The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: email@example.com (Grant Edwards)
Subject: Re: Looking for MFM endec description
References: <3DA2D6FC.B68FEBC@yahoo.com> <3DA2EF5A.E07DAA01@webaccess.net> <3DA3057C.9C4AE87D@yahoo.com> <3DA33E44.FE8D2642@yahoo.com>
User-Agent: slrn/0.9.7.1 (Linux)
Date: Tue, 08 Oct 2002 21:41:14 GMT
NNTP-Posting-Date: Tue, 08 Oct 2002 16:41:14 CDT
In article <3DA33E44.FE8D2642@yahoo.com>, CBFalconer wrote:
>> >> Contemplate some fun codes now. How about an RLL code that is
>> >> also DC free?
>> > Obviously impossible on a bit by bit basis. But it should be
>> > possible to encode each of a particular N bit sequence into
>> > some M > N bit pattern that would be DC free, and table drive.
>> > How to create a DC free sync pattern? The result would be
>> > useful, especially for long line communications.
>> Manchester encoding has been used for that for a long time (N=1
>> and M=2).
> But that doesn't get the reduced bandwidth/transitions of a longer
> The DC doesn't usually matter for such things as local disk
> drives, but the transitions do. Both are of interest for long
> distance transmissions.
OK. Taking a simple minded approach:
A code will have no DC if each of the M-bit patterns contain
equal numbers of ones and zeros.
For M==2, there are two such patterns: 01 and 10. That lets
you encode one bit. (M=2, N=1).
For some other even values of M:
M S B
-- ----- -------------
2 2 1.0
4 6 2.58496250072
6 20 4.32192809489
8 70 6.12928301694
10 252 7.9772799235
12 924 9.85174904142
14 3432 11.7448338375
16 12870 13.6517244331
M: number of bits per symbol
S: number of symbols with equal number of 0's and 1's
B: number of bits encoded: log2(P)
So (ignoring synchronization issues) it appears that there are
codes with no DC component for N:M values
My suggestion would be to reserve the low frequency patterns
for some sort of preamble. IEW, if using the 10-bit pattern,
reserve the symbols 0000011111 and 1111100000 for non-data use
as start or stop symbols. You'd know that if you saw 10 1's in
a row, the next transition marked the middle of a word.
I've ignored the issue of minimizing the number of transitions...
Grant Edwards grante Yow! I've got to get
at these SNACK CAKES to NEWARK
visi.com by DAWN!!
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup