I'm working through the equations for the provided AutoCal example and I don't get the same Tcal result as shown in Table 3. Using all the same inputs and getting the same Tmmt and Ferr values for row 1, but the Tcal value is about 2x those in the table. I've triple checked the equation. Can anyone confirm this too? Is the table incorrect?

Thanks.

Ref Freq. | 5.00E+07 | T_{xtal} | 2.00E-08 | |||||||

Output Freq. | 2.01E+09 | F_{PD} | 5.00E+07 | T_{PD} | 2.00E-08 | |||||

F_{VCO} | 1.25E+09 | |||||||||

Eq. 1 | AutoCal Time & Accuracy | Register | Value | Register | Value | |||||

T_{mmt} | 2.00E-08 | 0Ah[2:0] | n | 0 | 02h | R | 1 | |||

Eq. 2 | Ratio target VCO freq. to PD freq. | |||||||||

N | 2.50E+01 | 03h | N_{int} | 25 | 04h | N_{frac} | 0 | |||

Eq. 3 | FSM Time | |||||||||

T_{FSM} | 2.00E-08 | 0Ah[14:13] | 0 | -> | m | 0 | ||||

Eq. 4 | Expected # of VCO counts | |||||||||

V | 2.50E+01 | |||||||||

Eq 5. | VCO Frequency Measured | |||||||||

F_{VCOM} | 1.25E+09 | |||||||||

Eq 6. | Frequency Error | |||||||||

F_{err} | 2.50E+07 | |||||||||

Eq 8. | Calibration Time | |||||||||

T_{cal} | 1.06E-05 |

The error is due to the value of 'k'. k is not fixed at 3. Its value depends on the state of Reg0Ah[7:6]. The recommended k value is 01b, making k=1 (for all applications).

With m=0 & k=1 the Tcal time calculated will be correct.

You will note that the Tcal values calculated will differ by 0.56us from those listed in Table 3 of the datasheet (datasheet table is in error).

Note that the maximum VCO SPI clock speed is 50MHz so typically we have m=2 (allowing reference frequencies >50MHz). Reg0Ah[14:13]=

00b m=0;

01b m=2;

10b m=4;

11b m=5;

Typical recommended value for Reg0Ah = 2046h.