Tapped Horns - Simulation VS Reality

[Patrick Bateman]

Originally Posted by Brian Steele

One could argue that a "tapped-horn" isn't a real bass horn - the mouth size is way too small.

Did you do a similar test using something like the Autotuba?

Yes, I measured my autoTuba clone in and out of the car. The thing that sucks is that the PC crashed, and I lost all the measurements. This was about four months ago.

About six weeks ago I built a tapped horn and a front loaded horn, both with the exact same footprint, and with the exact same woofers. THIS is what's required to really A/B the two enclosure types; you need to use the same woofers to do a fair comparison.

In this comparison, the tapped horn was about 12dB more sensitive in the passband. TWELVE DB!!!

That's just a crazy increase in gain.

In a nutshell, you can get the same output in the same foot print with an array of sealed boxes, but nothing comes close to the efficiency of a tapped horn if you DON'T have the luxury of using multiple woofers and lots of power.

The AutoTuba *did* have an increase in efficiency, but it was fairly high in frequency. Below the tuning frequency, the AutoTuba behaved like a simple sealed box.

There IS one big advantage to the front loaded horn, and that is the increase in impedance. This effect is NOT subtle - the 4ohm woofer used in the AutoTuba basically becomes an 8ohm woofer in a front loaded horn. Due to this increase in impedance, your power handling basically doubles (since there's less current in the voice coil.)

Also, I agree with you - neither the AutoTuba nor any of the tapped horn designs are truly horns. They're closer to a transmission line really. I believe this is why cabin gain is in full effect - they're not really horns.

[davygrvy]

Originally Posted by Patrick Bateman

The AutoTuba *did* have an increase in efficiency, but it was fairly high in frequency. Below the tuning frequency, the AutoTuba behaved like a simple sealed box.

+1! Owning a few AT10s myself, I completely agree.

There IS one big advantage to the front loaded horn, and that is the increase in impedance. This effect is NOT subtle - the 4ohm woofer used in the AutoTuba basically becomes an 8ohm woofer in a front loaded horn. Due to this increase in impedance, your power handling basically doubles (since there's less current in the voice coil.)

Don't forget about phase angle and it's effect on Power Factor. If one was to look (or calculate) the load as viewed by the amplifier in such as way as described by a technique called Equivalent Peak Dissipation Resistance you'd see that at 108Hz where the phase angle is the most around +63 degrees and 9 Ohms, the load as viewed by the amplifier is still just 4 Ohms. As I recall, but lost the pix, the impedance humps happen in three places for the AT: 40Hz, 80Hz and a small peak somewhere around 120Hz. As the peaks are still just oasis-es in a 4 ohm desert, the rating required for the amplifier is still the minutia.

[Patrick Bateman]

Originally Posted by davygrvy

As I recall, but lost the pix, the impedance humps happen in three places for the AT: 40Hz, 80Hz and a small peak somewhere around 120Hz. As the peaks are still just oasis-es in a 4 ohm desert, the rating required for the amplifier is still the minutia.

Yep.

One of the things that's frustrating about debating box alignments is that people assume that simulations are accurate. And I've found time and time again that simulations will only get you "in the ballpark."

And I've been guilty of this myself - I dismissed tapped horns a couple years ago, because my initial results were poor. What I didn't realize at the time was that there are some simple tricks that can be used to bump up your output.

• When you build a horn, measure the impedance curve. You'll be surprised by what you see - a big bump in impedance. This bump in impedance allows you to use woofers with reaaaaaaaaaly low impedance. For instance, my Triple8 would dip down to 1.2 ohms if it was sealed. Yet my impedance is THREE ohms in a tapped horn. (Low passed at 80hz btw)
  The difference between 1.2ohms and 3ohms doesn't seem like a lot, but that's HUGE. That change in impedance lowers the power consumption by over FIFTY percent. Just a crazy improvement.
  
  I wonder if a lot of people build horns without considering the change in impedance? Because doing that really robs your SPL potential. Obviously a 1.2ohm load would blow up most amps, but a 3ohm load is quite manageable.
• The other "trick" is to use a MASSIVE flare. When you look at the sims, a big flare doesn't make a huge difference. But in the "real world", the boxes I've built with a substantial flare sound a lot more dynamic, and just plain louder.

I wish I had some measurements or experiments to illustrate the second point. It was one of those things where the improvement was instantly noticeable, and I just stopped doing things "the old way." (IE, I will never build another loudspeaker that doesn't use a substantially flared port.)

Here's a measurement of the impedance of a few subs I have. The purple line is the Triple8 Tapped Horn, which would have a 1.2ohm load if it was sealed.

[Brian Steele]

Originally Posted by Patrick Bateman

Here's a measurement of the impedance of a few subs I have. The purple line is the Triple8 Tapped Horn, which would have a 1.2ohm load if it was sealed.

Bear in mind that an impedance curve for a subwoofer can tell us a lot more than just the resonant frequencies of the alignment. For example, if the impedance at the dips is substantially higher than that predicted by the sims, this could be caused by air leaks between the internal sections of the box.

[Patrick Bateman]

[font=verdana]

Originally Posted by Brian Steele

Bear in mind that an impedance curve for a subwoofer can tell us a lot more than just the resonant frequencies of the alignment. For example, if the impedance at the dips is substantially higher than that predicted by the sims, this could be caused by air leaks between the internal sections of the box.

