Hello Team,
We are seeking to detect metal presence from several centimeters. We have this sensor at our main board design
So I tested this sensor response on its EVM, and got quite poor results – it works, but only at very close range, we can distinguish significant delta at both channel 0, and channel1 readings.
See the most optimal configuration I reached + data offset which is read when there is no metal target:
I only reached it by try and error , and have no technical explanation of why those parameters are the most sensitive ones. I still can detect a target from a very short distance – well below 1 cm (then the readings are ~ 0x3ff – 0x3d), overwise readings delta is no more than 3-4 LSB.
We are using external OSC of 40 Mhz, as an EVB onboard option.
Can you suggest a more proper configuration for metal presence detection or maybe some HW changes? I tried to review online design notes and calculators but did not succeed to reach any conclusion, suitable for my use case.
I also reached additional configurations that seem to be sensitive enough, but still cannot understand the theory behind those settings.
Those are recommended settings from a design spreadsheet, but with those settings, it responds to target is not optimal:
Should I run the FEMM tool instead? I’m using the coils design that exists on EVM, did not change nothing in HW. 
Can you please advise what are the optimal parameters for this coils design and why?
For which target distance those coils are optimized and what should the testing setup. What should be an expected resolution for target at uH vs. distance?
If you have any sample results recording this would be a great help.
Thanks in advance
Kind regards,
Shai
