diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Subwoofers (http://www.diyaudio.com/forums/subwoofers/)
-   -   How to Cram the Wrong Driver into a Tapped Horn (http://www.diyaudio.com/forums/subwoofers/224097-how-cram-wrong-driver-into-tapped-horn.html)

Patrick Bateman 22nd November 2012 04:36 PM

How to Cram the Wrong Driver into a Tapped Horn
 
In this thread, I am going to explore how to cram the wrong driver into a tapped horn.

Here are a few reasons you might consider doing this:

1) the 'ideal' woofer for a tapped horn is sometime unobtanium, or impractical. http://www.cadaudio.dk/12ps100.jpg
For instance, to achieve an F3 of 35hz with a tapped horn, you'll want a driver with an FS of about (35hz * 1.414), and as much xmax as you can possibly find. These parameters will lead you to woofers like the B&C 12PS100. But what if you can't afford the space? Or what if you don't need the high power handling of the B&C?
2) A lot of the best subwoofer drivers are physically beefy. They have sturdy metal or composite cones, heavy spiders, and gigantic surrounds for high excursion. Unfortunately, all of these features drive up the QMS of the driver. Due to this, there are a lot of great sub drivers that won't work in a 'conventional' tapped horn, because the QTS is too high. Or they'll work, but they require a box that's the size of a Volkswagen.
3) Perhaps you simply have a favorite woofer gathering dust in the garage, and you'd like to put it in a TH, but hornresp is giving you results that appear unacceptable.


If any of this applies to you, read on...

Patrick Bateman 22nd November 2012 04:46 PM

Before we cram the wrong woofer into a tapped horn, let's start by putting the woofer in a 'proper' tapped horn.

http://a248.e.akamai.net/pix.crutchf...-o_sideup.jpeg

Our driver for this experiment will be the Alpine SWR-843D. I can't say enough good things about this driver:
1) With 14mm of xmax, this driver has as much displacement as most of the prosound twelves that people use for tapped horns
2) it's BL curve is gorgeous, and similar to the now-discontinued CSS Trio 8 and Exodus Audio Anarchy
3) Because it's a mass market driver, it's stupid cheap. About $100. And available worldwide, from easily five hundred different retailers.
4) The build quality is fantastic
5) it's inductance is comparable to Scan Speak and Seas Excel, at just 1.23mh! That's much lower than the TC Sounds drivers that everybody loves so much*
5) It looks cool!



* I <3 TC Sounds too, and own four of their subs, but this Alpine is crazy good for a mass market driver

Jack Caldwell 22nd November 2012 05:07 PM

Interesting experiment! Needless to say I'll be following this thread! Where can I find the T/S specs for this driver? I looked at the Alpine website and those are nowhere to be found - are they hidden somewhere? Thanks.

Patrick Bateman 22nd November 2012 05:31 PM

When I design a tapped horn, I generally start by modeling the woofer in a vented box. For instance, if a vented calculator tells you to tune the vented box to 28hz, then the same woofer will generally work in a tapped horn tuned to 28hz.

Most of the time I find that this results in a tuning of about 0.707 multiplied by the FS of the woofer. For instance, the FS of the Alpine SWR-843D is 40hz, and will yield a tapped horn tuning of 28hz if you use this formula.

Another way of looking at it is to look for a woofer with an FS that's about 1.414 multiplied by the tuning of the tapped horn. For instance, if you want to tune your TH to 28hz, look for a woofer with an FS of 40hz. (28 * 1.414)

https://lh4.googleusercontent.com/-j...ksgiving01.jpg
Here's the SPL response of a vented box with a pair of the Alpines.*
It has an F3 of 24hz, an FB of 28hz (see above), and a box size of 67 liters. This sim is with 700 watts, which is the RMS power rating for two of the Alpines.

https://lh4.googleusercontent.com/-w...ksgiving02.jpg
https://lh3.googleusercontent.com/-q...ksgiving03.jpg
Here's the excursion and the group delay of the Alpine in the QB3 vented box. Note that we're exceeding xmax with the 700 watts that the RMS rating of the subs.

https://lh5.googleusercontent.com/-O...ksgiving07.jpg
Here's the SPL response of the Alpine in a properly sized tapped horn. The tapped horn is 134 liters, or twice the size of the QB3 vented box.

One thing that you may find interesting is that the tapped horn doesn't play any lower than the vented box. This is Hoffman's Iron Law at work; while the tapped horn is larger, the tuning is the same, so the efficiency gain is *above* the F3, not below it.

Due to Hoffman's Iron Law, the only way that we can make this tapped horn play lower is by tuning it lower, and trading efficiency at higher frequencies for efficiency at lower frequencies.

