Temperature drift in LC oscillator

[mattiad]

Hi everybody,

I'm studying the causes of temperature drift in an LC oscillator. I examined the variation of the varactor and the parasitic junction capacitance in the couple of MOSFET. In your opinion are there any other parameters which have a variation in temperature?

(Sorry for my English, I'm an Italian student )

Mattia

 

[hakeen]

Hi Mattiad,

Your english is very good, dont mind.

I can imagine there would be several sources of drift in an LC oscillator,. I can remember the most ones as below

1-) Thermal
2-) Voltage
3-) Mechanical

I believe #1 applies to every component you use. It might have high or low effects. You can check it from the datasheets with ppm values. With MOS you can have either thermal or voltage sourced drifts. With inductors all three might be consideration.

Mechanical ones can be come into play especially if you use wires to connect parts of your system.

The below webpage explains in more details and give further calculations. It might be good to give a look.

https://www.electronics-tutorials.com/oscillators/oscillator-drift.htm

Cheers

 

[mattiad]

Thank you for the information.
I forgot to say that the oscillator I'm studying is an Integrated circuit.
I'm focusing the study of temperature drift so I think I can not consider the mechanical causes. My problem is the following: I examined precisely the varactor and the mos' behavior but the theoretical results don't match the simulation results, expecially if I use a capacitive degeneration. The electronics articles shows that the variation of inductance can be negligible.
Thank you again

 

[BigBoss]

If your models have temperature effect, you should normally see the drift of VCO frequency by simulating the VCO against temperature.
But if your models haven't any temperature effect at all, you can only measure this.
The drift comes generally comes from heat around the component of the VCO, so the tempearature is increased, the behaviours of the varactor and MOS transistors are changed and this causes a frequency drift by time.