AW: CMOS/TTL chip questions (was: Cmos Latch Octal)

[Haible Juergen]

	>The ideal part for an oscillator, in any case, is
	>the
	>easily-obtained CD4069UBE.

Yes, that's the one that always comes unbuffered, and it's great
when you need a linear operating point (crystal oscillator with
smooth rather than steep slope rectangle output, for example).

For simple RC oscillators, my fav is a single schmitt trigger inverter
(40106, I think), with RC lowpass feedback. Can't beat that in terms
of efficiency, one gate, one resistor, one capacitor.
When I have two ordinary
gates left from the rest of the design, I build the 2-gate / 2-resistor /
1-capacitor oscillator you described. Never ran into problems
so far (with "B" type chips), but then again I have not built more
than a handful of these oscillators yet.

	>Buffered [B] CMOS devices are types in which the output "on"
impedance
	>is independent of any and all valid input logic conditions, both
preceding
	>and present.  All such CMOS product are designated by the suffix
"B"
	>following the basic type number."

I wouldn't dare to say that's wrong (in fact I don't know), but I would
have expected exactly the contrary ! I thought "B" gates are more
or less the same as "A" gates for the logical part, but with some
extra inverters at the output. This should make the gain higher, the
output slope steeper, and should provide *more*, not less, decoupling
from the input voltages that "A" devices. At least for simple gates
like 4001. Again, it seems you have done more research than I,
but I feel puzzled because it's so against "common sense".

	JH.