1 Feb
2007
1 Feb
'07
9:48 a.m.
woodelf wrote:
Chuck Guzis wrote:
One way: pick off some signal that fires once per instruction execution
-- say some signal that occurs in a particular decode phase of the
logic. Buffer it and send it through a simple RC filter, which feeds
the meter. Add a pot to scale things properly. The more instructions
executed per second, the more pulses per second, the higher the average
voltage put out by the filter. You might even be able to skip the RC
filter since the meter has inertia.
The old GE machines with the "Thousands of Operations Per Second"
analogue meters were kind of neat too.
So just how does one construct such a
meter?