Now that MPU6000 and MPU6500 are gone, and 90%+ use BMI270.

What is the consensus? Having issues? Works OK?
Did you have to use more filtering in BF 4.3?

Looks like BMI270 will be here to stay for years.

For me gyro noise in 40hz - 100hz is dead quiet. Way quieter than mpu6000 in that range. However I have a resonance at 136hz that I never had before. Tried lowering dynamic notch min to 100hz but it's still there. I'm going to try a new board to see if I just got a bad one.

Based on Chris Rosser's testing the MPU6000 came out on top compared to the BMI270, but keep in mind that those tests and the blackbox logs were performed with the quads on the bench and not during real flight.

Mark Spatz (UAV Tech) has just done the same testing using blackbox logs from actual flights and on those tests the BMI270 appears to beat the MPU6000. Also keep in mind that Betaflight 4.3 has specific code to better process the data from a BMI270 gyro whereas previous versions of Betaflight didn't.

The only downside to with the BMI270 is that it has a frequency of 3.2K (compared to 8K on the MPU6000) which in turn also restricts the maximum frequency of the PID loop of Betaflight to 3.2K as well. But most people won't even notice the difference. I even run my F7 FC's on a 4K PID look just to relieve a bit of unnecessary load / stress on the CPU and they fly great.

Relevant video links are below.

Hi - So I have been messing with this gyro over the weekend and got a good tune out of it. I figured I'd post a level flight slow throttle ramp PID toolbox spectral graph. So no flips or rolls in this flight so all you see is the motor noise across the full throttle range.

I left my PIDs on default for this, LPF1 off, LPF slider to 2.0, RPM filtering enabled, 2 dynamic notches min/max 100-400hz with a Q value of 400, and a single biquad filter on D at 101hz,

This gyro picked up 2 frame resonances that the MPU6000 never picked up before. Not really sure why. That's why I have the extra dynamic notch. The frame resonance is at a lower frequency than I'd like (136 hz) but the dynamic notch took care of it without adding hardly any delay. I did test 2 flight controllers with the same BMI270 chip and they both picked up this resonance. No one else has reported this though so it could just be my build.

Anyway - overall it flies great. I've been able to push my PIDs higher than with the MPU6000 due to the low noise between 50-200hz (with the exception of the frame resonance at 136hz). You can see a first and second harmonic in the motor noise that the RPM filter has no problem taking care of. 

I'm currently running this with the above filtering with the master slider at 1.5 and even on a 90 degree day the motors aren't hot (pretty warm but not hot). I couldn't do that with the last MPU6000 I had on the same build. Also because I could get more aggressive with my filtering I have almost no propwash using Nazgul F5 5" props.

Hope that helps!

   
   

sub 100hz
   

Method I used is detailed in this video

That is a super usefull reply!!