1. ZoneStar Z9V5Pro
    1. Printer internals overview
  2. Klipperize Z9V5Pro-MK4 (with non mixing E4 hotend)
    1. Found references
    2. Klipper Flashing procedure
    3. Changes from forked
    4. Heaters PID calibrations
    5. Bed Mesh
    6. Extrusion speed
    7. Pressure Advance (PA) calibrations
    8. Input Shaper

ZoneStar Z9V5Pro

photo

  • Core XY
  • 4 to 1 mixing/non-mixing hot-ends
  • 300x300x400 mm (actual bed 310x310mm)
  • STM32 based board

Printer internals overview

The covers are hold by 4 screws each:

  • Top contains stepper motors of extruders and connectors to printing head:
    Top of the printer
  • Bottom contains the main board:
    Bottom of the printer Board Image

The board is ZM3E4 V2.1:

  • Schematics and parts placement were found in Zonestar3D GitHub
  • In schematics, it is STM32F103VCT6, but on board, it is clone APM32 F103VET6 by Geehy

Klipperize Z9V5Pro-MK4 (with non mixing E4 hotend)

Found references

Klipper Flashing procedure

  1. Obtain Klipper by clonning it: klipper3D
  2. Configure by going to cloned folder in terminal and run 
     make menuconfig

    in menu choose:

    • MCU Architecture - STM32
    • Processor model - STM32F103
    • Bootloader offset - 20KiB bootloader
    • Communication interface (USB(on PA11/PA12)) and press Q and Y to save and exit
  3. Build FW file by running: 
     make

    it should complete without errors (otherwise install dependencies till it works). You will get klipper.bin in ./out/ folder.

  4. Prepairing uSD-Card:
    1. Copy the klipper.bin to root of uSD-Card and rename it to firmware.bin
    2. remove old_fw.bin if exists
    3. Unmount (eject) uSD-Card to cause fsync
  5. Flashing printer:
    1. Power off power supply (PSU) of the printer
    2. Insert uSD-Card
    3. Power on PSU
    4. Press and hold for few sec small power button on bottom main board cover (the Zonestar logo might blink few times but will stay dark after it).

In my case, the printer wasn’t found and power cycle didn’t help. Looking on Linux logs using journalctl -f showed that it had USB detection errors and it was solved by eliminating USB-hub and plugging-in printer directly into PC: Errors via hub:

Nov 02 02:44:20 pc-name kernel: usb 1-10.4: new low-speed USB device number 71 using xhci_hcd
Nov 02 02:44:20 pc-name kernel: usb 1-10.4: device descriptor read/64, error -32
Nov 02 02:44:20 pc-name kernel: usb 1-10.4: device descriptor read/64, error -32
Nov 02 02:44:21 pc-name kernel: usb 1-10.4: new low-speed USB device number 72 using xhci_hcd
Nov 02 02:44:21 pc-name kernel: usb 1-10.4: device descriptor read/64, error -32
Nov 02 02:44:21 pc-name kernel: usb 1-10.4: device descriptor read/64, error -32
Nov 02 02:44:21 pc-name kernel: usb 1-10-port4: attempt power cycle

Direct connection:

Nov 02 02:45:18 pc-name kernel: usb 1-8: new full-speed USB device number 77 using xhci_hcd
Nov 02 02:45:18 pc-name kernel: usb 1-8: New USB device found, idVendor=1d50, idProduct=614e, bcdDevice= 1.00
Nov 02 02:45:18 pc-name kernel: usb 1-8: New USB device strings: Mfr=1, Product=2, SerialNumber=3
Nov 02 02:45:18 pc-name kernel: usb 1-8: Product: stm32f103xe
Nov 02 02:45:18 pc-name kernel: usb 1-8: Manufacturer: Klipper
Nov 02 02:45:18 pc-name kernel: usb 1-8: SerialNumber: 28001100170000464C59504E
Nov 02 02:45:18 pc-name mtp-probe[732041]: checking bus 1, device 77: "/sys/devices/pci0000:00/0000:00:14.0/usb1/1-8"
Nov 02 02:45:18 pc-name kernel: cdc_acm 1-8:1.0: ttyACM0: USB ACM device
Nov 02 02:45:18 pc-name mtp-probe[732041]: bus: 1, device: 77 was not an MTP device
Nov 02 02:45:18 pc-name snapd[2187845]: hotplug.go:200: hotplug device add event ignored, enable experimental.hotplug
Nov 02 02:45:18 pc-name mtp-probe[732048]: checking bus 1, device 77: "/sys/devices/pci0000:00/0000:00:14.0/usb1/1-8"
Nov 02 02:45:18 pc-name mtp-probe[732048]: bus: 1, device: 77 was not an MTP device

