During measuring gas sensing parameters base-line resistance is not fixed and gets change sinusoidally before the target gas enters. What is the reason? Is there any way to fix it?
A good sensing material should have a reasonably stable base-line resistance. It should be fast in reaching to base line resistance after the test gas purge. It has to be optimised as a function of temperature and should be in the measurable range. You may find this article useful.
From the R(t) curve (1.jpg), the time interval is ~170s (gas modulation period?) for the peaks Rp(t) and valley Rv(t), the peak and valley values show some shifts with period/time might due to the transient effects of gas concentration, temperature etc. inside the sensor/system, which should not significantly affect the sensors research (optimize material, sensor, system and measurement etc, will be more application/engineering considerations).
For a given measurement conditions (gas concentration, purge rate/speed, modulation period, etc.), take/record 10~50 periods R(t), pick up the peak points Pp(t) or/and valley points Rv(t) could fit out the peak and valley transient curves (differential time constant, amplitude, etc.), for a desired/expected time (not the very early measured periods), use the fitted peak curve/envelope Pp(t) and valley curve/envelope Pv(t) for calculation (the base line shift effect should not significantly affect the analysed results).
If R(t) (1.jpg) related to the absence of target gas, still show 170 s period and R(t) changed from 2.6 MOhm to 3.3 MOhm, that might due to measurement or un-stable feature effects, which need more information about the sensor/system real time temperature (how to control and measure the real time sensor temperature), how to measure the resistance R(t) (use a normal ~3 MOhm resistance replace the sensor to check the resistance measurement) , details of the sensor structures (material, layers, sizes, contacts, etc.).
the most possible reason is caused by heating system changes. this change may caused by heating wire oxidizing or insulate layer leaking for MEMS chip. you can verify it by measure the cold resistance on heating wire or direct measure the temperature by FLIR..