Computation spreading utilizing dithering for spur reduction in a digital phase lock loop

• US 20090262877A1
• Filed: 04/17/2008
• Published: 10/22/2009
• Est. Priority Date: 04/17/2008
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Staszewski, Roman, Staszewski, Robert B., Shi, Fuqiang
• Granted Patent: US 8,134,411 B2
• Time in Patent Office: Days
• US Class Current: 375/376
• CPC Class Codes: H03L 1/022 by indirect stabilisation, i.e. by generating an electrical correction signal which is a function of the temperature H03L1/021 takes precedence

Abstract

A novel and useful apparatus for and method of spur reduction using computation spreading with dithering in a digital phase locked loop (DPLL) architecture. A software based PLL incorporates a reconfigurable calculation unit (RCU) that is optimized and programmed to sequentially perform all the atomic operations of a PLL or any other desired task in a time sharing manner. An application specific instruction-set processor (ASIP) incorporating the RCU is adapted to spread the computation of the atomic operations out over a PLL reference clock period wherein each computation is performed at a much higher processor clock frequency than the PLL reference clock rate. This significantly reduces the per cycle current transient generated by the computations. The frequency content of the current transients is at the higher processor clock frequency which results in a significant reduction in spurs within sensitive portions of the output spectrum. Further reduction in spurs is achieved by dithering the duration of the software loop of atomic operations and/or by randomly shuffling one or more non-data dependent instructions within each iteration of the software loop.

Claims

1. A method of reducing the generation of frequency spurs in the performance of a processing task normally performed within a reference clock period, said method comprising the steps of:

dividing said task into a plurality of atomic operation computations for execution in a software loop;

randomizing the execution of one or more atomic operations in each iteration of said software loop; and

wherein said atomic operations are clocked using a processor clock having a frequency significantly higher than that of said reference clock.

2. The method according to claim 1, wherein said computations of atomic operations are spread out in time throughout said reference clock period to reduce the effects of said frequency spurs.

3. The method according to claim 1, wherein said processing task comprises a digital phase locked loop.

4. The method according to claim 1, wherein said step of randomizing comprises the step of dithering the duration of each iteration of said software loop.

5. The method according to claim 1, wherein said step of randomizing comprises the step of inserting a plurality of random wait cycles throughout said software loop.

6. The method according to claim 1, wherein said step of randomizing comprises the step of shuffling the order of execution of one or more instructions within said software loop.

7. The method according to claim 1, further comprising the step of applying closed loop control whereby said step of randomizing is adjusted in accordance with one or more performance data collected from the execution of said processing task.

8. A method of reducing the generation of frequency spurs in a software based digital phase locked loop normally performed within a reference clock period, said method comprising the steps of:

dividing computation of said phase locked loop operation into a plurality of atomic operation computations for execution in a software loop;

dithering the duration of each iteration of said software loop; and

wherein said atomic operations are clocked using a processor clock having a frequency significantly higher than that of said reference clock.

9. The method according to claim 8, wherein the average period of execution of said software loop equal to said reference clock period.

10. The method according to claim 8, said step of dithering changes the spectrum of spurs generated by said atomic computations.

11. The method according to claim 8, wherein said step of dithering comprises the step of inserting a random wait period at the end of said software loop.

12. The method according to claim 11, wherein the duration of the random wait period in each iteration of said software loop is set in accordance with a predetermined wait pattern.

13. The method according to claim 8, wherein all computations required to perform said phase locked loop operation are completed within said reference clock period.

14. The method according to claim 8, wherein dithering the duration of each iteration of said software loop significantly reduces effects of current transients.

15. The method according to claim 8, further comprising the step of applying closed loop control to said method whereby said step of dithering is adjusted in accordance with one or more performance data collected from the execution of said digital phase locked loop.

16. A method of reducing the generation of frequency spurs in a software based digital phase locked loop normally performed within a reference clock period, said method comprising the steps of:

dividing computation of said phase locked loop operation into a plurality of atomic operation computations for execution in a software loop;

dithering the duration of each iteration of said software loop by inserting a plurality of random wait periods throughout said software loop; and

wherein said atomic operations are clocked using a processor clock having a frequency significantly higher than that of said reference clock.

17. The method according to claim 16, wherein the average period of execution of said software loop is equal to said reference clock period.

18. The method according to claim 16, said step of dithering changes the spectrum of spurs generated by said atomic computations.

19. The method according to claim 16, wherein the duration of each random wait period is set in accordance with a predetermined respective wait pattern.

20. The method according to claim 16, wherein all computations required to perform said phase locked loop operation are completed within said reference clock period.

21. The method according to claim 16, wherein dithering the duration of each iteration of said software loop significantly reduces the loop frequency spurs.

22. The method according to claim 16, further comprising the step of applying closed loop control to said method whereby said step of dithering is adjusted in accordance with one or more performance data collected from the execution of said digital phase locked loop.

23. A method of reducing the generation of frequency spurs in a software based digital phase locked loop normally performed within a reference clock period, said method comprising the steps of:

dividing computation of said phase locked loop operation into a plurality of atomic operation computations for execution in a software loop;

randomly shuffling the order of execution of one or more atomic operations in each iteration of said software loop; and

wherein said atomic operations are clocked using a processor clock having a frequency significantly higher than that of said reference clock.

24. The method according to claim 23, wherein said one or more shuffled atomic operations comprise non-data dependent atomic operations.

25. The method according to claim 23, further comprising the step of concatenating a plurality of reshuffled software loops to generate a modified software loop, wherein the atomic operation order of execution in each reshuffled software loop is unique.

26. The method according to claim 25, wherein higher numbers of concatenated software loops yields higher reductions in loop frequency spurs.

27. The method according to claim 23, wherein all computations required to perform said phase locked loop operation are completed within said reference clock period.

28. The method according to claim 23, wherein dithering the duration of each iteration of said software loop significantly reduces loop frequency spurs.

29. The method according to claim 23, further comprising the step of applying closed loop control whereby said step of randomly shuffling is adjusted in accordance with one or more performance data collected from the execution of said digital phase locked loop.

30. A phase locked loop (PLL) having a reference clock, said PLL comprising:

means for partitioning computation of said phase locked loop into a plurality of atomic operation computations for execution in a software loop;

means for randomizing the execution of one or more atomic operations in each iteration of said software loop;

a processor clock having a frequency significantly higher than that of said reference clock;

an oscillator operative to generate a radio frequency (RF) signal having a frequency determined in accordance with a tuning command input thereto; and

a processor operative to generate said tuning command, said processor clocked at said processor clock rate, said processor comprising a calculation unit operative to execute instructions, wherein each instruction is operative to perform one of said atomic operation computations.

31. The phase locked loop according to claim 30, wherein said means for randomizing comprises the step of dithering the duration of each iteration of said software loop.

32. The phase locked loop according to claim 30, wherein said means for randomizing comprises the step of inserting a plurality of random wait cycles throughout said software loop.

33. The phase locked loop according to claim 30, wherein said means for randomizing comprises the step of shuffling the order of execution of one or more instructions within said software loop.

34. The phase locked loop according to claim 30, further comprising a time to digital converter (TDC) for measuring a phase error between said reference clock and said RF signal.

35. The phase locked loop according to claim 30, further comprising a clock divider coupled to said oscillator for generation of said processor clock.

36. The method according to claim 30, further comprising a closed loop control unit whereby said means for randomizing is adjusted in accordance with one or more performance data collected from the execution of said phase locked loop.