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
As I recall you could not move a formatted ST506/412 HDD between controllers
without first reformatting. This is because the gap and header information
was likely different between different controller manufacturers. This
particularly applies to ECC but could include simple things like address
mark, sync byte, etc. So although your IDE drive gives you error free data
and a crystal will give you perfect serial timing (no pll required and the
pll in the controller never sees bit shift) the adaptor would have to
synthesize the particular format down to the bit and including the specific
ECC/CRC or the controller will post an error.
Since most of the ST506/412 controllers after the early ones were
"picocoded" state machines that did the serializing/deserialing I suspect
most any modern dsp can do the work, the real problem may turn out to be
getting the format information for the particular manufacturer and model
controller chip used in the system to which the adaptor attaches. Even
within manufacturers the format changed with generations and some
manufacturers did custom variants, some of which were rumored to be designed
to preclude generic ST506/412 drive attachment.
I suppose a very smart machine could learn by having a series of known data
patterns written to it, but that seems challenging.
So it may turn out to be an impossible task for other than the high volume
commercial controller chips and even there finding the specific ECC
algorithm might be difficult.
Just my 2 cents
Tom