Accelerometer-aided gyroscope

• US 8,612,146 B2
• Filed: 02/09/2011
• Issued: 12/17/2013
• Est. Priority Date: 02/15/2010
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Waters, Deric W., Pande, Tarkesh
• Primary Examiner(s): BADII, BEHRANG
• Application Number:
  US13/023,987
• Publication Number:
  US 20120158296A1
• Time in Patent Office: 1,042 Days
• Field of Search: 702/96, 702/141, 702/145, 702/151, 623/10, 907/62, 701/509, 701/2, 701/37, 178/18.01, 340/539.11, 345/173, 345/179, 356/459
• US Class Current: 701/509
• CPC Class Codes: G01C 19/04 Details
• CPC Class Codes: G01C 21/185 for gravity
• CPC Class Codes: Y10T 74/1229 Gyroscope control

Abstract

A system comprises a gyroscope configured to produce a gyroscope signal, an accelerometer configured to produce an accelerometer signal, and a filter unit coupled to the gyroscope and having a configurable bandwidth. The filter unit configured to filter the gyroscope signal. The system also comprises control logic that is configured to alter the bandwidth of the filter unit based on the accelerometer signal.

Claims

1. A system, comprising:

a gyroscope configured to produce a gyroscope signal;

a low-pass filter unit coupled to said gyroscope and having a configurable bandwidth, said filter unit configured to low-pass filter the gyroscope signal; and

control logic coupled to said filter unit and configured to alter the bandwidth of the low-pass filter unit based on an accelerometer signal by;

computing a metric based on the accelerometer signal to remove an acceleration due to gravity;

comparing of the metric to a threshold;

setting a bandwidth of the low-pass filter unit based on an accelerometer signal and the computed metric;

wherein computing the metric comprising;

squaring the acceleration signal for each axis;

summing the squares of the acceleration signals;

taking a square root of the sum; and

subtracting a gravitational constant; and

,

wherein computing the metric M uses the equation

M=√

{square root over (var(Ax)+var(Ay)+var(Az))}{square root over (var(Ax)+var(Ay)+var(Az))}{square root over (var(Ax)+var(Ay)+var(Az))}

where “

var( )”

is a variance computed over a predetermined number of samples.

2. The system of claim 1 wherein the gyroscope comprises a multi-axis gyroscope and said gyroscope signal comprises a separate, single axis gyroscope signal corresponding to each axis of the gyroscope, and said low-pass filter unit comprises a separate filter corresponding to each single axis gyroscope signal, each of said separate filters having a bandwidth configurable by the control logic based on the accelerometer signal.

3. The system of claim 1 wherein said control logic is configured to assert a signal, based on the accelerometer signal, to said gyroscope to cause the gyroscope to transition to a low power state comprising at least one of an off state and a sleep mode.

4. The system of claim 1 wherein, the control logic alters the bandwidth of the filter unit for a first bandwidth based on the accelerometer signal being above a first threshold and alters the bandwidth of the low-pass filter unit for a second bandwidth based on the accelerometer signal being below the first threshold but above a second threshold, and the control logic asserts a disable signal to the gyroscope based on the accelerometer signal being below the second threshold, said disable signal causing the gyroscope to power off or transition to a low power or sleep mode.

5. The system of claim 1 further comprising a buffer into which the gyroscope signal is stored and said low-pass filter unit retrieves and then filters the gyroscope signal from the buffer.

6. The system of claim 1 wherein said low-pass filter unit comprises a plurality of filters.

7. The system of claim 1 wherein computing the metric M uses the equation

M=max(std(Ax), std(Ay), std(Ay))

where std( )is a standard deviation computed over a number of samples, max( ) function means to select a maximum of a set of the standard deviation of the number of samples.