I have about 10 of these devices on about 6 computers, so this problem is common, the device runs at about 1200 to 1400 Hz. The picture shows a temperature, frequency and FFT results plot for one device. The typical is about 1400 +- 50. I take and FFT every 16384 data points, but I have to measure the time for each 16384 data points and adjust the FFT results. I have calibrated it against a standard accelerometer on a known vibration element.
Is this the expected behavoiur? it is not temperature dependent and only windows is running on the devices as they are remote.
I do not mind, I just want to confirm it.
One has to use the ticks on computer clock as the return on the time registers is not accurate enough nor calibrated.
