Tapped Horn corrections needed ?!

Greetings,

As a new member I would like your opinion (share experience) about the THAM12 tapped horn. I have been building speakers and horns since 1970 and just re-entered the speaker building arena . My target is to design and build equipment for musicians, me being a bass player, and the 2 bands I play in need a portable PA.

So the target is, small boxes with great sound, rock solid and affordable. I surfed the internet and ran upon Danley and the tapped horn. Small (check) great sound (apparently, as being commented as great speakers by many of you), rock solid (depends on material and construction), affordable (DIY).

Than I found the THAM12 and a ISO-view of the Danley TH115. The ISO-view is hard to read, but after some staring at what the construction was, I noticed 2 real differences (apart from a different way of folding).

1 - The speaker coupled to the horn is like it is done in all horns which are designed by the book in the Danley TH115, not so in the THAM12.
2 - The folded horn has 45 degree panels to help the air "around the corner" in the Danley Horn, partly NOT so in the THAM.

Question on 1 - Most designs found in forums do NOT have an adjusted loudspeaker panel to couple the speaker to the Horn. In the Horn design theory it says you need to calculate the effective surface of the speaker in question and use this number to determine the start surface of the horn. So a 12 inch speaker has a effective surface of say 530 square cm. If the tapped horn has an internal width of 32 cm, the opening for the speaker will be 530 / 32 = 16,5 cm and not a perfect circle of 28 cm.

What is the reason for NOT using this design theory (it is NOT complex to build, just a different shape of the loudspeaker panel (figure 1 shows this theory)

Question 2 - The same horn design theory describes that folding a horn, requires 45 degree panels on each fold of the horn (in fact one 45 degree panel = 2 45 degree corners making one 90 degree bend of the horn). Since all tapped horn are like folded horns, I would expects these 45 degree supporting panels to be in all designs. In most designs at least the first 2 or three bends of the horn lack these panels. In my opinion, this will cause standing waves in the horn, where the sound of the horn will change, adding more power. Played realy loud, the horn will not sound as a well designed horn, but become booming or boxy. I have read comments from people in forums about the Danley horns sounding smooth and loud, no matter how hard you drive them. That only makes sence because of the Danley horn following the design principle of a folded horn from beginning to the end.

WHY NOT apply these panels, it is just some more work, not really difficult as i have seen by the quality what some of you are building. Figure 2 shows and adjusted design of the THAM12 horn found on the internet, with a "correct" loudspeaker panel and 45 degree corners.

Figure 1


Figure 2



If I am talking bullshit, please provide me with facts on why the THAM12 is designed as i found it (which is like most of the other designs are).

kind regards

Frans


"do you still use Windows 8, I use Windows 95....."

Last week started the build. Here are some pictures. This is a test version, build from MDF, as I find this a handy material to work with. It showed that I have to correct some measures, things did not fit exactly.

I will post the new version of the drawing later.




This is the speaker to Horn throat




putting it together, i use 90 degree clamps, handy stuff!




The inside horn part...



The side panel with assisting lines, so I know where to put it.




Side panel and inside horn part connected. The panels to prevent vibrating of the cabinet are added.



Added bottom, back and top panel, also screwed and glued to anti-vibrating panels.



The future home of the Faital 12FH510




That;s it for now, next week more work to do.

Added panels today.


Included bracing




Tommorow, the other side panel and the first test without the 45 degree panels. Curious if this sounds OK or needs 45 degree panels anyway.

good stuff....

These are the tops, that should go with the THAM12 to make the portable PA. Will be used from about 120Hz upwards to 18 KHz.
It is a trial version, the official will be 12mm birch B/BB quality.

They hold Faital 4FE32 and will be pole-mounted (currently 35mm pole, but I think I will go for 25mm pole in the official version).
I used 2 programs to figure out how to make a suitable box for them, want to drive them hard. Should be able to handle 250 watts / each for the complete PA signal (instruments and vocals).

Fornt view




Rear-view




The somewhat strange panels on the back are to strengthen the pole mount panels on top and bottom.




I will add a drawing, but it's straight forward, 4 closed boxes of 0,6 liters each. Speaker serie-parrallel.

Hi JDFWessels,

I am really interested to know the outcome of your build with the adjustments made. Looking forward to the updated pics & reports.

first of all, I haven't build a Tham but few other tapped horns with pleasing results.

---

Question 1

---

Have you thought the space in front of the driver as front chamber, acting as mechanical low pass filter, described in horn design papers (was it Dinsdale papers)

---

Question 2

---

Danley has wrote about the subject, when they tried all rounded corner version of servodrive subwoofer, result was less lowest notes (less air space contributing in air movement) and no noticeable difference at higher end. Tham is very small in measures, so standing wave resonances are quite high up in frequency...


I'd like to see measurements your version compared to original design.

Personally, I have had much more trouble with standing waves in midbass horns than in subwoofer horns.

Best luck for Your build, proof is in the pudding....

