Woofer Resonance Notch Filters

[dirkwright]

So, I bought this super simulation software called Basta! I'm not advertising it, but I noticed while playing around with various woofers is that if I put a series RLC notch filter at the resonance point (higher one in a vented enclosure), the phase shift in that frequency range becomes much more mild. It doesn't change the SPL response, but I would like to ask if this is something that is worth doing? I can't tell from Basta! if it affects group delay or impulse response. I also do not know if any of these changes are audible. I've been wondering about this for a while and I'm glad I was able to finally model it.

In the attaches images, the phase response is the dashed line in the upper chart. The first one is with a notch filter tuned to the upper resonance, the second one is without a notch filter. In the lower charts, the dashed line is the impedance curve, the dotted line is the vent response, the solid blue line is the driver response and the thick black line is the system response. The two lines at the very bottom of the bottom charts are room gain to the left and baffle step loss to the right.

[jerome69]

Only listening can tell.

I done it with my open baffle, the sound became better. The impedance curve became better, less load for the amplifier and flattened the bass response.
Big passive OB project

Troëls Gravesen have done the same thing with 2 of its designs, the sound seems to be better.
PRELUDE
Jenzen-NEXT

[sreten]

Hi,

It only makes a difference if the amplifier has high output impedance.

A more succinct approach is to wire a say 33R 10w or 20W resistor
in parallel with the speakers, it will reduce phase shift for both the
vented peaks and typical midrange peaks, and flatten impedance.

rgds, sreten.

[dirkwright]

Originally Posted by sreten

Hi,

It only makes a difference if the amplifier has high output impedance.

A more succinct approach is to wire a say 33R 10w or 20W resistor
in parallel with the speakers, it will reduce phase shift for both the
vented peaks and typical midrange peaks, and flatten impedance.

rgds, sreten.

OK. Thanks .

I tried the 33R resistor and it does lower the impedance peaks and phase shift a little bit but not nearly as much as a series RLC notch filter tuned to one of the impedance peaks, according to Basta! anyway....

Otherwise, I can't see any effect from it in Basta! on things like cone excursion, amplifier voltage, maximum SPL or other parameters.

[dirkwright]

Originally Posted by jerome69

Only listening can tell.

I done it with my open baffle, the sound became better. The impedance curve became better, less load for the amplifier and flattened the bass response.
Big passive OB project

Troëls Gravesen have done the same thing with 2 of its designs, the sound seems to be better.
PRELUDE
Jenzen-NEXT

Thanks. I think I'll give this a try and see if I can hear a difference. It's nice to know that others have done something similar. I know I'm not completely crazy that way.

[dirkwright]

One bad thing I'm seeing is that the typical notch filter calculator makes the R value so low that it brings down the impedance too much. In Basta! I'm raising the R value of the RLC filter to keep the impedance above 3 or 4 ohms. But oh what a nice flat phase curve! I just need an amp that's good for 25 watts into 2 or 3 ohms...

[sreten]

Hi,

The phase of the loudspeakers response does not change one iota.

Flat phase regarding the drive current is fairly meaningless and
not worth chasing. The cost of the parts is typically prohibitive.

High phase angles with high impedance is not an issue for amplifiers.
High phase angles with low impedance is a typical amplifier killer.

Your not fixing anything. Your playing with the wrong things in Basta! IMO.

rgds, sreten.

[badman]

Originally Posted by sreten

Hi,

It only makes a difference if the amplifier has high output impedance.

A more succinct approach is to wire a say 33R 10w or 20W resistor
in parallel with the speakers, it will reduce phase shift for both the
vented peaks and typical midrange peaks, and flatten impedance.

rgds, sreten.

By midrange peak- I assume you mean the peaking from the M/H XO?

[sreten]

Originally Posted by badman

By midrange peak- I assume you mean the peaking from the M/H XO?

Hi, Yes. On paper you can make the impedance look more benign, rgds, sreten.

[dirkwright]

Adding a resistor in parallel with the woofer didn't do anything to the midrange SPL response according to my simulations in Basta! It only slightly lowered the resonance peaks. You presumably already have an RLC resonance notch filter on the midrange and tweeter, so why add a resistor there? Doesn't make sense to me.