Home » Sponsored » Pi Speakers » Reverse attenuation and HF comp. networks, active X-over
Reverse attenuation and HF comp. networks, active X-over [message #36550] Mon, 20 May 2002 21:30 Go to next message
Chris R. is currently offline  Chris R.
Messages: 82
Registered: May 2009
Viscount
Hi Wayne,

I'm about 3/4 way done building an active crossover for bi-amping
the 3677's I got from JBL. What I began thinking about is the HF
comp network. Since it ends up being a bypassed voltage divider,
I was trying to figure out what resistor values to use for a given
compensation. If I need ~10dB of relative boost at 20KHz, starting
above say 12KHz, it seems I'd need a divider with 10dB attenuation,
then bypassed for a 3dB point around where I want the boost to
kick in. Two questions. Using 10K ohms as the series resistor,
what value goes to ground? If I'm feeding the output to the high
pass filter, does it need another op-amp buffer between the HF comp
and the HP filter?
Thx, Chris

Re: Reverse attenuation and HF comp. networks, active X-over [message #36552 is a reply to message #36550] Mon, 20 May 2002 22:34 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)
The formula is dB = 20 log V1/V2, so you'll need approximately 1/3rd value shunt resistance for a 10dB attenuator. The thing is, you're doing more than just padding, so a resistive voltage divider isn't all there is to it. The crossover provides the traditional voltage splitter function and it also provides some response shaping. The HF filter should be slightly underdamped. Traditional 6dB/octave augmentation is provided for mass rolloff compensation, and the two transfer functions form a response curve that's flat up to about 4kHz, with 6dB/octave augmentation above that. I'd model your circuit with Spice, to verify it provides a transfer function that looks like this:


Re: Reverse attenuation and HF comp. networks, active X-over [message #36574 is a reply to message #36552] Wed, 22 May 2002 21:17 Go to previous messageGo to next message
Chris R is currently offline  Chris R
Messages: 133
Registered: May 2009
Master
Hi Wayne,

I haven't decided if you're nuts or you deserve a medal :^).
I posted the question pretty late PST and you still responded same
night in a later time zone. Anyway, I'm not sure I understand the
last comment about generating a shelving response before augmentation.
I'm assuming that what we are attempting to do is produce a filter
that has a mirror response to the published freq. response curves
for the horns. To me, that means I need flat up to ~12K, then
+6dB/octave (or more) on up. Am I missing something?

Thx, Chris

Re: Reverse attenuation and HF comp. networks, active X-over [message #36576 is a reply to message #36574] Wed, 22 May 2002 21:49 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

This is what we want:

Tweeter output from π passive crossover

Re: Reverse attenuation and HF comp. networks, active X-over [message #36598 is a reply to message #36576] Thu, 23 May 2002 16:50 Go to previous messageGo to next message
Chris R. is currently offline  Chris R.
Messages: 82
Registered: May 2009
Viscount
Hi Wayne,

Maybe I wasn't clear. What I'm using to divide the frequencies
are 18dB VCVS (Sallen-Key) filter stages. To control horn attenuation,
I'm putting a balance pot circuit between the inputs for
high/low (verses left/right). I was planning on putting the
HF comp between the output of the balance buffer and the input
of the HP filter. But now you're going to make me install spice
or write a perl script to verify what you are saying. Seems to
me that at freqs below 1/(2*Pi*Rs*C) (there's that Pi thing again :^),
the divider will work more like a divider because Xc will be > Rs.
Above that freq, the Xc will be less than Rs, hence provide HF
boost. It looks a lot like the one used in your x-overs. Wonder
where I got the idea. )^:

Rs
acsii art time: --\/\/\/\/\------
|___||_____| |
|| |
C #
# Rp
#
#
|

I got the filter circuits from an old Walter Jung OpAmp Apps book
from when I was in school. Gee, I know the formulas are close,
but I'll have to look them up AGAIN!

Thx, Chris


Re: Reverse attenuation and HF comp. networks, active X-over [message #36604 is a reply to message #36598] Thu, 23 May 2002 23:01 Go to previous message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

Adjust the Q of the high-pass filter, which would then allow you to adjust peaking slightly above 0.707, to around 1.0. Add a voltage divider with bypass capacitance. Then you'll be able to set the filter Q for a flat shelf followed by rising response.

Previous Topic: Mr. Parham
Next Topic: Lathe turned horn for PSD2002
Goto Forum:
  


Current Time: Fri Mar 29 07:28:06 CDT 2024

Sponsoring Organizations

DIY Audio Projects
DIY Audio Projects
OddWatt Audio
OddWatt Audio
Pi Speakers
Pi Speakers
Prosound Shootout
Prosound Shootout
Smith & Larson Audio
Smith & Larson Audio
Tubes For Amps
TubesForAmps.com

Lone Star Audiofest