- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I want to use altgx at 3G . most serdes I used before using 8b10b before parallel-serial trans,but there is 20% payload wasted.so I want to use scrambler like 3G sdi .the there is no 8b10b in 3G sdi protocl,only srambler . when parallel 20bits data arrived,I scramble them,and packet them use my own-defined frame header.and i can use the frame header to do word alignment.But what I am worried about is : will the cdr work ok when there is no 8b10b?Can My serdes work fine for a long time?
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
As far as I know, 8b10b encoding is not essential for serdes CDR.
As long as there are adequate line transitions there should be no problem. But you will need to make sure that you send transitions when the lines are idle (interpacket periods) otherwise it will fail.- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
As Kaz indicates, there is no requirement that 8/10B encoding be used.
You can use a pseudo-random binary sequence (PRBS) to generate sufficient transitions for the CDR to lock to the data. The Altera Embedded IP Users Guide has information on their pattern generator and pattern checker components. I believe there is an example design either in that guide, or perhaps with one of the transceiver webinars, that tests a link with the PRBS, but not 8/10B encoding. The PRBS patterns used by the generator/checker components might match the scrambler polynomial used by SDI. If they do, then if you can scramble zero valued data at your SDI source, you can run the Altera checker at the destination and make sure the link operates correctly. If that does, then you will be able to capture and descramble your SDI packets just fine. Cheers, Dave- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Besides providing sufficient bit transitions, 8b10b encoding has also unique "komma" characters for frame synchronization, idle signaling and similar. Without adding some redundancy to the data coding, you can't have it.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
--- Quote Start --- Besides providing sufficient bit transitions, 8b10b encoding has also unique "komma" characters for frame synchronization, idle signaling and similar. Without adding some redundancy to the data coding, you can't have it. --- Quote End --- I accidentally saw this post. May i know what is the advantage of this "Komma" character especially the k28.5? Is it the "Komma" character has longest continuous "1" among 256 of the 8b10b valid code group? :confused:

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page