ADF5610 datasheet(21/51 Pages) AD

21 / 51 page

Data Sheet

ADF5610

Rev. 0 | Page 21 of 51

Autocalibration can be disabled, thereby allowing manual band

selection and even faster VCO tuning. Refer to the Manual

VCO Calibration for Fast Frequency Hopping section for more

information about manual tuning.

Autocalibration Using Register 0x05

Autocalibration transfers band control data to the VCO

subsystem via Register 0x05. The address of the VCO subsystem

in Register 0x05 is not altered by the autocalibration routine.

The address and ID of the VCO subsystem in Register 0x05

must be set to the correct value before autocalibration is

executed. For more information, see the VCO Serial Port

Interface (VSPI) section.

Automatic Relock on Lock Detect Failure

It is possible, by setting Register 0x07, Bit 13, to have the VCO

subsystem automatically rerun the calibration routine and

relock itself if the lock detect bit indicates an unlocked condition

for any reason. With this option, the system attempts to relock

only once.

VCO Autocalibration on Frequency Change

Assuming Register 0x0A, Bit 11 = 0, the VCO calibration starts

automatically whenever a frequency change is requested. To

rerun the autocalibration routine for any reason at the same

frequency, rewrite the frequency change with the same value,

and the autocalibration routine executes again without

changing the final frequency.

VCO Autocalibration Time and Accuracy

The VCO frequency is counted for the period of a single

autocalibration measurement cycle (tMMT).

tMMT = tXTAL × R × 2n

(1)

where:

tXTAL is the period of the external reference (crystal) oscillator.

R is the reference path division ratio currently in use, set in

n is set by Register 0x0A, Bits[2:0] and results in measurement

periods that are multiples of the PD period, tXTAL × R. The

frequency resolution used by the FSM during autocalibration is

dependent on the value of n and the PD frequency. Higher values

of n result in reduced frequency error (higher frequency

resolution) during autocalibration. The highest value is

currently recommended.

The VCO autocalibration counter, on average, expects to

for every PD cycle.

N is the ratio of the target VCO frequency (fVCO) to the

frequency of the PD (fPD) where N can be any rational number

supported by the N divider.

N is set by the integer register and fractional register contents

using Equation 2.

N = NINT + NFRAC/224

(2)

where:

NINT is the integer set in Register 0x03.

NFRAC is the fractional part set in Register 0x04.

The autocalibration FSM and the data transfers to the internal

VCO VSPI run at the rate of the FSM clock (tFSM) where the

FSM clock frequency must not be greater than 50 MHz.

tFSM = tXTAL × 2m

(3)

where m is 0, 2, 4, or 5 as determined by Register 0x0A, Bits[14:13].

The VSPI and FSM clock rate are determined by the value of m.

It is recommended to set Register 0x0A, Bits[14:13] = 2.

The expected number of VCO counts is given by

VCO Counts = floor(N × 2n)

(4)

The nominal VCO frequency measured (fVCOM) is given by

fVCOM = VCO Counts × fXTAL/(2n × R)

(5)

where the worst case measurement error (fERR) is

fERR ≈ ±fPD/2n + 1

(6)

For example, the autocalibration time (tCAL) for an 8-bit step

tuned VCO (where the total number of bits includes those needed

to switch VCO cores) is as follows. First, a 20 × (k + 1) bit wait

state occurs. Register 0x0A, Bits[9:3] set the value of the wait

state (k) for wideband devices to k = 1. Wideband VCOs like

the ADF5610 require additional wait time or one additional PD

cycle (k = 1), whereas narrowband VCOs operate with no

additional wait time (k = 0). The wait period is followed by the

tMMT for calibration, then the VSPI data transfers of 20 clock

cycles each. This process repeats eight times because there is

one bit to select the VCO core and seven bits to select the VCO

band for a total of eight bits. Total calibration time, worst case,

is given by the following equation:

tCAL = tXTAL (8R × 2n + (160 + (128 × (k + 1)) × 2m))

(8)

For guaranteed lock retention across temperature extremes

using a single subband (lock and leave), the frequency error

must be less than 1/8 of the frequency step required to induce a

VCO subband change.

VCO Autocalibration Example

The VCO subsystem must satisfy the maximum fPD limited by

the two following conditions:

N ≥ 16 (fINT), N ≥ 20.0 (fFRAC)

where:

N = fVCO/fPD. fPD ≤ 100 MHz.

fINT is integer mode.

fFRAC is fractional-N mode. The minimum N values change

depending on the operating mode.

For example, if the VCO subsystem output frequency is to

operate at 7.01 GHz and the crystal frequency is fXTAL = 50 MHz,

R = 1, and m = 2 (see Figure 22), then tFSM = 80 ns (12.5 MHz).

When using autocalibration, do not allow the autocalibration

FSM clock to exceed 50 MHz (see Register 0x0A, Bits[14:13]).

The FSM clock does not affect the accuracy of the measurement.

ADF5610 Datasheet(PDF) 21 Page - Analog Devices

ADF5610 Datasheet(HTML) 21 Page - Analog Devices

Html Pages

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Datasheet Download

Link URL