resistor choice for attenuation

[Wolf]

I asked you twice if this was in/for a xover, and was the Q of the filter, with no reply, and now I see you are referring to something completely different. I see now you are using a 3rd order acoustic HP of the driver in a sealed box + cap that has been added.

That is completely different!

Here I was thinking you were using BW equations to get a 12dB rolloff for a woofer xover LP point, and then figuring the point in to add a 3rd cap to make it 3rd Bessel rolloff, and using the T/S to get there.

Now we know why I didn't understand your process, even though some of the terms were apparently old-jargon and unfamiliar to me.

Now- to address the 3rd order acoustic HP combination; if the sub has a 12dB rolloff, then it will indeed make it easier to blend. The one-order acoustic difference allows phase to be easier aligned.
Room issues aside- the phase still shouldn't have to be aligned at a single point to be well blended. If you can hear the difference between 94 Hz and 95Hz, and identify them on the fly without them being adjacently heard, I applaud you. I still don't see the need for your insistance on being at one certain frequency for 'best results'.

Later,
Wolf

[moron#99]

I asked you twice if this was in/for a xover, and was the Q of the filter, with no reply, and now I see you are referring to something completely different. I see now you are using a 3rd order acoustic HP of the driver in a sealed box + cap that has been added.

That is completely different!

Here I was thinking you were using BW equations to get a 12dB rolloff for a woofer xover LP point, and then figuring the point in to add a 3rd cap to make it 3rd Bessel rolloff, and using the T/S to get there.

Now we know why I didn't understand your process, even though some of the terms were apparently old-jargon and unfamiliar to me.

Now- to address the 3rd order acoustic HP combination; if the sub has a 12dB rolloff, then it will indeed make it easier to blend. The one-order acoustic difference allows phase to be easier aligned.
Room issues aside- the phase still shouldn't have to be aligned at a single point to be well blended. If you can hear the difference between 94 Hz and 95Hz, and identify them on the fly without them being adjacently heard, I applaud you. I still don't see the need for your insistance on being at one certain frequency for 'best results'.

Later,
Wolf

i'm confused. we are still talking apples and oranges i think.

let me try again.
I will model a speaker as a perfect voice coil with no resistance. It has infinite xmax with perfectly uniform magnetic field. It is mounted to a perfectly rigid piston with no mass.

In order to make it real world I add a capacitor equal to the spider and suspension. I add a resistor equal to the copper windings. I add an inductor to represent the inertia of the cone.

If I mount the speaker in a sealed box then the trapped air changes the value of capacitor and inductor. It is still just one capacitor, one inductor, and one resistance but the values change depending upon the size of the box.

the physical electronic capacitor that I am adding interacts with the speaker and changes its behavior. It is not a crossover. It is a modification to the speaker itself. It interacts with the above and creates a third order bessel high pass.

I do not use high pass crossovers on my woofers or mid-bass. I modify the driver to be a third order bessel and then design a low pass filter to match it. That is why I sometimes need to pad a woofer. My choice of a mid-bass driver will only have one valid frequency where it can be crossed over into a woofer.

[Wolf]

the physical electronic capacitor that I am adding interacts with the speaker and changes its behavior. It is not a crossover. It is a modification to the speaker itself. It interacts with the above and creates a third order bessel high pass.

My choice of a mid-bass driver will only have one valid frequency where it can be crossed over into a woofer.

Nope, we agree and are talking of the same thing. You just basically restated what I said above.

What we are disagreeing on is that it's a range of freq's that the midbass and woofer 'share', even though one frequency is the point in which they xover.

Trust me, I fully understand what you are doing right now.

We also disagree in the point of having to pad a woofer. That's never a good idea for this type of application.

Later,
Wolf

[moron#99]

the woofer will have a low pass filter as required by the midbass driver. The f3 of woofer lowpass typically a overlaps f3 of midbass in a bessel system.

The F3 of the midbass driver can not be changed without abandoning the third order bessel transfer equation for acoustic output.

if you are concluding that these things are moveable then I don't think you are understanding.

maybe i should combine this with another of my ideas an mail it off to the speaker thing-a-ma-bob. I have been dying to build a cabinet-less port out of PVC pipe for months now. Vac=Vap. hehe.
but I also want to build a pvc cabinet with threaded plug so I can adjust the cabinet volume. that would lend itself well to the butterworth transform.

[Wolf]

the woofer will have a low pass filter as required by the midbass driver. The f3 of woofer lowpass typically a overlaps f3 of midbass in a bessel system.

The F3 of the midbass driver can not be changed without abandoning the third order bessel transfer equation for acoustic output.

if you are concluding that these things are moveable then I don't think you are understanding.

Like I said- room effects will swamp this stuff. You will likely have room modes that change the effect of the summation to not be at the same point your equations tell you it is.

If you are in fact not measuring the acoustic output somehow to verify that you are hitting this target exactly on the mark, nor are you measuring where the reflections will not swamp the effect of this transition (likely outside), then you cannot for a fact absolutely verify that the F3 of the woofer and midbass meet at exactly the same single and unique ideal frequency.

I'm sorry if you don't understand what I'm saying, but I find it absolutely hard to believe that you can target and verify that you are truly hitting this very important frequency without a true semi-anechoic measurement of acoustic output.

Equations are not enough when you have to be this picky. I'm not sure I understand why you say it has to be at this important frequency, other than 'the equations tell you it's so'.

Later,
Wolf

[moron#99]

here is a site with pole/zero plots
http://cnyack.homestead.com/files/af...lbuttpoles.htm

the distance of a pole or zero from center does not matter as it can be easily changed. the angle, however, is immovable without causing fundamental changes in circuit bevavior. this applet will allow you to click through various nth order filters. as long as the higher order filter includes the sames poles at the same angles as lower filter then you can simply add paerts to create it. if the poles are at different angles then the result will never be perfect and your only choice is to run a conglomerate of narrow bandpass filters. that operate between pole/zero frequencies.

For example, if you desire a fourth order bessel then you could configure the second order speaker with Qtc around 0.5 and add two poles.

room acoustics do not stop or start driver ringing. your frequency charts are a steady state test. they do not portray how a system will behave at a point of discontinuity - such as a drumstick striking a skin.

[Wolf]

Then how are you verifying the match? If you're being that picky, it has to be right, and you have to verify it somehow. Equations will not do this.

That is what I'm having issues with. How are you absolutely positive you are getting what you say you are?

Later,
Wolf

[moron#99]

i have already built them, put them in a lab and tested. they work perfectly, the solution set is scaleable, and the result reproduces on all platforms attempted. that was a long time ago though. back before pc's. lab equipment was not cheap and it was not possible after leaving my job as audio engineer to access test equipment again. These days it seems like all you need is a mic and a laptop. that's why i want to build the pvc pipe speakers. i can use the threaded insert to "dail a Q".

at any rate, the pole zero plots give you a quick way of identifying what is possible. you still have to sit down and grind out the math. for example if you take a pair of overnight sensations and plot their pole zeros then you could start filpping through bessels looking for a match. You would not find one until it was unreasonably complicated and expensive. Hence, OS would not be deemed a correctable speaker. the pole zero would have saved you days of trying to divide one polynomial into several others. with the pole zero you can see that it's not possible within 10 minutes.