SensorsのトピックRe: FXOS8700CQ High Pass Filter Implementation

FXOS8700CQ High Pass Filter Implementation

mikeweir — Tue, 03 Feb 2026 22:21:54 GMT

Hi,

Thanks for your clear and prompt answer to my question about table 64 in the FXOS8700CQ data sheet. Could you further elaborate the signal flow when the high pass filter is enabled? In particular, I'm interested in correctly utilizing the vector magnitude configuration, so that is the specific mode I'm asking about. For example, is the filter implemented in the early analog stages, in the DSP, before or after the squaring operation, the offset subtraction, etc. A block diagram, or a signal flow diagram, showing the relevant software-controlled variable parameters, or even a comprehensive narrative tracing the signal from the source through the vector amplitude limit detector, would be most helpful. Do you have anything along these lines you can share?

Thank you for your assistance,

MPW

Re: FXOS8700CQ High Pass Filter Implementation

david_diaz — Thu, 30 Nov 2017 17:44:57 GMT

Hello Mike,

If you want to use the Acceleration vector-magnitude detection from the FXOS8700CQ device, I highly recommend reviewing the application note below where the vector-magnitude of the accelerometers is explained.

AN4692 - Using the Xtrinsic FXLS8471Q Accelerometer Vector-Magnitude Function

Uses cases such as Detection of linear acceleration exceeding a threshold or Detection of chance of tilt angle exceeding a threshold are mentioned on the document above.

An example code can be found here. I consider you may find it useful. Please use it as a reference.

I hope you find useful this information.

Regards

David

Re: FXOS8700CQ High Pass Filter Implementation

mikeweir — Sat, 02 Dec 2017 15:06:55 GMT

Thank you, David, I'd found that app note already, and another also by Talat Ozyagcilar that is helpful (though hadn't found FXLS8471Q - Accelerometer vector-magnitude function example code before). They're tantalizingly close to what I'm looking for, yet you see that the high pass filter is not addressed in the figure, nor in the register settings. That makes sense to me, as the "tilt" detection is probably intended to respond to sustained (i.e. low frequency) changes in orientation. The use of the initial values (in registers 0x63 thru 68) deal with the initial offset, but do not form a true HPF. In my case, I want to ignore sustained changes, and operate on the rapid changes: thus, I want to enable the HPF offered in register 0x0F. Figure 1 in app note 4692 might be a good starting point, but it will require a bit more elaboration than just adding the register names or values to the lists shown feeding the control logic and CalcVECM (where I suspect the filter is implemented) blocks. I hope you can continue to investigate this for me.

Sincerely,

Mike