Collaborative Tapped horn project

Disregarding any enclosure related limits, the power handling for a multi-driver enclosure is calculated by multiplying the power handling per driver by the number of drivers. Double-click on the Eg box and Hornresp converts watts and nominal impedance to volts for input power.

As others have said, the best way to check your design is to run your hornresp model with the output of your amp (or amps you are considering) at max power (max RMS wattage into your impedance) and look for excursion issues. Hornresp has a max SPL function - when on the SPL response, select tools, then Max SPL. Enter your maximum wattage and xmax, hornresp does the rest. The red portion of the SPL curve mean that you are excursion limited, black indicates that excursion is within xmax.

For example, I am currently constructing a tapped horn subwoofer for my brother. It will use 4 Tang Band W6-1139SI. They are wired series-parallel, so the net impedance is 4 ohms. These drivers are rated at 50 W RMS power, or 100 W music power. When summed, 4 of them will handle 200 W RMS or 400 W music.

According to the spec sheet, the BASH amp is capable of delivering 500 watts into a 4-ohm load. In this enclosure, applying all 500 W into the 4 drivers results in a maximum of 12 mm of excursion within the design's pass-band (30 Hz-100 Hz), but applying 500 W of test-tones to coils that can only handle 200 W will result in thermal failure (the magic smoke comes out). Applying all 500 watts also indicates that there will also be excursion problems below 25 Hz and from 34 to 46 Hz.

Realistically, my brother will probably not be listening to test tones at 120 dB, so in this case, the oversized amp is not a concern, and is actually a benefit, as it will be able to supply clean power for transients without clipping. When set in the final location, it will likely be listened to with the amp supplying less than 1 watt average to the sub, with 10 watt peaks. This will provide almost 100 dB, which is louder than he generally listens. Excursion at these levels is a non-issue (under 1 mm regardless of frequency).
 
littlemike said:
Realistically, my brother will probably not be listening to test tones at 120 dB, so in this case, the oversized amp is not a concern, and is actually a benefit, as it will be able to supply clean power for transients without clipping. When set in the final location, it will likely be listened to with the amp supplying less than 1 watt average to the sub, with 10 watt peaks. This will provide almost 100 dB, which is louder than he generally listens. Excursion at these levels is a non-issue (under 1 mm regardless of frequency).
I suspect you will be even better than that.
Listening with just 250mW to the TH and peaks of +20db (25W) that are passed through cleanly. They will not "sound loud" because there is almost no distortion indicating over-driven, but they sure will have an audible effect.
 
Hi,
since it is some kind of quiet here in regards to design new suggestions,
I like to contribute what's spooking in my mind for some time:

I modeled a TH which is (net volume wise) similar to the TH-Mini.

As a driver I choose the B&C 12TBX100.

TH-Mini%20para.JPG

TH-Mini%20spl.JPG


I think this is a pretty good sample for that you have to put the right driver into the right sized TH
in order to make it work (at least in the simulation).

The typical two resonance peaks are not existing and the it seems
that you do not have to worry about the response above the upper xover frequency.

Some lining inside the horn (as seen in the DTS-50 or the TH-SPUD)
might be required.

What do you think?
Erik
 
Greets!

Yes, as TD has repeatedly stated WRT horn design in general and TH design in particular and as I've 'proved' with some theoretical sims much earlier in this thread, it's possible when the right driver specs are used for the intended app to get a ~ideal maximally flat band-pass response across ~three full octaves with its FL6-Fh6 being 'close enough' to the driver's calculated mid-band efficiency without the sharp phase inversion in its pass-band that HR normally predicts, so factor in the real world smoothing effect of the TH and while some have questioned the validity of at least some of DSL's TH product's claimed performance, I never did.

GM
 
@Tinitus
He He... you're right. Good point.

Actually one thing is the interest in the matter, the other is how such
a thing might sound compared to my Tangband 38Hz TH. The later is
based on a heavy mass "car style" sub with weak motor, whereelse the the B&C driver is more or less the oposite of that.

How much might this change the ability of similar TH's in regards to punch and attack (but not max. possible SPL)?

What do the experts think?

Erik
 
Just an update, in case anyone is interested in my progress. I'm hopelessly stuck on measuring driver parameters, and until I figure out what I'm doing wrong I'll stay stuck.

I've refined my procedure quite a bit, and now within either method the results are basically identical in consistancy but the 2 methods still don't agree with each other.

For example, every single time I run the closed box method I get ~6L vas.

