Very often when using a midrange in a 3-way, Fs impedance correction is required to get the proper behavior of the driver. These can be very expensive, given how large the filter values must be. I'm wondering if there is anything we can do with the box to tame the resonant peak? Something like an aperiodic damper, or? Or possibly Helmholtz arrays?
I thought I saw a design where the midrange isolation chamber was "split" with an aperiodic damper. Unsure if "tame the Fs" was the design goal.
Well stuffed ~Fs/0.71 TL is my way and doesn't need to be very large if using a high Qt driver (~0.7 with the vintage drivers I used), but as motor strength rises the density required can get to the point where I've had to just wrap/seal off the basket with really dense 3/4" carpet underlayment to get it 'close enough' for my ears.
Well, historically I normally used so called full range drivers, so fiddled with HR to see if my dim memory had failed me yet again, so using a MA Alpair 12, changed its Qt to 0.7 and its TL is < 1/2 its Vas Vs its stock 0.28 Qt that HR couldn't go high enough in density no matter how big it was, proving why I had to resort to so much trial n'error grief way back when I did some Altec 12" ~0.2 Qt mids (horn) woofers.
Regardless, I forgot to mention band stop filters.
Regardless, I forgot to mention band stop filters.
I see. So these are just tapered volumes? What's the math? I see 10:1 mentioned. I'm looking at using this for a SB17CAC and MW16TX.
Why is it desired to flatten the impedance of a sealed driver? Just thinking about ported two ways where generally the woofer always has two huge peaks either side of the tuning frequency, and they usually aren’t dealt to. Is this simply a matter of reducing current distortions? Or does the drivers output collapse at the impedance peak when sealed?
This is specific to a midrange and its interaction with the high pass of its crossover filter. There the Fs is often a real problem that prevents you from hitting your acoustic targets, and just simply puts out more energy (and excursion) when you don't want it too. But when your Fs is 35hz, or even 235hz, the values for an impedance trap are really big. So I was hoping to do it acoustically.
Yep, the pioneer's design mantra of acoustic solutions for acoustic problems is the way when wanting max efficiency.
We are flattnin gthe resonant peak so that the high pass XO looks into a flatter impedance.
The math is usually, the line has a large sealed volume, given how low given a mid.length is longer than if you were tuning to the Fs, especialy with 10:1. Damp lightly near the driver to very dense mear the terminus. Test by lowering the resonant peak or using GM’s click-test.
dave
The math is usually, the line has a large sealed volume, given how low given a mid.length is longer than if you were tuning to the Fs, especialy with 10:1. Damp lightly near the driver to very dense mear the terminus. Test by lowering the resonant peak or using GM’s click-test.
dave
To damp the Fs, must the line length have to be 1/4wl? That's really long for the drivers I'm looking at with 28hz-40hz Fs.
EDIT: dug up MJK's paper. With a 10:1 taper, there is a correction factor of .62, so that helps. But I'm still looking at 50-60" lines.
EDIT: dug up MJK's paper. With a 10:1 taper, there is a correction factor of .62, so that helps. But I'm still looking at 50-60" lines.
Printed Meta material: https://www.diyaudio.com/community/threads/3d-printed-metamaterials.360739/
The thread also contains a link to a relevant JAES publication.
The thread also contains a link to a relevant JAES publication.
Right; I never felt the need to taper since it's more about stuffing density. That said, I previously noted ~Fs/0.71, though don't know why since the ~Fs/0.31 I did the test sim is correct, so 28 Hz = ~36.75" axial length.
Yes i'm thinking about that option. I have a paper which uses an array of little Helmholtz resonators with a depth of 1/20 wl. I could print them pretty easily.
Isn't it the best way to use a midrange with higher effiency and then use a resistor network with a resistor in series and one parallel just to flatten completely the impedance of the driver?
You can put a smaller R in series and an appropriate R with lower value in parallel just to flatten the impedance of the driver. As long as impedance does not drop too low.
As I learned now from the simple resistor method of Esa Merilainen (website current drive info) many two or three way boxes profit from resistances in line with the speaker to bring distortion down at no cost.
Here is a german website for calculating this network.
https://lautsprechershop.de/tools/t_w_teiler.htm
You can put a smaller R in series and an appropriate R with lower value in parallel just to flatten the impedance of the driver. As long as impedance does not drop too low.
As I learned now from the simple resistor method of Esa Merilainen (website current drive info) many two or three way boxes profit from resistances in line with the speaker to bring distortion down at no cost.
Here is a german website for calculating this network.
https://lautsprechershop.de/tools/t_w_teiler.htm
Attachments
I never liked too much damping in loudspeakers or using Variovents (controlled damped leakage of the box) as a loudspeaker dominated by mechanical damping sound usually lifeless. Maybe this is less important for true midrange drivers than for the bass drivers in two way boxes.
However I would always use minimal passive filtering and just add some EQ or active filtering for bringing frequency response to the desired curve. Today you can get very cheap but good quality EQs. Most simple would be a tiny custom made DIY filter for putting between preamp and amp. Would be less expensive than the filtering at several hundred hertz done passively.
However I would always use minimal passive filtering and just add some EQ or active filtering for bringing frequency response to the desired curve. Today you can get very cheap but good quality EQs. Most simple would be a tiny custom made DIY filter for putting between preamp and amp. Would be less expensive than the filtering at several hundred hertz done passively.
Where do you live? What i do is closed box stuffed with a special sheeps wool, does dampen the resonance very well as well as the cavity resonances. I could send some wool.