
miklschmidt•2mo ago
If everything is updated (especially the beacon module), this is most likely a wiring issue yeah, or you have a loose connection in your USB ports etc. I'd need to see the results of BEACON_RATOS_CALIBRATE sounds like axis twist
variable_beacon_contact_z_homing: True # Make all G28 calls use contact instead of proximity scan
variable_beacon_contact_bed_mesh: True # bed mesh with contact method
variable_beacon_contact_bed_mesh_samples: 2 # probe samples for contact bed mesh
variable_beacon_contact_z_tilt_adjust: True # z-tilt adjust with contact method
variable_beacon_contact_z_tilt_adjust_samples: 2 # probe samples for contact z-tilt adjust
variable_beacon_contact_z_homing: True # Make all G28 calls use contact instead of proximity scan
variable_beacon_contact_bed_mesh: True # bed mesh with contact method
variable_beacon_contact_bed_mesh_samples: 2 # probe samples for contact bed mesh
variable_beacon_contact_z_tilt_adjust: True # z-tilt adjust with contact method
variable_beacon_contact_z_tilt_adjust_samples: 2 # probe samples for contact z-tilt adjust
Are you using a glass bed? you're using a 6x6 mesh... that's not gonna do much, if you have lots of inconsistent bad spots.
armearme.
armearme.OP•2mo ago
Yes, glass bed, 6x6 mesh. For example, I have printed a small part now (picture of mesh is below) and first layer looked exactly like this small mesh - on the left side nozzle was too far away, on the right side it was good/maybe a little too close. So basically it does not compensate anything.
No description
armearme.
armearme.OP•2mo ago
How do I send you all the results? debug.zip? I have made a beacon ratos calibrate, there is bunch of info
armearme.
armearme.OP•2mo ago
Okay I can not do a beacon calibration
No description
armearme.
armearme.OP•2mo ago
Okay I have this info:
No description
miklschmidt
miklschmidt•2mo ago
Tons of mechanical issues You have axis latency (there's is noise in your setup, wiring issues), you have >120 microns of axis twist and you're not getting consistent probing results.
armearme.
armearme.OP•2mo ago
Are there any tutorials on how to fix these issues? 🙂 Also, I will try to run the same calibration without glass and not with contact beacon mode, maybe something will be different
miklschmidt
miklschmidt•2mo ago
There's no easy way, start inspecting. You can run the individual parts of the beacon calibrations one by one. The poke test will check electrical noise. The gantry twist measuremennts will show you how parallel your gantry is in Z (or how poorly your glass bed fits on the metal beneath it). The inconsistent probing is because something in your z axis is moving when it's not supposed to, so something isn't rigid or something is loose. Start by finding and solving the inconsistent probing it can be everything from hotend wobble to loose z couplers.
armearme.
armearme.OP•4w ago
so the "Beacon_ratos_calibrate" is a series of tests that show how well or bad the printer is assembled?
armearme.
armearme.OP•4w ago
What about now?
No description
miklschmidt
miklschmidt•4w ago
That's sort of the side effect of it yeah, it can be used to analyze a whole bunch of things, so we do 🙂 significantly better Still some twist (around 80 microns in range now, so much better), but this could just be you flex plate, a bed mesh would help diagnose further It's possible your umbilical is tugging on your carriage at the front (say, if it's not quite long enough)
armearme.
armearme.OP•4w ago
The only difference between these two "beacon_ratos_calibrate" is that first one is done with glass bed and contact mode, and the second one without glass bed, on the aluminum plate and with scan mode 😄 Yeah, I think that's the case, I will try to give it more length
miklschmidt
miklschmidt•4w ago
Makes sense, i'm guessing the glass is flexing too much because it doesn't conform to the plate properly.
armearme.
armearme.OP•2w ago
Hello. I don't really see this issue as a wiring issue. USB seems to be fixed pretty solid, nothing moves. Unless you have in mind any other wires that can cause it.
armearme.
armearme.OP•2w ago
No description
miklschmidt
miklschmidt•2w ago
Yeah all wires can cause this (not the USB cable), if it shorts anything etc and makes the MCU hang due to protections or actual damage, you get timeouts. It can also be significant noise and bad impedance matching on the USB cable you're using, but that wouldn't be my first guess as there's very little data being moved between the toolboard and the pi, so less than ideal correction doesn't have much of an impact. I would go over your toolhead wiring and disconnect stuff that doesn't prevent it from running and see if that solves the issue. Start with the 4028 for example (if you have any of those wires connected to the toolboard, the default is to connect the PWM wire to the toolboard).

Did you find this page helpful?