The Designer's Guide Community
Forum
Welcome, Guest. Please Login or Register. Please follow the Forum guidelines.
Sep 27th, 2021, 12:51pm
Pages: 1
Send Topic Print
Alexander PFD (Read 6121 times)
neoflash
Community Fellow
*****
Offline

Mixed-Signal
Designer

Posts: 389

Alexander PFD
May 26th, 2010, 11:59am
 
Hi,

Almost all BB-PD used in CDR have no frequency detection capability.

What kind of behavior they will show if there is large frequency offset between data and local sampling clock?

Thanks,
Neo
Back to top
 
 
View Profile   IP Logged
love_analog
Senior Member
****
Offline



Posts: 101

Re: Alexander PFD
Reply #1 - May 31st, 2010, 7:56am
 
They will a parallel Frequency detector loop. Once the VCO is brought within range, you will switch over to the PD loop
Back to top
 
 

loveanalog.blogspot.com
The Power of Analog
View Profile   IP Logged
supermoment
Senior Member
****
Offline



Posts: 167

Re: Alexander PFD
Reply #2 - Jun 2nd, 2010, 9:49pm
 
typically BB-PD has a narrow LPF bandwidth compared to linear PD
Back to top
 
 
View Profile   IP Logged
supermoment
Senior Member
****
Offline



Posts: 167

Re: Alexander PFD
Reply #3 - Jun 11th, 2010, 5:37am
 
Only feedforward CDR architecture (aka oversampling type) does not have frequency detector.
Back to top
 
 
View Profile   IP Logged
KangSub
New Member
*
Offline



Posts: 9

Re: Alexander PFD
Reply #4 - Jul 24th, 2010, 1:32am
 
hi,

followings are my opinion.

If there are large frequency offset between clock and data, BBPD will produce the random sequence of UP and DN outputs.

BBPD uses at least 3-sample in consecutive 2-UI for phase detection between clock and data and produces valid UP and DN according to the position of data transition.
on the other words, BBPD detects the position of data transition relative to that of clock, using consecutive 3-samples in 2-UI.
so, the frequency of clock and data have to be equal.

if there exist large frequency offset between data and clock, then, BBPD cannot sample 3-sample in consecutive 2-UI and cannot operate properly.  
in that case, the 3-samples have no information about the position of the data transition.
therefore, the output of the BBPD is not meaningful.
if the input data is PRBS, because the probability that sampled bit is 0 or 1 is always same, the UP and DN output of BBPD will also random sequence.

I don't think that Alexander BBPD  can detect any frequency information.

thanks,

- KangSub
Back to top
 
 
View Profile   IP Logged
loose-electron
Senior Fellow
******
Offline

Best Design Tool =
Capable Designers

Posts: 1638
San Diego California
Re: Alexander PFD
Reply #5 - Aug 31st, 2010, 4:39pm
 
Um... this is covered in a dozen textbooks...

look into pull in range, and frequency acquisition vs phase acquisition.

Back to top
 
 

Jerry Twomey
www.effectiveelectrons.com
Read My Electronic Design Column Here
Contract IC-PCB-System Design - Analog, Mixed Signal, RF & Medical
View Profile WWW   IP Logged
neoflash
Community Fellow
*****
Offline

Mixed-Signal
Designer

Posts: 389

Re: Alexander PFD
Reply #6 - Aug 31st, 2010, 5:31pm
 
Thanks for bring this reference source up.

Would you like to list 1-2 text books which you think are best in describing this issue?

Thanks.

loose-electron wrote on Aug 31st, 2010, 4:39pm:
Um... this is covered in a dozen textbooks...

look into pull in range, and frequency acquisition vs phase acquisition.


Back to top
 
 
View Profile   IP Logged
Pages: 1
Send Topic Print
Copyright 2002-2021 Designer’s Guide Consulting, Inc. Designer’s Guide® is a registered trademark of Designer’s Guide Consulting, Inc. All rights reserved. Send comments or questions to editor@designers-guide.org. Consider submitting a paper or model.