After building a few tapped horns, I've noticed that more "extreme" designs are less forgiving.

For instance, here is the measured and the predicted impedance curve of a prosound twelve in a four cubic foot tapped horn. Hornresp predicted a strong resonance at 48 and 110hz. In the measured results we see a strong resonance at 100hz and a slightly deficient resonance at 48hz.

Above is a pic of a tapped horn with three eights, and it's measured impedance. (The purple trace.)

The impedance curve is screwed up. There should be a strong resonance at 35 and 85hz, but there's only one resonance. This "tapped horn" is behaving like a bandpass box. Sorta.

Here's the exact same subwoofer, but I took it apart, and went CRAZY sealing gaps. I literally went up and down the seams with a flashlight, and found about a dozen pinholes. And then I dumped half a tube of liquid nails into all the seams, to insure they're airtight.

Now the resonance is deeper, and more pronounced. But it still doesn't match the prediction.

On the upside, tapped horns seems to be a lot more forgiving of mistuning. For instance, if you mistune a bandpass box, it could create a six dB peak, or it could drop your efficiency by ten dB. When a bandpass box goes wrong, it seems to wrong in a spectacular way.

One the downside, I'm starting to believe that ultra-small tapped horns are VERY unforgiving. For instance, the build quality on my 40hz tapped horn is shoddy at best, and it works just like the prediction. But my 23hz tapped horn, which is half the size, is anything but forgiving.

[bentoronto]

I am puzzled by a number of thoughts posted and hope someone can help me see the light, please.

1. With a TH with much absorbent and with a signal an octave or two from the resonance, I thought there'd be little difference in output or impedance between enclosure types? In the tuned region, a big difference in output would mean big boomy sound because the rest of the speaker compass is not big?

2. At low frequencies, I take impedance to correspond to motion of the voice coil and with highish impedances meaning loosish, boomish, less controlled motion? Is it true that an "acoustic suspension" is more linear than the restoring forces of the driver's mechanism?

Thanks.
Ben

[Patrick Bateman]

With a couple more years of building under my belt, I threw together this post to describe what I think one should expect when they sim speakers:

Audio Psychosis • View topic - How to Make You Speakers Sound Better (for Free)

In a nutshell -

• IMHO, WinISD should only be used if you're building a sealed box
• If you're doing anything more complex, Hornresp is the easiest program that will give you accurate results.
• There are so many variables in all other box designs, that using software that ignores these variables is a recipe for disaster. For instance, simply moving the port on the loudspeaker from the front of the box to the back of the box will alter the frequency response.
• Narrow 'blips' in the predicted response will likely be absent or reduced in the real world
• Broad peaks and dips in the response are what we need to look out for

Sometimes it can be helpful to use an iterative process. For instance, when doing 3D design I will sometimes start with a design in 2D, get it correct, then export it to a 3D package. The simpler 2D interface makes it possible to 'rough out' and idea before I do the heavy lifting with a more complex program. It's the same way with speakers. It's fun to 'knock out' a simple design in WinISD or the like, but I always fire up hornresp or Akabak before I get ready to make sawdust.

The latter two programs are time consuming, but not as time consuming as a failed or poorly engineered project.

[Lord Baccus]

how do you feel about placement of an enclosure? as in Loading the sub or port?

I feel after reading this entire thread, that if I had done a lil more research I would have had better results? but I was trusting in someone that was suppose to know what they were doing? I was trying to jump into the SQ game without doing the testing myself & well we see the results... for the RECORD Lambros @ Ultra did replace the driver as it seemed that the pole piece was shifted and was affecting the movement of the coil, damaged in shipping / the slugs on the motor were shifted as well. I sent the New sub to my Daughter in Florida and it was installed in a 1cf^3 enclosure with a 500watt mono-block amp and lives on happy and healthy in her Mother-mobile C=

[gorgon53]

Originally Posted by Eva

What I see in these pictures is a piece-of-junk car audio driver with many design compromises, it's intended to be used only in sealed boxes. A vented box with a single big port in one corner could have produced exactly the same result.

Excursion capability is high, power handling is high, but the design is extremely weak when it comes to handle the cone tilt resulting from uneven acoustical load around the cone and high excursion. That's what produced the rubbing of the voice coil against the pole piece. The driver is not intended to work against a hard and uneven air load (the scenario in most horns).

To make it worse, the spider is very far from the pole piece, so any minor cone tilt results in great deviation of the voice coil from the center. The donut style rubber surround makes it even worse because it does not provide any cone centering effect either.

Such a driver would require at least a symmetrical taped horn or symmetrical ports in a vented box. Anyway, a dual spider driver, or any driver with the spider closer to the pole piece, or just any driver with a cloth surround would be much more reliable in such a tapped horn (for example, almost all pro-audio subs).

Furthermore, one layer of the voice coil got almost completely unwrapped but there are no signs of darkening, so heat was not involved at all. I think it was just due to the lack of quality and quantity in the enamel used to put the magnet wire together and/or improper curing.

Low cost Chinese car-audio subs? No thanks

Spot on.

[Brian Steele]

Originally Posted by Patrick Bateman

Another thing that I've noticed in my measurements is that a sealed box mounted in the trunk of a car sees an increase in the impedance curve. My hunch is that the air in the trunk "presses" against the cone, raising impedance and lowering excursion.

Impedance should go down as excursion is reduced. Perhaps you meant the Fb is reduced?