[BUG] CORE One INDX: Tool Offset Sensor zero-sample failures disappear with LDC1612 FREF_DIVIDER=2

[jkopriva]

Printer model

Prusa CORE One INDX, 8 tools(Founders)

Firmware version

6.6.3, 6.9.0, Custom build with LDC1612 fref_divider changed from 1 to 2

Upgrades and modifications

No response

Printing from...

Calibration issue

Describe the bug

Tool Offset Calibration repeatedly fails during the contactless XY measurement with:
ContactlessOffset ... Failed to get first sensor sample
The failure is intermittent. An axis/tool may successfully return sensor data, while the next measurement can immediately fail to obtain the first sample.
Hardware troubleshooting
The same failure has been reproduced after testing:

• two Tool Offset Sensor PCBs
• multiple independent TOS cables
• cable routed outside the printer
• both PUB connectors on xBuddy Extension
• two xBuddy main boards
• replacement X/Y motors
• replacement toolhead carriage
• replacement PSU
• repeated gantry/belt checks and squaring
• calibration without PTFE tubes attached
  Cable and main xBuddy changes did not eliminate the failure.

How to reproduce

Try Offset calibration - it was failing for me almost always. Sometimes one tool passed, but the next one failed immidietly.

Expected behavior

Calibration of tool offset tool succeeds on first try.

Files

Official firmware behaviour

6.6.3: repeated failures. Most recent test: 0 successful calibrations in 5 attempts.
6.9.0: intermittent. Individual tools can sometimes calibrate after several attempts, and calibration can progress to later tools, but a complete run remains unreliable.
LDC1612 configuration
The Tool Offset Sensor uses an LDC1612 with a 40 MHz external reference clock.
The default channel configuration contains:
.fref_divider = 1,
CORE One INDX uses TOOL_OFFSET_SENSOR_GEOMETRY=SINGLE_COIL, and the contactless calibration uses CH1 as the single active channel.
Therefore the effective LDC1612 reference frequency is:
40 MHz / 1 = 40 MHz
TI specifies a maximum fREF of 35 MHz for LDC1612 single-channel operation.

Test

I built firmware changing only:

- .fref_divider = 1,
+ .fref_divider = 2,

giving:
40 MHz / 2 = 20 MHz
No hardware was changed between the unsuccessful stock-firmware testing and this test.

Result

Tool Offset Calibration successfully completed all 8 tools on the first attempt.
This is the first configuration I have tested that completed the entire calibration immediately.
I am going to reassemble the printer, perform a hard reset, rerun all printer calibrations and repeat the test to establish reproducibility.

Possible explanation

This does not yet prove that the 40 MHz reference frequency is the root cause.
However, CORE One INDX appears to operate the LDC1612 in single-channel mode with fREF=40 MHz, while the LDC1612 single-channel specification limits fREF to 35 MHz.
The result may therefore indicate that some LDC1612/TOS units operate successfully at 40 MHz while units with less margin can intermittently fail to produce samples.
This could potentially also explain why replacement Tool Offset Sensor boards fix the same Failed to get first sensor sample error for some users while the problem is not widespread across all INDX printers.

I have generated the bug report with AI(it was pretty long conversation about it) and also sent an email about this issues directly to Bondtech and Prusa support.