ADS4225: CLKOUT Jitter and SNR

Hideki Hayashi1

Expert 7520 points

Part Number: ADS4225

Hi teams,

I have questions about ADS4225 jitter.

I guess the clock jitter from CLKOUT is calculated from aperture jitter 140 fs.

It results in 1 / (140 * 77.7MHz) =92ps where fs is 77.7MHz. And it also equals t_jitter.

Is it correct?

Also SNR_jitter is calculated by an equation 10.2.2.4 (2).

If the SNR_jitter is calculated by the above t_jitter value, it results in 45dBc, where f_in is 10MHz. is it correct? I think it's too low.

Best regards,

Hideki

over 3 years ago

0 RJ Hopper over 3 years ago

TI__Mastermind 18095 points

Hideki:

Yes, the aperture jitter is 140 fs. The total jitter from the system is the RMS sum of the aperture jitter and the external clock jitter per equation 3 in section 10.2.2.4. I do not follow the equation 1/(t_aperture * Fs) and the total jitter will not be 92 ps.

If you use a really good external clock then the jitter performance is dominated by the aperture jitter. That limits SNR to better than 100 dB at 10 MHz due to jitter contribution alone. In reality, SNR performance at this frequency will be dominated by thermal noise and not jitter.

--RJH

0 Hideki Hayashi1 over 3 years ago in reply to RJ Hopper

TI__Expert 7520 points

Hi RJ,

Thank you for the advice.
Customer wants to use CLKOUT as a clock input for FPGA and needs to calculate the jitter to meet the FPGA clock requirements (< ±750ps). Of course, the external clock jitter needs to be included in the calculation, but how is the contribution of the aperture jitter calculated?

Best regards,

Hideki

0 RJ Hopper over 3 years ago in reply to Hideki Hayashi1

TI__Mastermind 18095 points

Total Jitter = sqrt( (aperture jitter)^2 + (clock jitter)^2 ).

--RJH

0 Hideki Hayashi1 over 3 years ago in reply to RJ Hopper

TI__Expert 7520 points

Hi RJ

Sorry for confusing you. I want to know how much the t_jitter (aperture jitter) in time domain which should be calculated from 140fs since fs is frequency. I guessed it was reciprocal in first question, but you said that is not correct. So I'd like to ask you.

Hideki

0 RJ Hopper over 3 years ago in reply to Hideki Hayashi1

TI__Mastermind 18095 points

The aperture jitter *is* in the time domain. It is 140 fs .... i.e. 140 femto-seconds. I understand the confusion since "fs" often refers to the sampling frequency. --RJH

0 Hideki Hayashi1 over 3 years ago in reply to RJ Hopper

TI__Expert 7520 points

Yes...I'm confused about as you think. I understand the jitter is written by femto seconds.

Thank you so much.

Regards, Hideki

Data converters

Data converters forum

ADS4225: CLKOUT Jitter and SNR