circuit too temperature sensitive

I assume this is the device in question:

Page 4 shows the temperature sensitivity of the oscillator negative resistance but it isn't extremely large (maybe 9% over more than 160 degrees C range).

The operation is also affected by the temperature sensitivity of the resonator coil resistance, the resonator capacitance and losses and probably most important, the core permeability and losses.

What kind of capacitor are you using in the resonator?

What kind of core material are you shaping the coil field with?

Yes, that's the part. newer spec sheet:

The resonator cap is a .001uf multilayer ceramic, currently only an X7R. The core is "P" ferrite, 2500 permeability:

That should be pretty good. I was afraid you might be using a Y5V or Z5u, which are terrible.

Since this is a gapped structure, it makes little difference in the shape of the magnetic field if the permeability is

What frequency is your resonator running at?

What temperature range are you hoping for?

The core is a pot core design. I'm running at around 100K Hz right now. Higher freqs were giving lower sensing distances. I'll need to cover a wide outdoor temperature range: -20F to 120F (-28C to 50C)

I guess I would try to total up all the temperature sensitive losses in the resonator structure and compare that to to the temperature sensitivities of the chip, to determine where the main sensitivity lies.

But if you have a choice of ferrites, you might pick the one with the most temperature independent losses and try substituting a much better capacitor, as an experiment to see how much the ceramic one is involved. At the very least, use an X7R with a voltage rating 10 to 20 times the peak voltage applied.

By the way, what value capacitor are you using?

.001uf

Well, that is small enough to be replaced with a C0G type, to eliminate the capacitor from the problem, pretty conclusively.