Every single time I run the added mass method I get ~9L for vas.

(Factory vas spec is almost 12L. Qts stays constant regardless of the testing method.)

So it's a simple task of finding my error but I'm WAY behind schedule now with no idea how long it might take.

And no, I'm not buying a woofer tester, this should be something I can do myself.

I won't litter this thread with any more mundane details unrelated to tapped horns, just wanted to say don't hold your breath waiting on me.
 
Hi Tapped Gurus,

I have been lurking at this thread for a while now but I must admit I still do have difficulties figuring out if tapped horn is a design for me..

For this project , I will soon be looking at a good sub to take over from my (Beyma 10G40 powered) mains from about 100Hz/150Hz down to as low as possible (and as cleanly as possible). Say 30Hz if I'd have to throw a figure...

The use will be 50% HC 50% hifi at "modest" SPLs (home use, not a fan of high SPLs).

I would like the budget to remain... frugal (as it is a word I read here ;) )... Say 150€/$ for the driver (preferabily "sourceable" in Europe) and the same for the amp (give or take 50 on the total).

I am going the hardcore DIY route for my mains (starting from scratch) and having the hell of a job at filtering the mains but I must say that I would happily go the "lazzy route" and use a proven design for the sub, this has been too much hard work and I need a rest :eek:.

My first actual question :
-do you think tapped horn is design for me ?

Collateral question :
-in a nutshell, what are pros and cons of tapped horn vs. more "traditional" designs (vented). I vaguely remember reading some posts about this earlier in the thread, but i maust say that it is still fuzzy for me and I am not sure there is a consensus (furthermore I cannot find these posts anymore).

My second actual question :
-is there somewhere a list of design "that work" ?

Even a list of designs (speaker + cabinet sketch/plan) with expected/measured output would be really great (like the one from steve71 on page 95).

If you have any other pointers at tapped hor designs in "reasonably sized" enclosures that are proven and useable that would be fantastic.

The design from steve71 where the cabine is about 130 x 80 x 50 cm is OK but a maximum for me. Ideally, I would rather have something that could be a bit taller (more than 130cm, up to 160 cm is still good) and less wide (bringing back the 80cm to 60/50cm).

Many thanks (many many many) in advance for your help and advices.
 
@ just a guy

You're on the right track.

My TB W6-1139SI measured specs are very similar to what you posted under the added mass method. I recall correctly, my Vas averaged about 7.5 L for the 4 drivers I tested. For my measurements, I used my WT2 and the added mass method. I measured the T/S parameters with the cones facing forward and the motor clamped to my workbench. I tested Vas with the cone facing up and the voice coil vent unobstructed. Rather than use nickels, I added lead rings of known weight to the cone according to the WT2 software suggestions for an ideal test mass. My results are very consistent with your added mass tests.

I've never used a closed box for measuring Vas, only added mass. I don't know if that is right or wrong, I just know that the added mass method was what I learned back when I got my original Woofer Tester. So far, my results have generally paralleled the simulations.

I've got most of the wood cut for the tapped horn I plan to use these in, we'll see how close the real thing measures when compared to the simulation once I get it finished up. At the slow rate I am progressing - it may take me another week or two, but I will share my results with everyone once I have them.
 
littlemike - thanks very much. I'm sure I'll get this figured out sooner or later. As far as I'm concerned the added mass and closed box methods should yeild the same result - if done correctly - so that's what I'm shooting for - and the only way I know to actually cross verify that the results are correct.

jm_kzo - if "easy and predictable" is the goal I'd copy one of William Cowan's published designs. AFAIK he's the only guy that's posted easy to find designs complete from hornresp predictions to measured frequency response. Volvotreter also has some nice designs but AFAIK no measured frequency response.

If "cost effective" is the goal, it has to start with drivers that are available to you locally, and hornresp. This is a crap shoot if you don't actually measure the t/s parameters. Tapped horns are very complex in that designing around the wrong t/s specs could easily end in disaster.
 
There is something to be said about consistency

I've refined my procedure quite a bit, and now within either method the results are basically identical in consistancy but the 2 methods still don't agree with each other. For example, every single time I run the closed box method I get ~6L vas. Every single time I run the added mass method I get ~9L for vas. (Factory vas spec is almost 12L. Qts stays constant regardless of the testing method.) So it's a simple task of finding my error but I'm WAY behind schedule now with no idea how long it might take.

Just a Guy

It may sound really stupid but are you subtracting the volume of your drivers magnet and cone structure from the volume of the test box?

