LTC6955 datasheet(18/22 Pages) AD

18 / 22 page

LTC6955

Rev 0

For more information www.analog.com

Preliminary Technical Data

Advance Product Information Subject to Change

Rev PrA

APPLICATIONS INFORMATION

These calculations are also theoretical. They assume a

noiseless ADC with infinite resolution. All realistic ADCs

have both added noise and a resolution limit. The limita-

tions of the ADC must be accounted for to prevent over-

specifying the sampling clock.

Figure 8 plots the previous equations and provides a

simple, quick way to estimate the sampling clock jitter

requirement for a given input signal or the expected SNR

performance for a given sample clock jitter.

Figure 8. SNR vs Input Signal Frequency vs Sample Clock Jitter

The RMS jitter of an ADC clock source can be indirectly

measured by comparing a jitter dominated SNR measure-

ment to a non-jitter dominated SNR measurement. A jitter

dominated SNR measurement (SNRjitter) is created by

applying a low jitter, high frequency full-scale sine wave

to the ADC analog input. A non-jitter dominated SNR mea-

surement (SNRbase) is created by applying a very low

amplitude (or low frequency) sine wave to the ADC analog

input. The total clock jitter (tJ(TOTAL)) can be calculated

using Equation 4.

tJ(TOTAL) =

log10 10

−

SNRjitter

⎛

⎝

⎜

⎞

⎠

⎟

−10

−

SNRbase

⎛

⎝

⎜

⎞

⎠

⎟

⎡

⎣

⎢

⎤

⎦

⎥

πfIN

(4)

Assuming the inherent aperture jitter of the ADC (tJ(ADC))

is known, the jitter of the clock generator (tJ(CLK)) is

obtained using Equation 2.

ADC Sample Clock Input Drive Requirements

Modern high speed, high resolution ADCs are incredibly

sensitive components able to match or exceed labora-

tory instrument performance in many regards. Noise or

interfering signals on the analog signal input, the voltage

reference or the sampling clock input can easily appear

in the digitized data. To deliver the full performance of

any ADC, the sampling clock input must be driven with a

clean, low jitter signal.

Figure 9 shows a simplified version of a typical ADC sam-

ple clock input. In this case the input pins are labeled

ENC± for Encode while some ADCs label the inputs CLK±

for Clock. The input is composed of a differential limiting

amplifier stage followed by a buffer that directly controls

the ADC’s track and hold stage.

The sample clock input amplifier also benefits from a

fast slewing input signal as the amplifier has noise of its

own. By slewing through the crossover region quickly,

the amplifier noise creates less jitter than if the transition

were slow.

TOTAL CLOCK

JITTER (RMS)

10fs

20fs

50fs

100fs

200fs

500fs

1ps

FREQUENCY OF FULL-SCALE INPUT SIGNAL (GHz)

0.01

0.1

104

114

124

6955 F08

Measuring Clock Jitter Indirectly Using ADC SNR

For some applications, integrating a clock generator’s

phase noise within a defined offset frequency range (i.e.

12kHz to 20MHz) is sufficient to calculate the clock’s

impact on the overall system performance. In these situ-

ations, the RMS jitter can be calculated from a phase

noise measurement.

However, other applications require knowledge of the

clock’s phase noise at frequency offsets that exceed the

capabilities of today’s phase noise analyzers. This limita-

tion makes it difficult to calculate jitter from a phase noise

measurement.

LTC6955 Datasheet(PDF) 18 Page - Analog Devices

LTC6955 Datasheet(HTML) 18 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

Datasheet Download

Link URL