LDO Orbiscan - Eddy Probe

[PFarm]

Found this on Facebook. Link Here

[atrushing]

This looks interesting. Being so low profile it will be a lot easier fitting different toolheads.

[polyneutron]

Now mellow also has a probe. Seems everyone is comming out with a eddy current probe.

[PFarm]

2 hours ago, atrushing said:

This looks interesting. Being so low profile it will be a lot easier fitting different toolheads.

Yes, and an I2C connection, which most toolhead boards have. No extra cable run from the motherboard will be nice.

[PFarm]

 

Update from Facebook

The Orbiscan Project

This is a project I originally started about two years ago  — Orbiscan.

The goal from the beginning was to create a better bed sensing solution based on the inductive sensing principle. A simple, cheap and effective design.

 

Key design features

• I2C interface - compatible with Orbitool O2S

• Integrated digital temperature sensor

• 20 mm coil diameter
  The coil size directly defines how far the sensor can “see.” A larger diameter improves accuracy at greater distances, allowing for a larger gap between the sensor and the bed.

• Dual-layer inductive coil design
  Helps reduce temperature-induced drift of the sensing coil, resulting in more stable distance readings—especially in heated chambers.

• Improved coil drive circuit
  The voltage across the coil is less dependent on target distance, preventing out-of-range errors and improving robustness.

 

Why didn’t it go public back then?

Despite all these improvements, the sensor never reached the level of precision I was aiming for.

There were always small drift issues that I couldn’t fully eliminate. Even though temperature compensation reduced the problem, it never removed it entirely. The original Klipper temperature calibration process was also quite cumbersome and didn’t deliver consistent results.

I developed my own temperature drift calibration macro, which helped but still wasn’t perfect.

The biggest challenge remained zero-contact point detection:

• It drifted with sensor temperature

• It was also affected by X-gantry deformation (mostly due to bimetallic effects)

• And more generally, all printers deform to some extent as the chamber heats up, which adds another layer of variability

 

What changed?

Things improved significantly with recent updates to Klipper, particularly the introduction of tap detection. with inductive probes.

The early versions weren’t much better than my custom macros—but the latest algorithm is a game changer.

It effectively eliminates the previous sources of inaccuracy:

• The sensor now behaves very reliably

• First layer consistency is excellent, regardless of chamber temperature

• Complex temperature calibration is no longer necessary

At this point, the only required calibration is:

• Sensor current

• Distance reading

 

Final thoughts

A huge thanks to Kevin O'Connor and everyone involved in developing this algorithm. Their work made it possible to finally achieve the level of reliability I was originally aiming for.

 

Would you like this for your printer?

 

 

 

[polyneutron]

Any of these inductive probes hold up to pure PC, PEEK, ULTEM. 300 to 400 degree nozzles and 120 to 150 degree beds with chambers 40 to 50C. Many claim 108C and only register 70 at most. My inductive probes have meet a slow and agonizing death. Do any really have a max temp of 100 plus for 8 to 12 or more hours.

Carto worked great until bed Temps hit 128C for 13 hours. Works with most blends of PV due to lower bed and nozzle Temps. Open to suggestions.

Edited May 4 by polyneutron

[polyneutron]

What is the max temperature if you have determined it yet. Can it hold up running bed temperatures at 130 and nozzle temps 300 for 12 hours or more in an enclosed chamber at 40°C. That and I have been running it this way for more than a week before the other inductive probe rated to 108 degress and never got close to half of that value started to fail and finally failed. No visible damage but it no longer works.