Virtually all modern pre-power combinations work on the principle of a
constant voltage output, that is, a low output impedance driving a high
input impedance. This has the benefit that as there is very little current
flowing, the cable resistance has no bearing (not that there is any of any
consequence in short lengths), but more importantly, the cable capacitance
will have a miniscule effect on the frequency response. (again for short
"domestic" distances)

As I understand it, Krell's CAST works on a different idea, that is driving
current from a high output impedance into a low input impedance, effectively
constant current.
This was also the principle behind the unpopular DIN standard, although the
output Z wasn't very high, and the input Z wasn't very low!. Although
keeping the input impedance of the power amp very low will also avoid any
noticable low-pass filtering due to the connecting cable, I don't see any
benefit of this approach, except commerically, to ensure that Krell sell a
complete Pre-power system.

Perhaps any devotees of Krell CAST may like to explain how it's better than
the conventional approach.

Serge

"Midlant" wrote in message
...
When matching a pre-amp to a power amp what is the correct match IRT
impedences? ie: low out on pre with high in on amp or vice-verse? What
are the benefits of the design?
I have also recently read that there are three (more?) different pre-amp
design set ups that also effect how the pre is matched to the amp. Could
anyone explain this, please?
Thanks.
John

"Midlant" wrote in message ...
When matching a pre-amp to a power amp what is the correct match IRT
impedences? ie: low out on pre with high in on amp or vice-verse? What
are the benefits of the design?

Most "normal" equipment works on the principle that the signal is an
AC voltage of varying amplitude & frequency. Thus, you want each
device to have low output impedance and high input impedance. The high
input impedance means the downstream device doesn't draw much current
(or power) so the upstream device doesn't have to work very hard. In
other words, it's easier (takes less power) to swing a given voltage
across a higher impedance load. The low output impedance means the
upstream device can deliver some current if necessary and is
relatively immune to cable effects.

I played around with impedance ratios when I designed & built my
preamp (ladder stepped passive attenuator switchbox) and found that if
the ratio of impedances is at least 100:1 (meaning input of downstream
is at least 100 times output of upstream) then I could not detect any
sonic differences. But when the ratio got down to 20:1 then I could
hear some subtle differences.

Based on that I'd guess 50:1 to be a reasonably safe rule of thumb,
meaning anything more is probably overkill as the effects would not be
audible.