"Tim Padrick" wrote in message ...
Not so. In a series-pass crossover, the inductor would be in parallel
with the tweeter and the capacitor in series with the woofer. Two
SPICE files indicate the topological differences. First, the more
common parallel model (assume MyWoofer and MyTweeter are both
appropriate Spice sub circuit models of the woofer and tweeter, and
ignore the fact that the values may or may not be appropriate):

* Parallel crossover net list
Vin 1 0 AC SIN 1.0 0.0

Lwoof 1 2 1MH
Xwoof 2 0 MyWoofer

Ctweet 1 3 8UF
XTweet 3 0 MyTweeter

Now, the same as a series-pass model:

* Series crossover net list
Vin 1 0 AC SIN 1.0 0.0

Cwoof 1 2 8UF
Xwoof 1 2 MyWoofer

LTweet 2 0 1MH
XTweet 2 0 MyTweeter


An inductor in parallel with a tweeter, without a series
capacitor, would be seen by the amp as a short at low
frequencies. A cap in series with a woofer would roll off
the low frequencies.

Too bad you didn't try to understand or analyze the circuit.

If you were to put JUST a tweeter and JUST an inductor in parallel
across an amp, you'd be right, but that's VERY clearly, from the
topology above, NOT what is happening. You cannot look at a circuit
and simpy pick little pieces of it and expect a quick analysis based
on one piece to give you a coherent picture of the whole. Let's, in
fact, do a more complete analysis and see what is REALLY happening

We'll make some simplifying assumption just to make the analysis
easier: we'll replace XWoofer and XTweeter with resistive loads
RWoofer and RTweeter.

Start at frequencies well below the crossover point. At these
frequencies the impedance of shunt capacitor CWoof is very high,
and that of the shunt inductor LTweet is very low. As a result,
the current flows through the woofer leg (RWoofer) and through
the inductor LTweet) And the amplifier sees, essentially, the
woofer as the load.

Now, at high frequencies well ABOVE the crossover point, the
impedance of the shunt capacitor CWoof is very low and that of
the shunt inductor LTweet is very high. As a result, the current
flows through the shunt capacitor (CWoofer), bypassing the woofer
and flowing through the tweeter leg. And the amplifier sees, essentially,
the tweeter as the load.

Around the crossover, the impedance of the two shunt reactances
LTweet and CWoof) are about the same, and are also about the same
as the impedances of the woofer and tweeter (assuming we don't pick
values out of thin air, like I did). In such a condition, equal
amounts of the current flow through each leg of the mesh, meaning
the power to the woofer and tweeter are about the same. And the load
seen by the amplifier is the parallel combination of the woofer mesh
and tweeter mesh.

Contrary to your analysis at NO point does the amplifier EVER see
a short circuit.

Let's in fact do a more precise analysis. I have adjusted the
values to something a bit more practical, aiming for a 1 kHz
crossover. Here's the new circuit:

* Series crossover analysis
Vin 1 0 AC SIN 1.0 0.0

CWoof 1 2 25UF
RWoof 1 2 8

LTweet 2 0 1MH
RTweet 2 0 8

.AC OCT 3 20 20K
.PRINT AC VDB(1, 2) VDB(2,0)
.END

Now, here's the output response across the woofer terminals
and the tweeter terminals, plotted every 1/3 octave:

Frequency Woofer Tweeter
Response Response
---------- -------- --------
20 Hz 0.0 dB -36.1 dB
25.2 0.0 -34.1
31.7 0.0 -32.1
40 0.0 -30.1
50.4 0.0 -28
63.5 0.0 -26
80 0.0 -24
101 0.0 -22
127 0.0 -19.9
160 0.0 -17.9
202 0.0 -15.8
254 0.0 -13.7
320 0.0 -11.6
403 0.0 -9.4
508 -0.1 -7.3
640 -0.4 -5.2
806 -0.9 -3.3
1020 -1.9 -1.8
1280 -3.4 -0.9
1610 -5.3 -0.3
2030 -7.4 -0.1
2560 -9.6 -0.0
3230 -11.8 0.0
4060 -13.9 0.0
5120 -16 0.0
6450 -18.1 0.0
8130 -20.1 0.0
10200 -22.1 0.0
12900 -24.2 0.0
16300 -26.2 0.0
20500 -28.2 0.0

As you can see, the response problems you predict simply do not
happen.

Now, as to the assertion of there being a "short circuit" across the
amplifier, let's look at the impedance:

Frequency Impedance
---------- ---------
20 Hz 8 Ohms
25.2 8
31.7 7.99
40 7.99
50.4 7.98
63.5 7.97
80 7.96
101 7.93
127 7.89
160 7.83
202 7.74
254 7.61
320 7.41
403 7.16
508 6.84
640 6.51
806 6.25
1020 6.15
1280 6.26
1610 6.53
2030 6.87
2560 7.18
3230 7.43
4060 7.62
5120 7.75
6450 7.84
8130 7.9
10200 7.93
12900 7.96
16300 7.97
20500 7.98

At no point does the impedance EVER drop below 6 ohms. Hardly a short
as you claim. Let's even look at DC, where the inductor impedance is
0 and the capacitor impedance is infinite: the imepdance of the total
cricuit is that of the woofer. At infinite frequency, where the
impedance of the capacitor is 0 and the inductor is infinite, the
impedance is that of the tweeter.