Printer Model: Ender 5 Pro
MCU / Printerboard: SKR Mini E3 V3
Host / SBC: RPI 5
klippy.log
I recently upgraded my Ender 5 to the Endorphin setup, which is a hybrid-corexy kinematics setup. I haven’t been able to home it successfully.
X homes fine and properly. Y goes in a diagonal direction at 45° moving away from the end-stop in the positive y direction, or at 45° moving away from the end-stop in the negative y direction (if that makes sense) depending on which way I have the dir_pin or !dir_pin set. Its my understanding the X stepper is supposed to move in tandem to the Y stepper to foster an actual true Y movement.
I’ve not had a klipper issue I couldn’t resolve in the past - where the documentation is fantastic. I was reluctant to post, but I’ve reached the end of troubleshooting possibilities. I’ve found some slightly similar issues on reddit as well as this discourse forum, they were related to messing around with “+”'s and “-”'s in the ~/klipper/klippy/kinematics/hybrid-corexy.py file. I haven’t had luck with that yet but may not be trying all possibilities out with possible modifications. I wanted to ask about this before making too many adjustments.
Initially I thought maybe something was off with the motor, or it wasn’t supplying enough current. That’s seemingly not the case, the STEPPER_BUZZ x and y commands make the steppers behave in the proper direction, which from my understanding is 10 1mm oscillations away from the endstop location.
Looking for any suggestions to use this new kinematics. Thank you
Attached what I believe is a reset klippy log just before home command (and subsequent shutdown) as well as a picture of the endstop locations. Video of the failing homing process. Forgive the wiring, I just want to get a successful home before cleaning it all up.
I’m not sure that is true for this setup. Looking at the pic below taken from the Endorphin site, it appears X is controlled by one motor and Y the other.
Id love it if it worked that way, but if I just move the Y stepper manually, I get the same behavior, the X moves as well. There is something thats fundamentally not working for me, Im hoping someone has special knowledge of this. Heck if I could get a hold of the original project creator, Id pay him. I can find no way for additional support. I’ve spent lots of time and real world money to purchase and print petg parts for his project, only to be disappointed following all the directions to a T.
If nothing else, is there a guide to turn what I have now into a true corexy setup? Im willing to buy additional rails and motors if necessary. I’m committed at this point and want to make the most out of this upgraded Ender 5 pro.
I can’t make certain sense of it, but looks like you define the carriages, then define the steppers with which carriage they move. So stepper x would be just carriage x and stepper y would be some combination of carriage x and carriage y.
But I have no experience with this, so what I say here should be taken with a grain of salt. I just thought this was an interesting problem and am trying to wrap my head around it…
Otherwise, maybe this is your solution for going full corexy…
If I manually move Y, the X will move as well - the same goes if I issue a home command. The X stepper is turning as well. There is something I’m missing where the motors work together properly for Y movements. Ill have to review your posts to see if changing the kinematic will help here. I’ve reached out to several people on Reddit that have had success with this Endorphin mod, but have yet to hear back. At this point, with what I’ve put forth into time and parts I’d pay someone to assist.
If Y motor is stationary, rotating X motor clockwise will move point A closer to the motor which will drag point C closer to the motor which will drag the carriage which will drag the point D which will drag the point B and take up the slack in the belt.
So if Y motor is not moving, carriage moves only in X
If Y motor pulls gantry towards itself and X is not moving. The point A and B cannot move without X motor turning or one end would develop slack and the other would break. At the same time the gantry moves the kink in X motor belt between points A and C closer to A, so since A is not going anywhere the distance A to kink grew smaller therefore distance from kink to C must grow larger.
So if Y motor is moving and X motor is stationary the carriage will move in X and Y.
To move the carriage in Y only both motors must spin at exact rate and in the correct direction. If I’m looking at this correctly, the X motor has to maintain the same distance between point A and the kink for the carriage to stay fixed in X and it does that by turning clockwise (moving point A towards the motor)
Thank you bozzo for the run down on how its supposed to work. I feel I have a good understanding of that. It’s a sound reference for this post.
Many people in other posts suggested putting the X stepper belt clips in the front, but mathematically that doesn’t seem like it would make any difference.
It does. I decided to forgo the aesthetics I was looking for and moved the belt clips to the front. It just works as intended now. Slapping my head here, but perhaps someone brilliant will correct my mathematical thinking error.
Additionally, if I look at the picture 2 responders have posted (the one from the endorphin website I’ve looked at dozens of times) it shows the clips in the front. Re-reading my OP, I see now that I didn’t follow the instructions to a T and stand corrected. I had one job to do.
Belt clips at the back, like, C side is short and D side wraps around?
that reverses the action of the X motor needed to counteract the Y movement.
Clips at the front, pulling C towards the X motor moves carriage away from the X motor.
Clips at the back, pulling C towards the X motor moves carriage towards the X motor.
I _think you could do it your way, and live with a mirrored X axis (need to move the endstop position to 0), so all prints would come out wrong-handed needing a mirror operation in the slicer.
You’re possibly right, though I tried every dir pin possibility and the Y still moved in a 45degree motion every time instead of straight for the homing process. At this point, I’m leaving it alone so I get to testing out this new hardware finally! Thank you.
