12 Jun
2010
12 Jun
'10
11:21 p.m.
It should be observed that the typical "time since last pulse" sort
of data collection done by most of these things (e.g. Catweasel) and
inerpreted by a "threshold" sort of algorithm does constittute a PLL
of sorts, since phase errors tend to self-correct long-term speed
variations (i.e. every sample is relative only to the previous one).
In other words, if the drive is spinning too slowly and pulses (1t,
1.5t and 2t for MFM) come a bit late, the error isn't cumulative.
FWIW,
Chuck