27 Jun
2012
27 Jun
'12
11:06 a.m.
On 27 Jun 2012 at 10:41, Tom Gardner wrote:
There has been a lot of discussion on this subject and
I have tried to
read all of it, but I may have missed a discussion of what I think is
the key problem in such a product. Apologies in advance if I am
repeating something.
I think you've got it right--my implementation seeks to simply sample
bit transitions without getting involved in the implementation of the
formatting details of IAMs, CRC/ECC--although one would expect that
MFM data is still MFM data. Hence the need for speed and RAM.
At least data rates are fairly standard--5MHz for MFM, 7.5MHz for 2,7
RLL.
--Chuck