I've been working on this for a while using the raw data. I finally decided I needed to start making plots to show what I am talking about.
The Blue line in these plots represents the time between crank pulses. The variation is visible at all RPMs, but fades in my plots because the duration of my sample rate approaches the frequency of the variation. The data is grainy, but outside the error function for the sensors and all. So it is not noise.
What these plots show is something happening to vary the crank speed drastically on a consistent basis. I don't always catch every instance of the event in a sample frame, but the variation is always there on all my plots and is always ~12°-18° (2-3 reluctor teeth). The speed variation is about 10% and is happening very very quickly.
My Question is: What is happening ~20-30 times a revolution? 10% is a huge variation.
The Blue line in these plots represents the time between crank pulses. The variation is visible at all RPMs, but fades in my plots because the duration of my sample rate approaches the frequency of the variation. The data is grainy, but outside the error function for the sensors and all. So it is not noise.
What these plots show is something happening to vary the crank speed drastically on a consistent basis. I don't always catch every instance of the event in a sample frame, but the variation is always there on all my plots and is always ~12°-18° (2-3 reluctor teeth). The speed variation is about 10% and is happening very very quickly.
My Question is: What is happening ~20-30 times a revolution? 10% is a huge variation.