My printer is now printing slanted models. I have run various speeds all the way down to 25mm/s. and the results are almost identical. I have changed out stepper motors and drivers. The system was working with a Manta M8P V1.0 board but I was having some problems with the temp sensor on the EBB42 board so I upgraded everything to take advantage of an easier hook up of the CAN bus. Any suggestions of what I should be looking at?
I would agree if the printer hadn’t been printing square prior to the board change. I do wonder if it has anything to do with the highspeed of the board processor and the pulse widths for the steps it creates. I 'm just grasping at straws with this. It’s almost like the direction change pulse is coming too soon or too late and the driver isn’t fast enough. I have tried with and without hold current.
This is where it gets interesting. If I change the microsteps the fewer the number of micro steps the greater the lean. The cube on the left is 256 micro steps and the one on the right is 32.
It was pointed out to me that I should have interpolate: False in the stepper config. I had it previously but when I copied the new example for all the pin changes I didn’t notice this was missing. It’s always the simple things
Actually it’s not perfect but it’s better. I have some TMC5160 drivers coming so I’ll work with what I have for now and then see what happens with them. Some others have suggested the TMC2130 drivers were a problem.
Could you describe the mechanics of your printer (ideally with photographs)?
If it’s always slanted on the X axis (even after swapping out the steppers and drivers) I would first want to be convinced that the problem wasn’t with something mechanical before looking at the drivers/software.
The last bed mapping showed a total variance of about 0.5mm. Each print runs a probe cycle over the area to be covered by the print. The control and steppers run at 24V.
When you say “The X axis at this time has a single stepper driving a belt to the three ball screws” I presume you mean the z axis?
Looking through your klippy.log, three comments.
You’re only setting one of the steppers to your homing current and not both during sensorless homing. As both steppers are running when homing an axis, the current on both must be reduced.
Your run current is quite high. I run at 0.8A for my 300x300 Voron 2.4 and my custom CoreXY printers. I found running at higher currents really heated up the steppers and lead to layer shifts. You might want to try reducing them.
You have a lot of macros. When somebody is having operational problems with their printers, one of the first things I suggest is to remove all the macros as generally people pull in macros from various sources and aren’t sure of their provenance or how they actually work. Doing a quick scan of your macros, it looks like you need many of them for basic operations but I wonder what would happen if you took out the M204 and M205 macros.
Finally, what is your host system and have you made any changes there?
Yes I meant Z instead of X. I fixed the homing, thanks for that. I find my steppers run cool I wonder if it’s because they are metal mounted to the metal frame = one giant heatsink. I’ll take a look at all the macros I don’t use many of them they were just there.
This week I will receive the tmc5160 drivers I ordered a few weeks ago and then I’ll start all over. I’ll look at running at a lower amperage for the steppers.
The host system would be Klipper, Mainsail on the CB1. When I changed boards I didn’t reflash my CB1. I thought that shouldn’t change. I’m not familiar with eMMC so I didn’t go there and from the manual I thought that was only for the CM4 Raspberry Pi.
Was the config basically the same with the old Manta board?
Though not related to this issue the feeder current seems way too low in my opinion and then it is running in StealthChop mode.
Did you perform the tests without the hold currents already?
Your acceleration seems quite moderate but as X and Y steppers are running in StealthChop as well this can be an issue.
Can you check with SpreadCycle again?
And maybe print a bigger cylinder in vase mode/spiralized to exclude any travel moves.
Sometimes a 3D printer is running just a bit below the limits all the time and one might think that all is fine.
And then just a slight change of a parameter makes the issue visible.
Otherwise this is a really nice and solid printer!
And your print shown above looks really nice!
The pictures you posted seem to indicate that there are lost “step pulses” between the micro-controller and the tmc driver. The log is also reporting messages like the following Stepper stepper_y phase change (was 387 now 203), which is also indicative of lost “step pulses” between mcu and stepper driver.
Alas, it is not clear what the root cause of lost “step pulses” is in your setup. In the past we’ve seen this with boards that have “level shifters” on the dir/step lines that are too slow to handle the necessary high speeds. It could also result if the dir/step lines have too high electrical capacitance.
It’s possible you could force Klipper to use a slower pulse rate by adding step_pulse_duration: 0.000001 to your [stepper_x] and [stepper_y] config sections. However, doing this may limit the maximum speeds you can move the steppers (that is, if you try to move faster than the now limited step pulse rate then Klipper will report Stepper too far in the past errors).