If you keep getting consistent results it is usually some weird little problem in the calculations not you test method.

Try testing Vas with the box by sealing the driver to the box with the magnet and chassis out of the box. I do this when I have to by sealing it with cheap hot glue between the test box and the driver chassis. It peels off quite easily but holds the driver in place as a quick and dirty method.

Mark
 
It may sound really stupid but are you subtracting the volume of your drivers magnet and cone structure from the volume of the test box?

Not stupid at all. I'm testing with the driver facing into the box, and adding the volume of air trapped in the cone. I'm pretty sure I'm accurate within a few cc's. The driver is very well sealed with foam weatherstripping.

There's only a few variables (computer and soundcard, user inputs, physical procedure, resistor, weight and volume measurements, etc), so I'm just going to revisit them all one by one until my test results agree with each other as time and frustration level permit.
 
Volvotreter

Your is close to what I came up with and built (2 of em)
An externally hosted image should be here but it was not working when we last tested it.

That's at full power = 60v.
Here is the original I came up with
An externally hosted image should be here but it was not working when we last tested it.

but didn't want to deal with the inductor. (5mh)

First listed is 106 Lt. Which is slightly larger than the TH Mini and the other is 101 Lt which is pretty close to the original.
I still have to do real measurements but that is waiting on me to make a mic. instead of using an old Shure SM57.
 
just a guy said:
jm_kzo - if "easy and predictable" is the goal I'd copy one of William Cowan's published designs. AFAIK he's the only guy that's posted easy to find designs complete from hornresp predictions to measured frequency response. Volvotreter also has some nice designs but AFAIK no measured frequency response.

If "cost effective" is the goal, it has to start with drivers that are available to you locally, and hornresp. This is a crap shoot if you don't actually measure the t/s parameters. Tapped horns are very complex in that designing around the wrong t/s specs could easily end in disaster.

Hi just a guy, thanks for your answer.

Does this mean when designing a tapped horn you'd rather not trust T&S parameters found in the driver specifications but ALWAYS measure the driver yourself ?

I know Volvotreter's site and designs. Unfortunately his mini horns are probably a bit too mini and the one with the Emminence 15' is too big :xeye:.

I'll have a closer look at William's designs on his site.

Finally, any pointer at any literature or article with pros and cons of tapped horns ?

Many thanks again.
 
jm_kzo said:

Does this mean when designing a tapped horn you'd rather not trust T&S parameters found in the driver specifications but ALWAYS measure the driver yourself ?
That's the surest way... TS specs are to often to much 'off', and a TH being Tricky & Pesky... Give it a try, fire up HornResp, choose a driver, and fiddle a bit with the driver params: the response curve can react heavyly on slight changes.
Do you source your drivers locally? I do want to build a nice TH later this year, but I can't seem to find a good shop nearby (Entre Toulouse et Montpel quoi...).

Regards, Paul
 
Finally, any pointer at any literature or article with pros and cons of tapped horns ?

You could read the patent but most people come away from that document with a worse understanding of how tapped horns work than before (including myself initially) and I'm not really sure why since it's all plain english.

IIRC though (it's been a long time since I saw how it started), Patrick Bateman's "Tapped Horn for Dummies" thread ended up turning into a goldmine of great information from the likes of Tom Danly, GM, Mark Seaton, and an interesting counterpoint from Gedlee. Some really great info there, scattered about.

As for VERY basic pros and cons -
PRO - tapped horns can be loud like a 6th order bandpass, and can do it over greater bandwidth (if designed properly)

CON - tapped horns are complex and can be intolerant of any error in t/s specs and/or construction sloppiness

There's more but in my mind those weigh pretty heavy.
 
@ just a guy:
Many thanks for the short pros and cons... It seems that if I can survive the design or find a design that works for me, TH might be the way to go for me as it seems difficult to find other design that can cover 30Hz to 150Hz with a quite flat response (and a "contained" budget).

@ bibster:
When I spoke about locally sourceable drivers, I meant drivers that you can easily buy online in France (or Europe) at worst. I have bought my Beyma's and DE250 from www.bcorde.com. Except La Maison Du Haut Parleur, I would not know a good shop in the Toulouse-Montpellier area. Audiophonics are in Bordeaux, but I am not sure they have a "brick & mortar" shop as well.

All of this stated, if the good driver or good amp for me is sold by PartsExpress at the right price, I will happily order there.

Have you settled on design yet yourself ?

Cheers.