Calibration method for phase-locked loops and related circuit

1. A method of calibrating a voltage-controlled oscillator (VCO) in a phase-locked loop circuit, the method comprising a dichotomous search in a set of candidate frequency bands via a sequence of subsequent halving steps with index i=1, . . . , N that produce reduced subsets of the set of candidate frequency bands, the reduced subsets having respective upper bound values, lower bound values, and central values, a central value of a subset resulting from the halving step of the index i in the sequence being a function of an average of an upper bound value and a lower bound value of a subset resulting from the halving step of index i−1 in the sequence, the method comprising:

identifying a desired calibration frequency for the VCO, f_des_cycle, as a number of cycles counted in a reference clock signal;

reading a current frequency value, read_freq, which is a function of a current operation frequency of the VCO;

performing the dichotomous search by:

storing a stored frequency value store_read_freq[i]=read_freq for i=1; and

calculating a first modulus value, |f_des_cycle−store_read_freq[i]| and a second modulus value, |f_des_cycle−store_read_freq[i−1]|;

checking whether the first modulus value, |f_des_cycle−store_read_freq[i]| is lower than the second modulus value|f_des_cycle−store_read_freq[i−1]|;

in response to the checking yielding a positive outcome, performing a further halving step in the dichotomous search with the index i increased by one; or

in response to the checking yielding a negative outcome:

reducing by one the respective upper bound and lower bound of a current subset in the dichotomous search prior to resuming the dichotomous search in response to read_freq being higher than f_des_cycle plus a tolerance value; or

increasing by one the respective upper bound and lower bound of the current subset in the dichotomous search prior to resuming the dichotomous search in response to read_freq being lower than f_des_cycle minus the tolerance value; and

calibrating the VCO to a frequency resulting from the dichotomous search.

2. The method of claim 1 , comprising starting the dichotomous search by setting a first value for store_read_freq[O]=read_freq without comparisons made.

3. The method of claim 1 , wherein the central value of the subset resulting from the i-th halving step in the sequence of subsequent halving steps in the dichotomous search is a nearest integer to the average of the upper bound value and the lower bound value of the subset resulting from the (i−1)-th halving step in the sequence.

4. The method of claim 3 , wherein the central value of the subset resulting from the i-th halving step in the sequence of subsequent halving steps in the dichotomous search is a lower nearest integer to the average of the upper bound value and the lower bound value of the subset resulting from the (i−1)-th halving step in the sequence.

5. The method of claim 1 , wherein the sequence of subsequent halving steps in the dichotomous search comprises:

a first, coarse search section performed with a first value for the tolerance value; and

a second, fine search section performed with a second value for the tolerance value, the second value being smaller than the first value.

6. The method of claim 5 , wherein the first value and the second value for the tolerance value are six and unity.

7. The method of claim 1 , wherein calibrating the VCO includes selectively activating and de-activating capacitors in a set of 2N capacitors in a capacitor bank according to a binary code.

8. The method of claim 7 , wherein the dichotomous search includes up to N subsequent halving steps.

9. The method of claim 1 , further comprising:

setting at least one threshold value for a number of subsequent halving steps in the dichotomous search, and

issuing a calibration error message in response to the number of subsequent halving steps in the dichotomous search reaching the threshold value.

10. The method of claim 5 , further comprising:

setting a first threshold value for a number of subsequent halving steps in the first, coarse search section in the dichotomous search, and issuing a first calibration error message in response to the number of subsequent halving steps in the first, coarse search section in the dichotomous search reaching the first threshold value, and

setting a second threshold value for the number of subsequent halving steps in the second, fine search section in the dichotomous search, and issuing a second calibration error message in response to the number of subsequent halving steps in the second, fine search section in the dichotomous search reaching the second threshold value.

11. The method of claim 1 , further comprising performing the dichotomous search in an open-loop condition of the phase-locked loop circuit.

12. A phase-locked loop circuit comprising:

a voltage-controlled oscillator (VCO);

a counter configured to read a current frequency value, which is a function of a current operation frequency of the VCO; and

a finite state machine configured to:

identifying a desired calibration frequency for the VCO, f_des_cycle, as a number of cycles counted in a reference clock signal;

receive from the counter the current frequency value;

perform a dichotomous search in a set of candidate frequency bands via a sequence of subsequent halving steps with index i=1, . . . , N that produce reduced subsets of the set of candidate frequency bands, the reduced subsets having respective upper bound values, lower bound values, and central values, a central value of a subset resulting from the halving step of the index i in the sequence being a function of an average of an upper bound value and a lower bound value of a subset resulting from the halving step of index i−1 in the sequence, the dichotomous search comprising the finite state machine configured to:

store a stored frequency value store_read_freq[i]=read_freq for i=1; and

calculate a first modulus value, |f_des_cycle−store_read_freq[i]| and a second modulus value, |f_des_cycle−store_read_freq[i−1]|;

check whether the first modulus value, |f_des_cycle−store_read_freq[i]| is lower than the second modulus value|f_des_cycle−store_read_freq[i−1]|;

in response to the check yielding a positive outcome, perform a further halving step in the dichotomous search with the index i increased by one; or

in response to the check yielding a negative outcome:

reduce by one the respective upper bound and lower bound of a current subset in the dichotomous search prior to resuming the dichotomous search in response to read_freq being higher than f_des_cycle plus a tolerance value; or

increase by one the respective upper bound and lower bound of the current subset in the dichotomous search prior to resuming the dichotomous search in response to read_freq being lower than f_des_cycle minus the tolerance value and

apply calibration signals to the VCO to calibrate the voltage-controlled oscillator in accordance with a frequency resulting from the dichotomous search.

13. The phase-locked loop circuit of claim 12 , wherein the finite state machine is configured to start the dichotomous search by setting a first value for store_read_freq[O]=read_freq without comparisons made.

14. The phase-locked loop circuit of claim 12 , wherein the central value of the subset resulting from the i-th halving step in the sequence of subsequent halving steps in the dichotomous search is a nearest integer to the average of the upper bound value and the lower bound value of the subset resulting from the (i−1)-th halving step in the sequence.

15. The phase-locked loop circuit of claim 14 , wherein the central value of the subset resulting from the i-th halving step in the sequence of subsequent halving steps in the dichotomous search is a lower nearest integer to the average of the upper bound value and the lower bound value of the subset resulting from the (i−1)-th halving step in the sequence.

16. The phase-locked loop circuit of claim 12 , wherein the sequence of subsequent halving steps in the dichotomous search comprises:

a first, coarse search section performed with a first value for the tolerance value; and

a second, fine search section performed with a second value for the tolerance value, the second value being smaller than the first value.

17. The phase-locked loop circuit of claim 12 , wherein the finite state machine configured to calibrate the VCO includes the finite state machine configured to selectively activate and de-activate capacitors in a set of 2N capacitors in a capacitor bank according to a binary code.

18. The phase-locked loop circuit of claim 17 , wherein the dichotomous search includes up to N subsequent halving steps.

19. The phase-locked loop circuit of claim 12 , wherein the finite state machine is configured to:

set at least one threshold value for a number of subsequent halving steps in the dichotomous search, and

issue a calibration error message in response to the number of subsequent halving steps in the dichotomous search reaching the threshold value.

20. The phase-locked loop circuit of claim 12 , wherein the finite state machine is configured to perform the dichotomous search in an open-loop condition of the phase-locked loop circuit.