28 Dec
2012
28 Dec
'12
4:29 p.m.
On 12/28/2012 04:03 PM, John Wilson wrote:
Not. In the slightest. Instant resync on every
passing glitch.
Works fine in practice! But I'd love to be educated on how to do
better -- a lot of the disks people would read with something like
this will be in much worse condition than what I'm testing with.
One thing that helps is to implement a bandpass filter on your
pulses--just reject everything outside of the pass band. If you can
determine the data rate (and hence the average interval between pulses),
it appears to work even better.
I tried a bunch of PLL software simulations and found that in nearly all
cases, the bandpass approach worked better on FM, MFM and MMFM as well
as GCR.
YM of course MV.
--Chuck