There can be other differences between SMT and DIPs however.
In the early 1980's, the computer industry was switching from 14 and 16 pin DIPs to 20 and 24-pin DIPs, in the transition from TTL to ASTTL and FAST (Fairchild's equivalent). Things did not go well -- the manufacturers had always ignored the inductance of the bond wires inside the packages, and for 14 and 16 pin DIPs they were right. But the combination of much faster switching times, and substantially longer GND and VCC bond wires inside the packages meant that when high-current-transient events were occurring on the die (simultaneously switching several buffers in the same direction, for example) a very high voltage could be induced on the ground bond wires. In effect, for a short time the entire die was referenced up to 2V higher or lower than ground, while the input voltages were as before. This meant clocked devices could re-clock themselves, or output buffer settling times would be much longer than advertised.
These problems went away only when the industry went to surface-mount technology. Yes, the silicon die were the same size as before, but now there were no long bond wires and therefore no induced voltage transients.
I have no knowledge of whether any of the above directly applies to any of the chips in question here.
-BobC
> I have no opinion about SMT in general so this is not a critical post - just a
> question. I have wondered for some time if the physics of smaller parts -
> like SMT parts - and I guess transistors especially, causes them to respond
> differently in VCOs, VCFs,
Errr....no.
a) for one thing, SMT is 99% a ∗mechanical mounting∗ thing, not a silicon thing.
For example, the axial 0.1uf bypass caps in the kits today are ∗actually∗ SMT
chip
caps with leads attached :) In fact, I can argue that the leaded bypass caps are
∗degraded∗ SMT caps.
Also, the DIP ICs were are using now are using the ∗exact same die∗ inside as
the SMT parts do. All that changes are the 'pins', and that comes into play at
frequencies over 400Mhz.
b) so, the only 'concern' is caps (in the audio path) and the resistors. Well,
resistors
in SMT are just as stable and just are reliable (if not more, because SMT reflow
soldering is much more reliable than thru-hole, especially in
high-shock/impact). That
leaves the caps. In situations where I feel the circuit is ∗better served∗ with
a certain
thru-hole cap, then that's what I will do. There is no "rule" that every part
HAS to be SMT,
it's just cheaper that way. But the pots and wires are thru-hole, so much for
that :)
Remove the tin-foil hats!
Paul S.