Hello everyone,
I would like to add an important update to my original post because I have now performed several firmware-level tests, and the problem still remains.
This is the original CFS supplied with the Creality Hi Combo, not the CFS-C.
The main symptom is still the same:
-
The CFS loads filament normally.
-
The CFS unloads and retracts filament normally.
-
The extruder starts consuming filament.
-
The buffer flag moves from the upper position to the lower position.
-
The printer correctly detects that the buffer is empty.
-
The CFS does not start feeding again.
-
The extruder continues pulling against the resistance of the CFS until the PTFE tube bends and the extruder begins clicking or slipping.
I confirmed the buffer state directly through the Klipper console.
With the buffer flag in the upper position:
BOX_GET_BUFFER_STATE ADDR=1
buffer_state: 0x3
After the extruder consumes the filament stored in the buffer and the flag moves down:
BOX_GET_BUFFER_STATE ADDR=1
buffer_state: 0x2
empty
This confirms that the printer is receiving the buffer state change correctly.
I also tried manually enabling CFS printing:
BOX_ERROR_CLEAR
BOX_ENABLE_CFS_PRINT ENABLE=1
However, even after this command, the CFS still did not feed when the buffer changed to 0x2 / empty.
Firmware tests performed
Test 1 — Printer 1.1.0.73 and CFS 1.4.2
The printer was originally running:
Printer firmware: 1.1.0.73
CFS firmware: 1.4.2
CFS hardware/boot variant: G32
I performed a forced and targeted reinstall of the same CFS firmware version, even though version 1.4.2 was already installed.
The firmware file was verified before flashing, and the process completed successfully:
old version: cfs0_050_G32-cfs0_000_142
will update: cfs0_050_G32-cfs0_000_142
update: ok
Immediately after the reinstall, the CFS fed filament correctly once. The buffer flag moved down, the CFS motor activated, and the flag moved back up.
However, when the printer started the normal print preparation sequence, it retracted the filament and then loaded it again. After that cycle, the problem returned.
The extruder consumed the filament in the buffer, the flag moved down to 0x2 / empty, and the CFS stopped feeding again.
Therefore, reinstalling CFS firmware 1.4.2 did not permanently solve the problem.
Test 2 — Printer 1.1.0.65 and CFS 1.1.3
To rule out a compatibility problem with CFS firmware 1.4.2, I downgraded the printer to:
Printer firmware: 1.1.0.65
I then performed a targeted downgrade of the CFS from version 1.4.2 to version 1.1.3 G32:
old version: cfs0_050_G32-cfs0_000_142
will update: cfs0_050_G32-cfs0_000_113
update: ok
After a complete shutdown and restart, the versions were confirmed as:
Printer firmware: 1.1.0.65
CFS firmware: cfs0_050_G32-cfs0_000_113
I then loaded the filament normally through the CFS.
The buffer initially reported:
buffer_state: 0x3
I heated the nozzle to 220 °C, enabled CFS printing, and manually extruded filament in small increments.
When the buffer flag moved down, the printer correctly reported:
buffer_state: 0x2
empty
However, the CFS motor did not start feeding.
The exact same problem therefore occurred with both firmware combinations:
Printer 1.1.0.73 + CFS 1.4.2
Printer 1.1.0.65 + CFS 1.1.3
Test 3 — Printer firmware 1.1.0.50
I also installed printer firmware version 1.1.0.50.
With this version, the CFS initially worked. During manual extrusion, the CFS fed additional filament approximately three times. Each time, the buffer flag moved back up normally.
After those first few successful cycles, the CFS suddenly stopped feeding again.
The buffer flag moved down, the system detected the empty condition, but the motor no longer activated.
This makes the problem appear intermittent rather than being caused only by one specific firmware version.
What is already confirmed
The following functions work normally:
-
CFS detection and communication;
-
Manual filament loading;
-
Manual filament unloading and retraction;
-
Motor operation during loading and unloading;
-
Filament movement through the CFS hub;
-
Filament movement through the buffer;
-
Buffer flag mechanical movement;
-
Detection of the upper buffer position;
-
Detection of the lower/empty buffer position;
-
Printing from the external spool holder.
The failure occurs specifically when the CFS should automatically feed more filament after the buffer reaches the empty position.
Parts under consideration
The buffer PCB is marked:
MF003_BUFFER_V23
The complete buffer part number appears to be:
4004010263
I have already purchased a replacement buffer, but it has not arrived yet.
I am also considering the complete central CFS movement assembly:
CFS Hub Kit
Part Number: 4004010265
This assembly appears to include the central motor, filament drive system, local sensors and odometer-related components.
My main questions now are:
-
Is the buffer itself responsible for sending the automatic “feed more filament” request to the CFS, or does it only report the upper and lower positions?
-
Since the printer correctly detects both 0x3 and 0x2 / empty, can the buffer PCB still be defective or intermittent?
-
Could the CFS Hub Kit, PN 4004010265, be responsible for this behavior even though manual loading and unloading work?
-
Could the odometer or one of the photoelectric sensors inside the hub cause the CFS to feed a few times and then stop?
-
Is the automatic feeding motor controlled directly by the CFS mainboard, meaning that the mainboard may be the actual defective component?
-
Which part would be the most logical replacement:
-
Filament Buffer, PN 4004010263;
-
CFS Hub Kit, PN 4004010265;
-
Hub photoelectric/odometer board;
-
CFS mainboard;
-
Complete CFS unit?
- The CFS also has an external power input. Is there a safe diagnostic procedure using the official external 24 V power adapter, or could this create a dual-power condition when the CFS is connected to the Hi Combo?
At this point, the same failure has occurred after a clean firmware reinstall, after a CFS firmware downgrade, and with several printer firmware versions.
Any technical information about which component interprets the buffer signal and commands the hub motor would be extremely helpful before I continue replacing parts.
Thank you.