Clock switching in always-on component

• US 9,653,079 B2
• Filed: 02/12/2015
• Issued: 05/16/2017
• Est. Priority Date: 02/12/2015
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Kosut, Alexei E., Jones, Girault W., Gulati, Manu, Herbeck, Gilbert H., Millet, Timothy J.
• Primary Examiner(s): Goddard, Tammy Paige
• Assistant Examiner(s): Leland, III, Edwin S
• Application Number:
  US14/621,093
• Publication Number:
  US 20160240193A1
• Time in Patent Office: 824 Days
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command

Abstract

In an embodiment, a system on a chip (SOC) may include one or more central processing units (CPUs), a memory controller, and a circuit configured to remain powered on when the rest of the SOC is powered down. The circuit may be configured to receive audio samples and match those audio samples against a predetermined pattern. The circuit may operate according to a first clock during the time that the rest of the SOC is powered down. In response to detecting the predetermined pattern in the samples, the circuit may cause the memory controller and processors to power up. During the power up process, a second clock having one or more better characteristics than the first clock may become available. The circuit may switch to the second clock while preserving the samples, or losing at most one sample, or no more than a threshold number of samples.

Claims

1. A method comprising:

monitoring audio samples in a first component in an integrated circuit during a time that a remainder of the integrated circuit is powered down, wherein the first component is powered on during the time and is operating according to a first clock;

detecting a key phrase in the audio samples by the first component;

powering up at least a portion of the integrated circuit responsive to the detecting;

determining that a second clock used within the integrated circuit is available, wherein the second clock has at least one characteristic that is closer to an ideal clock characteristic than a corresponding characteristic of the first clock, wherein the second clock becomes available responsive to powering up the portion of the integrated circuit, and wherein the at least one characteristic comprises a duty cycle, and wherein the second clock has a duty cycle that is closer to a 50% duty cycle than a duty cycle of the first clock; and

switching to the second clock in the first component at a time that is prior to an arrival of subsequent audio samples to ensure that no more than a threshold number of audio samples is lost or corrupted during the switching.

2. The method as recited in claim 1 wherein the one or more characteristics include phase noise.

3. The method as recited in claim 1 wherein the one or more characteristics include frequency accuracy.

4. The method as recited in claim 1 wherein the one or more characteristics include period jitter.

5. The method as recited in claim 1 further comprising:

capturing a first plurality of the audio samples by the first component operating according to the first clock;

capturing a second plurality of samples by the first component operating according to the second clock;

writing the first plurality of samples and the second plurality of samples to a memory to which the integrated circuit is coupled; and

processing the first plurality of samples and the second plurality of samples from the memory as a single phrase.

6. The method as recited in claim 5 wherein the processing is performed by a processor that is part of the portion of the integrated circuit that is powered up responsive to detecting the key phrase.

7. The method as recited in claim 6 wherein the powering up comprises booting an operating system on the processor.

8. The method as recited in claim 1 further comprising generating the first clock in an oscillator in the integrated circuit.

9. The method as recited in claim 8 further comprising generating the second clock in a phase lock loop (PLL) in the integrated circuit.

10. An integrated circuit comprising:

one or more processors; and

a first circuit coupled to the one or more processors, wherein the first circuit is configured to;

remain powered up during times that the one or more processors are powered down;

receive a first plurality of audio samples captured by one or more audio input devices during a time that the one or more processors are powered down, wherein the first circuit operates according to a first clock during the time that the one or more processors are powered down;

detect a predetermined pattern in the first plurality of audio samples;

cause the one or more processors to power up responsive to detecting the predetermined pattern, wherein a second clock activates responsive to the power up; and

switch to the second clock responsive to the second clock becoming operable at a time that is prior to an arrival of a subsequent second plurality of audio samples, wherein the second clock has a duty cycle that is closer to a 50% duty cycle than a duty cycle of the first clock.

11. The integrated circuit as recited in claim 10 wherein the predetermined pattern represents sound comprising a verbal utterance of one or more words.

12. The integrated circuit as recited in claim 10 further comprising a memory controller coupled to the first circuit, wherein the first circuit is configured to provide one or more memory controller parameters to the memory controller to program the memory controller for operation responsive to the power up.

13. The integrated circuit as recited in claim 10 wherein the first circuit comprises a buffer configured to store the first plurality of the audio samples and a second plurality of audio samples received during and after the power up, wherein the first circuit is configured to write the first plurality of samples and the second plurality of samples to the memory.

14. The integrated circuit as recited in claim 10 wherein the second clock has a lower phase noise than the first clock during use.

15. The integrated circuit as recited in claim 10 wherein a period jitter of the second clock is lower than the period jitter of the first clock.

16. The integrated circuit as recited in claim 10 wherein a frequency accuracy of the second clock is higher than the frequency accuracy of the first clock.

17. A system comprising:

an audio input device;

an audio coder/decoder (codec) coupled to the audio input device and configured to generate audio samples from sound detected by the audio input device;

a memory; and

an integrated circuit coupled to the audio codec and the memory, wherein the integrated circuit includes an audio filter circuit, one or more processors, and a memory controller coupled to the memory, and wherein;

the audio filter circuit is configured to detect a predetermined pattern in the audio samples from the audio codec during a time that a memory controller and the one or more processors are powered down, wherein the audio filter circuit operates according to a first clock during the time;

the audio filter circuit is configured to cause the memory controller and the processors to power up responsive to detecting the predetermined pattern, wherein the power up causes a second clock to activate in the integrated circuit; and

the audio filter circuit configured to switch from the first clock to the second clock responsive to activation of the second clock at a time that is prior to an arrival of a subsequent audio sample, wherein the second clock has a duty cycle that is closer to a 50% duty cycle than a duty cycle of the first clock.

18. The system as recited in claim 17 wherein the integrated circuit includes an on-chip oscillator that sources the first clock and a phase-lock loop that sources the second clock.