I think this is important to note; there is nothing magical about tapped horns, and a carefully engineered vented box will have comparable SPL per cubic foot. No way to get around Hoffman's Iron Law.

https://lh3.googleusercontent.com/-3...ksgiving08.jpg
https://lh6.googleusercontent.com/-O...ksgiving09.jpg
Here's a pair of sims comparing the excursion and group delay of the tapped horns to a vented box. Here's one place the TH shines; it has reduced excursion at multiple frequencies, as the TH resonates above the passband, whereas the helmholtz resonance of the VB occurs at only one frequency. The group delay of the TH at resonance is also lower, because the rear wave of a TH is 90 degrees out of phase, while the rear wave of a VB is 180 degrees out of phase. (That's why the group delay at resonance of the VB is so much higher than the TH.)


In summary - with an identical tuning, the F3 of the two boxes is similar. The TH is more efficient above the tuning frequency, but the box is also twice as big!

In my next post, I'll explore how to use the Alpine in a box that's not 'correct'





* in all of these sims I've simulated the Alpine using two woofers. This is because I generally prefer push-pull loading, as it gives you a 'free' reduction in 2nd harmonic. IMHO, I'd generally prefer to use a pair of eights over a single twelve due to the reduction in 2nd harmonic.

Patrick Bateman 22nd November 2012 05:59 PM

Quote:

Originally Posted by Jack Caldwell (Post 3252073)
Interesting experiment! Needless to say I'll be following this thread! Where can I find the T/S specs for this driver? I looked at the Alpine website and those are nowhere to be found - are they hidden somewhere? Thanks.

There's a monster of a thread on diyma with some info on the driver; but the basic t/s specs are here:

http://support.alpine-usa.com/produc..._S06212011.pdf

If I'm not mistaken, the best features of the driver are common across the line. (Shorting rings, low price, a BL curve that looks like Alpine stole a page from Adire Audio's XBL technology.)

I believe the reason that the driver is so deep is that it's likely hiding a dual-gap motor, but I'm not sure if Alpine has copped to it. (A few years back Pioneer and Orion were selling dual gap drivers too.)

Patrick Bateman 22nd November 2012 07:16 PM

In post number four, I demonstrated that we can use vented box math to build a tapped horn.

In this post, I will do the same, but with a twist. This box will use the same woofer, but at a higher tuning frequency.

The reasons we might do this are illustrated in post one of this thread.

Some of you many be familiar with QB5 alignments. A QB5 alignment is a clever way to combine a 'boomy' vented alignment with a filter. Without the filter, the 'boominess' would make the box nearly unlistenable. But when you add the filter, you get a long list of advantages:

1) Power handling goes up in a huge way. This is because the high pass filter reduces excursion. But this is different than a typical subsonic filter. Typically, a subsonic filter is used to reduce excursion BELOW the pass band. In a QB5 alignment the filter is set much higher. Because it's set higher, it changes the response shape, and improves power handling to a greater degree than if it was set lower in frequency.
2) Distortion goes down. Generally when you reduce excursion, you reduce distortion.
3) Box size is generally smaller than a QB3. As the box is designed to be tuned higher than QB3, box size is smaller too.


You put all of these things together, and you can create some compelling alternatives than the 'typical' boxes we see here, which tend to use big prosound drivers. For instance, a pair of the Alpine eights in a push pull configuration need a box that's comparable to a B&C twelve. But the push-pull mouting reduces second harmonic, the Alpines have more xmax than the B&C, and the Alpines are more readily available. The QB5 alignment reduces distortion even further.

So let's investigate this further:

https://lh6.googleusercontent.com/-i...ksgiving10.jpg
Here's a hornresp simulation of two of the Alpines in a QB3 and a QB5 box. Both boxes have two Alpine SWR-843Ds. Both boxes are getting 700 watts into 8ohms (74.83V) The QB3 box is 67 liters and tuned to 28hz. The QB5 box is 46.5 liters and tuned to 40.6hz.

We see the following:

1) The F3 has gone up a lot - about 10hz. From 23hz to 33hz. But again, this can be an advantage in some situations. For instance, let's say you have enough space for a subwoofer that plays to 35hz. Instead of using a 6.5" woofer, the QB5 alignment allows you to use an eight! The key is that the filter flattens out the 'boom'
2) The peak has gone from a couple dB to four dB. Getting kinda boomy, but we'll fix that...

https://lh5.googleusercontent.com/-v...ksgiving11.jpg
Here's my favorite graph. The excursion. We see that with 700 watts into 8ohms, the QB5 sub is almost within it's mechanical limits, whereas the QB3 is blowing through them. And the addition of the highpass filter, which is fundamental to the design, will only improve this further.


https://lh4.googleusercontent.com/-N...ksgiving12.jpg

If I'm not mistaken, there isn't a way to add a subsonic filter in hornresp. Due to this, I had to export the data to excel, then sim the filter there.

The graph above shows the QB3 box versus the QB5 box, both filtered and unfiltered.
The QB3 response is green.
The unfiltered QB5 response is blue.
The filtered QB5 response is red.


