Tweeter Crossover

[John Busenitz]

I am helping a friend build a 2-way loudspeaker with a cabinet,
8" woofer, and a premade crossover he bought (we know we need
to construct an impedance compensation network). The "problem"
stems from the fact that the crossover is 1250 Hz 2nd order,
which is on the low side for most tweeters.

I've heard that the tweeter Fs should be at most an octave below
the crossover frequency. I've seen tweeters with Fs ranging from
650 Hz to 900 Hz. I would pick the one with the lowest Fs, but
for example, the 650 Hz Fs tweeter has a published "frequency
response" (no limits) of 1500-20000 Hz, while one with a higher
Fs (880 Hz) has a frequency response of 1-20 kHz, which would
seem to be a better response for the system. Which tweeter would
be a better match?

I guess it depends on which spec is the more reliable and demands
more weight: Fs or "response"? I would guess that the Fs would
be a better spec to go on, but I'm not sure. Any advice?

Thanks

_____________________________________________________________
John Busenitz buse...@ecn.purdue.edu
P.U. ECE http://cernan.ecn.purdue.edu/~busenitz
Disclaimer: My statements do not represent Purdue University.

[Dave Dal Farra]

In article <Pine.SOL.3.91.970216...@shay.ecn.purdue.edu>,
John Busenitz <buse...@ecn.purdue.edu> wrote:

>I am helping a friend build a 2-way loudspeaker with a cabinet,
>8" woofer, and a premade crossover he bought (we know we need
>to construct an impedance compensation network). The "problem"
>stems from the fact that the crossover is 1250 Hz 2nd order,
>which is on the low side for most tweeters.
>
>I've heard that the tweeter Fs should be at most an octave below
>the crossover frequency. I've seen tweeters with Fs ranging from
>650 Hz to 900 Hz. I would pick the one with the lowest Fs, but
>for example, the 650 Hz Fs tweeter has a published "frequency
>response" (no limits) of 1500-20000 Hz, while one with a higher
>Fs (880 Hz) has a frequency response of 1-20 kHz, which would
>seem to be a better response for the system. Which tweeter would
>be a better match?
>
>I guess it depends on which spec is the more reliable and demands
>more weight: Fs or "response"? I would guess that the Fs would
>be a better spec to go on, but I'm not sure. Any advice?

Hi John,

I'd opt for a tweeter with acceptable power handling. Neither "fs" or
response will tell you that. Power handling is thermally limited, or
excursion limited. Thermal limits are based upon magnet design, voice
coil wire gauge and other design aspects too complicated to discuss here.
Excursion limits are based upon Sd times Xmax, the swept volume.

Thermal limits must be respected so the driver doesn't burn out, excursion
limits in a tweeter so that distortion remains low (its hard to bust a
tweeter through over excursions). The manufacturer tries to cover them
all by specifying a maximum continuous power (thermal) and a maximum music
power with a certain xover (max thermal and excursion).

A metal dome with a dissipative voice coil former will have good
thermal power handling. Crossing over so low, I'd also
gun for one with decent excursion handling. A 1" Vifa metal
dome might work, the ones with the back volume, but that零 still one
awfully low crossover point.

Also realize that a second order crossover will (typically)
give a 3 dB bump in the crossover range if the drivers are in phase or a 3
dB null if they're connected out of phase.

If I were in your situation, I'd connect the drivers in phase to start,
then play with the tweeter compensation network to see if you can bring
down the peak. This way you could get a flattish response with less
voltage delivered to the tweeter around crossover, giving better power
handling. The drawback is that you will now not be nulling out the
impedance rise at fs, but by using a ferrofluid version of the Vifa, the
resulting out of band peak should be small as the tweeter will have a
marginal impedance rise at fs. If you're still worried, you could
parallel two compensation networks: one to work at fs, and one to kick in
around crossover. You can tweak using a signal generator and voltmeter to
monitor the xover behaviour electrically (to help you design the
compensation networks), and your ears to get an indication what零
happening acoustically.

Hope this made sense to you.

--
Dave Dal Farra (gpz...@nortel.ca) "I was moving so fast
Nortel Technology I started using Him
Audio Design Group as a braking marker"
GPz750/FatJ1200

[Dave Dal Farra]

>>I've heard that the tweeter Fs should be at most an octave below
>>the crossover frequency. I've seen tweeters with Fs ranging from
>>650 Hz to 900 Hz. I would pick the one with the lowest Fs, but
>>for example, the 650 Hz Fs tweeter has a published "frequency
>>response" (no limits) of 1500-20000 Hz, while one with a higher
>>Fs (880 Hz) has a frequency response of 1-20 kHz, which would
>>seem to be a better response for the system. Which tweeter would
>>be a better match?
>>
>>I guess it depends on which spec is the more reliable and demands
>>more weight: Fs or "response"? I would guess that the Fs would
>>be a better spec to go on, but I'm not sure. Any advice?

Hi John,

In retrospect, here's a bit more information for you that will more
directly answer your question.

For a given piston, if response is flat, excursion increases 4 times as
frequency is halved.

So, for a tweeter that rolls off at 12 dB/octave, the excursion is
constant for a constant input through roll off. Take 2 tweeters. One
has fs1, the other fs2, where fs1<fs2, and all else remains equal. Both
are flat above their fs, and both roll off at 12 dB/octave. Tweeter fs1
will show more excursion below fs2 than tweeter 2, for the same input.
So, to control excursion with a given xover, it could make sense to use
the tweeter with the higher fs. This is only true if the compensation
network makes the impedance flat. If it doesn't and there is an impedance
rise, the impedance rise would create poorer power handling as more
voltage is delivered to the impedance bump.

There's one last thing to consider. Since your xover is so low, I'd use
the tweeter with the lower fs anyways. Around fs, tweeter acoustic phase
is going through some serious changes, even though the magnitude looks
flat. Crossing over with a second order near fs will bring trouble in the
form of a very non flat combined (woofer/tweeter) frequency response.

More confusion for your morning coffee.

[Guenter A. Scholz]

talking about tweeter x-overs reminds me that I recently heard a
speaker by scientific fidelity which had a 3dB down frequency of 1kHz. I
could not believe it at first. But the low x=over did make sense from a
dispersion viewpoint and integration to the mid-driver (no lobing). Is this
low a x-over truly `pushing' it for a tweeter or can some indeed perform well
to this low (and necessarily below) a frequency. Just wondering

-regards, guenter

[Dave Dal Farra]

Hi Guenter:

Its simple physics. Just as in a sub woofer where you need swept volume,
a tweeter digging down low needs swept volume for low distortion. I've
yet to see a commercial driver outside a decent sized ribbon that
adequately cover 1 K to 20 k. Most 1 " domes are designed for a second
order or better above 2 kHz.

As for lobing, you don't need to push a tweeter down to 1 kHz to achieve
decent polar response. As qualification, I'm a strong oponent to the
school of omni directional radiation. I believe controlled dispersion
provides the best sound in the home.