The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Measurements >> Phase Noise and Jitter Measurements >> CDR transfer function & BW
https://designers-guide.org/forum/YaBB.pl?num=1281923334

Message started by casual on Aug 15^th, 2010, 6:48pm

Title: CDR transfer function & BW
Post by casual on Aug 15^th, 2010, 6:48pm

When we measure CDR transfer function, what is the input data pattern to the CDR? Is it a random data or 10101 data?

CDR BW for oc192 SONET is at least 4MHz for jitter tolerance requirement. Loop BW formula is alpha*K_total*R where alpha is data density.

If i use 101010 data pattern, the BW is K_total*R since alpha =1
If i use random data pattern, the BW is 0.5*K_total*R because alpha =0.5.

It gives me two set of BW.

Title: Re: CDR transfer function & BW
Post by casual on Aug 16^th, 2010, 10:41pm

random data is used in transfer function measurement. So i think alpha=0.5 should be used.

Title: Re: CDR transfer function & BW
Post by Mayank on Sep 7^th, 2010, 9:09pm

Quote:

random data is used in transfer function measurement. So i think alpha=0.5 should be used.

For actual results, use the transition density of the data stream which you are recovering. Transition density of incoming data stream differs for different protocols & is chiefly dependent on the type of coding used. [ 8b/10b OR 64b/66b ] In some cases, it can go as low as 0.2

--Mayank.

Title: Re: CDR transfer function & BW
Post by casual on Sep 23^rd, 2010, 3:07am

I am referring to PRBS input without any coding. It can test the robustness of the receiver circuit like long consecutive 1 or 0.
( line coding like 8B10B is for system level test).

So the data density is 0.5 for PRBS data input.

I think the 8B10B also has data density 0.5
64B/66B has much lower data density since it depends on data scrambler and data.
Am i correct?

Title: Re: CDR transfer function & BW
Post by love_analog on Oct 11^th, 2010, 6:38am

Data dentisity refers to #1s or #0s over the total number of bits.
Whether you do 8b/10b or 64b/66b, this only dictates max run lenght and not the transition densisty
As long as you have scrambled/PRBS data, you will get .5

Title: Re: CDR transfer function & BW
Post by Mayank on Oct 22^nd, 2010, 2:02am

Quote:

Whether you do 8b/10b or 64b/66b, this only dictates max run lenght and not the transition densisty

Defining max. no. of #1s or #0s in a particular bit stream also defines the min. transition density that can be encountered in a certain time window.
Every encoding specifies with itself a min. transition density that it will maintain.

@ casual: I think its 0.3 for 8b/10b encoding. Pls. confirm.

-
Mayank.