We see the following:

1) the subsonic filter reduces the 'boom' in the QB5 box by about a third. There's still a peak, but it's just 2.5dB. IMHO, this would be barely noticeable at such low frequencies; your room will add peaks and dips that are a lot bigger than 2.5dB
2) The F3 of the QB3 box is much much lower than the QB5 box - about 15hz. But, again, this is by design. QB5 boxes aren't a method of extracting every last bit of bass out of a woofer; QB5 boxes are an alignment that let you use a woofer which would typically be too large for a give box. Or a method of using a larger woofer in a given box. You might argue that QB5 boxes are almost the polar opposite of the typical huge DIY sub.

kevinkr 22nd November 2012 07:36 PM

I'm following too, often find your threads thought provoking Patrick.. Keep 'em coming.. :D

Cordraconis 23rd November 2012 08:09 AM

Quote:

Originally Posted by Patrick Bateman (Post 3252215)

Some of you many be familiar with QB5 alignments. A QB5 alignment is a clever way to combine a 'boomy' vented alignment with a filter. Without the filter, the 'boominess' would make the box nearly unlistenable. But when you add the filter, you get a long list of advantages:

If I'm not mistaken, there isn't a way to add a subsonic filter in hornresp. Due to this, I had to export the data to excel, then sim the filter there.

Hello Mr. Bateman,

I tried to send a PM but it doesn't seem to get trough, so I'll post some of it here.


HornResp indeed doesn't have filtering capabilities (yet), aldough they have been suggested recently. ;)

As you'll probably have noticed, I'm also playing around with a (fullrange) QB5 alignment, but having to re-learn Akabak after all those years doesn't help my speed ... How good are you with Akabak? I think I have found a way to design custom (variable Q) passive filters according to the alignment tables, but they have a slight recession above the tuning freq. Now I have noticed that your design has a slight rising ... :)

I feel Akabak is giving me a hard time and maybe we can give each other a push in the right direction. (Aldough I know a passive highpass in a sub would be massive!)

Thanks for any PM (so as not to pollute your thread),

Cordraconis

Patrick Bateman 24th November 2012 07:44 PM

Quote:

Originally Posted by Cordraconis (Post 3252784)
Hello Mr. Bateman,

I tried to send a PM but it doesn't seem to get trough, so I'll post some of it here.


HornResp indeed doesn't have filtering capabilities (yet), aldough they have been suggested recently. ;)

As you'll probably have noticed, I'm also playing around with a (fullrange) QB5 alignment, but having to re-learn Akabak after all those years doesn't help my speed ... How good are you with Akabak? I think I have found a way to design custom (variable Q) passive filters according to the alignment tables, but they have a slight recession above the tuning freq. Now I have noticed that your design has a slight rising ... :)

I feel Akabak is giving me a hard time and maybe we can give each other a push in the right direction. (Aldough I know a passive highpass in a sub would be massive!)

Thanks for any PM (so as not to pollute your thread),

Cordraconis

Cool! I'll follow up with you.

Patrick Bateman 24th November 2012 07:53 PM

I hate to be a 'wet blanket', but I'll likely need to put this project on hold for a bit. Here are the problems I'm running in to with this design:

1) The main reason that I'm trying to build a sub is that I'm trying to design a sub for my car that I can put on the back seat, without ruining the leather seats
2) My original plan was to build a very lightweight box, and that plan worked pretty well. (check out 'tapped horn for the lazy and impatient' for details/plans for that box
3) Unfortunately, the box from the aforementioned thread was designed as part of a larger system, and I ran out and bought a new car two weeks ago

so...

I'm trying to come up with a sub that won't wreck my seats, but also one that I can basically just plug into the existing stereo.

My plan to use the Alpine in a QB5-style alignment, but a tapped horn, isn't working for a few reasons:

1) the 'ripple' in the tapped horn is much larger than the ripple in a vented box, so the subsonic filter doesn't work as well to reduce the ripple
2) the much bigger problem is that the ripple in the tapped horn is much higher in frequency than in a vented QB5. For instance, a vented QB5 box might have an F3 of 40hz with ripple at 45hz. Due to the F3 and the ripple being close together, the subsonic filter smooths it out nicely.
In the tapped horn, this is not the case; when trying the same trick with a TH, the gap is much wider. For instance, the F3 of the TH might be 40hz, but the ripple is at 60hz! It basically calls for a unconventional filter, or perhaps a parametric EQ combined with a high pass. Way more complex.

I think part of the reason why the QB5-style alignment is so challenging with the Alpine is that it's QTS is so high. With a QTS of 0.55, as you get close to the driver's resonance the ripple gets quite extreme.

I'm not convinced that QB5-style alignments are impossible with a tapped horn; but I *am* convinced that the alignment is better suited to a woofer with a QTS that's much lower.

http://www.soundsplinter.com/images/...ight-image.jpg
http://www.soundsplinter.com/images/...-right-sm1.jpg
I have some TC Sound TC2+ woofers sitting in the garage, might dust them off and try it...


All times are GMT. The time now is 05:20 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2