2 Dec
2011
2 Dec
'11
2:53 p.m.
In time, it was quantized when it was recorded and
retains that
quantisation when read (note I said discrete time slots relative to
each other).
Well, it retains it modulo various timing errors which we've already
discussed.
Hwowever, that is not my point. Yes, if you record time-quantized data,
then you get time=-quantized data back. But nothign i nthe drive enforces
that time quantization, other than having minumum nad maximum pulse
separation times.
Think of the following system -- nothign to do with disks. Suppose you
have 2 old-fashionerd (really old-fashioned) modems. 300 baud Bell 103
devices, say. The data input goes to the FSK modulator and selects one of
2 frequencies. The output of that is fed down a telephone line to the
demodulator at the other end which switches the receive data output
appropriateyl.
Now, is that receive data output time-quantized or not?
My view is that when used in the conventional way to transfer
asynchronous seiral data, the receiver data line is time-quantized withi
na characeter (once you have the start bit, the other bits follow at
reugalr time intervals), but not between characters (there is no
requiement for the start bits of successive chracters to be separated by
an integral number of cycles of the buad clock). And of course I don't
haev to send any form of serial data lon that link. I could wire the
input to a simple 2-stat sesnor (say a light barrier) and say use it to
signal when my cat enters my machine room. In that case it's a big
stretch to say the output of the demodulator is time-quatized.
-tony