Changes from forked

  1. Z-sensor wasn’t mapped so added: 
     [probe]
 pin: !PB13
 ...
  2. Fixed bed dimensions
    • due to wiring Y axis getting stuck into motor connector and so Y movement limited by 300mm (bed is 310mm)
    • screw locations also fixed and added delta of z-sensor position (for some reason it doesn’t get compensated automatically)
  3. Calibrated rotation_distance for all 4 extruders

Heaters PID calibrations

To calibrate both PID loops of heaters used:

    PID_CALIBRATE HEATER=extruder TARGET=200
    SAVE_CONFIG
    restart
    PID_CALIBRATE HEATER=heater_bed TARGET=60
    SAVE_CONFIG
    restart

results:
graphs of temperatures

Bed Mesh

Created initial bed mesh before fixing anything using GUI HEIGHTMAP clicking CALIBRATE and then running SAVE_CONFIG in console result:
initial bed mesh

After that used

SCREWS_TILT_CALCULATE

to adjust screws and bring all corner to the same plane, and running mesh again:
final result

Interestingly, I added plastic rectangle below PEI flexible top and the mesh didn’t change, then added metallic ruler and the sensor was triggered even before going down. It means, that the sensor measures metal plate/magnetic sticker and not real bed position. Hopefully, switching to steel PEI plate will solve the issue. Another option is to switch to BLTouch sensor that uses mechanical contact instead of inductive proximity sensor.

Extrusion speed

Repeating experiment from this video:
https://www.youtube.com/watch?v=0xRtypDjNvI
extrusion of 300mm at various speeds

F [mm/min] E300mm [gr]
100 0.82
200 0.81
250 0.79
300 0.74
300 0.71
400 0.63
500 0.5

Graph of weight vs speed

So it seems that printing above F200 starts extrude less material.

The 200mm/min of 1.75mm filament is equal 8.0176 mm^3/s

200/60*(1.75/2)^2*pi = 8.0176 mm^3/s

Later tried to increase tension between gears and filament using adjustment screw and the results:

F [mm/min] E300mm [gr]
300 0.79
400 0.70

So 300mm/min seems ok:

300/60*(1.75/2)^2*pi = 12.026 mm^3/s

Pressure Advance (PA) calibrations

Uses to start pushing fillament a little before starting printing and stopping extruding a little earlier, to compensate for long Bowden tubing that connects extruder and printing head and gives slack for filament to move and compress like spring inside tubing.

Using Orca slicer’s Calibrations menu created PA pattern for Bowden printers:
PA calibration pattern

Looking at the corners, at 0.6 it looks sharp and the under-extrusion just starts.

The value was put into config and into slicer settings.

[extruder]
...
pressure_advance = 0.6
...

After configuring this I started to get Klipper errors of MCU 'mcu' shutdown: Stepper too far in past. Few evening past, I understood that it happens at short intervals and interplays between feed rate (F) and PA values. Still, didn’t solve it completely.

Input Shaper

Used to modify frequency of pulses provided to motors to avoid generating oscillations due to mechanical resonances in printer construction.

Calibrated accordingly to Klipper manual

My initial prints and comparison of 2 methods EI and MZV. Haven’t learned them yet, but sticked to recommendation to use MZV as less smoothing.

3 prints:

  • - - no input shaping, starting accel=1000, step 500
  • MZV - starting accel=500, step 1000
  • EI - starting accel=500, step 1000

X axis:

Input shaping X axis

Y axis:

Input shaping Y axis

Initial input shaper print failed because vibrations and probably bad cleaning of printer head before print. Solved by enabling skirt in slicer and so few loops of filament are extruded around model and so all leakages are wiped off before starting.

Continue to OrcaSlicer configuration…