Hi Peter, I have no input on operation of the PAPER bit code, just some tips on 1PPS distribution. Are you using a passive power divider of the type used for RF signals? Mini circuits or similar? This, or cascaded BNC or SMA T-s are not well suited to distributing multiple copies of a 1PPS signal with integrity.
There are a few solutions. We use CMOS distribution amplifiers from Stanford Research Systems. http://www.thinksrs.com/ You can configure the distribution amplifier you need from this page: http://www.thinksrs.com/products/FS730-3.htm (Can’t link directly to a sample product but be sure to select a base or main frame, and add the CMOS distribution amplifier “Internal Option”. Good luck. Jonathan > On Jan 29, 2015, at 9:00 PM, Peter Niu <[email protected]> wrote: > > Hi, David, > We are facing a sync problem.We deployed 8 ROACH2s in our system now.Taught > as the PAPER model ,I connect a 1pps in the sync_in port of pf1,and put > signal out from sync_out of pf1 into the sync_in of pf2,and so on. When I run > the paper_init.rb to initial all the 8 ROACHes,However The different ROACH > has their own sequence of MCNT . Different ROACH(different FID) send the > different MCNT in the same time,which should stay the same. > I thought it is a 1PPS sync problem ,I want to use a power divider to split > the 1PPS signal into 8 ways to each ROACH.But the power divider did not work > in our lab. (I thought the power divider need to be special) Do you have some > good suggestions? > Thanks for your help again! > Peter > >