Andy

Hi Andy,

First, before answering your questions, let me tell a bit of background for this threat.

I am a bass player (biggest hobby) and ex sound engineer and have build lots of boxes in the past. I am very critical on low end sound, that's my area in the sound field. For the 2 bands i play in, I wanted to design and build a compact PA that should fit in my car, or in a very small van.

3 design principles;

- MUST be SMALL and scalable (small venue, few boxes, larger venue, scale up number of boxes).
- Pro sound quality, equal or better them most brands. Quality before SPL, unless we end up with not enough SPL.
- Affordable in price (cost of material)

So I designed a bass-reflex box based on my 10-inch cabinets for my bass guitar as PA low end and mid/high satellites on a pole (4 * 4 inch) like the HK Elements design. HK use 3,5 " but I wanted a bit more bite so 4 inch was the driver. The 10 inch bass reflex was kind of OK, but having a full drumkit on these showed the need for to many boxes. So I went up the scale to design a 12 inch vented enclosure, with much better result. The satellites were OK from the first design attempt. Then I came across some audio design forums and found out that developments had gone much further than my knowledge was and I became aware of the Tapped Horn, sir Danley and stuff like that. So why not use this design for the low-end part of my portable PA. Tapped horn could be kind of small and, as claimed with Pro audio quality, simply produce more dB with the same (crown digital) amp I plan to use. So I started reading the topics and threats and the more I read, the more confused I got.

Point here is, that there is a unknown gap between using modeling software to design a box, which sounds good on paper (if at all possible to have that conclusion) and building that design. In almost all threats I noticed that (without me wanting to insult any individual who contributes to these forums) people draw conclusions based on assumptions or "theory" and as a result not apply new idea's to old concepts, thus NOT improving. As my strong feeling that this is NOT the way to design, it looks to me that Pro's like Tom Danley most likely uses a method / approach, which I also believe we should, by combining theory and practice (even if we don't fully understand why) to improve a product. I have read numerous times, that people praise his products (I have not heard them yet) and state they are not able to match them, while on the other hand restrict themselves to the outcome of design software. I think NO-ONE is capable of stating (let alone proving) that this software is fully correct or covers all design issues and rules of physics, since there are way more (combination of) variables influencing the character and quality of sound than we all are aware.

MAYBE, applying most / all of these idea's TOGETHER will result in a better design, even each idea by itself will not!! :whistle:

Having said that, I wanted to do the following;

1 - Use modeling software to start the design " as good as we can get on paper) so we know what driver, principles frequency response, predict "problem area's"
2 - Use the "old school" horn theory to convert the principles design in a combination of principle and theoretical design (like how to fold a horn, use pressure chambers,etc.)
3 - ADD common practice of what we / the market have found so far to enhance the design
4 - Use knowledge from patents to improve the design
5 - End with a design that is a combination of points 1 - 4 and that is as close as possible to the theoretical design. This is probably better understood by the following: When we fold a horn designed in Hornresp, we run into a design conflict at each corner. The horn is designed as a straight line, the box is folded. Unfolding the end result does NOT give you the original design, but a horn with triangles on the side of the horn at each point where we have folded the horn. The bigger the surface that is bended gets. the bigger the mismatch. So WHY keep on "Hornresping" for day's when what you build NEVER MATCHES WHAT you design?
This point has been brought up numerous times, but "ignored" by most people stating like "that is only a small mismatch, you will not notice that in the end result" or "this extra volume in the horn at the corner adds to more air mass, so will even improve the ned result". My Point is, you could correct this by using extra panels in the corners to get closer to the design. If it would NEED more air mass, why did the design software not suggest to add these parts of volume to the design in the first place?
6 - MIS-Use common practice to solve other issues. Well MIS-use maybe is a strange word, but I think you will get the point.
7 - Build, listen AND measure AND correct till we get what we want. I have heard sound systems there are measured in full detail and show nice graphs and high SPL but SOUND CRAP.

I do not have enough paper here to explain all that is in my head, but ONE principle stands out:
- If a theory or practice does not support your idea's, DONT ignore it or through it away, but keep it insight and examine if it could be used otherwise, OR WILL work when combined. It is called empirical design and is to my opinion a very good methodology to improve my product. As an example, the "reflecting or extra" panels in corners are ignored by most people, because the theory e.g. states that these panels are to small for the low frequency wave size "to be seen by the wave", "so they don't work". As a result, we don't use them. The theory is correct. Another theory on these panels state that "there are no standing waves in our folded tapped horn, because the length of each segment is to small for the wave length to generate a standing wave", so we don't use them. MY point is, I will use them, because;

1 - they bring me closer to the original design,
2 - The will help air to flow more evenly (no bounce back of air against a 90 degree surface) which can be heard in my opinion
3 - I DO see those panels in designs of e.g. Tom Danley, so why are they there when they don't work?
4 - They help fighting panel resonance resulting in very stiff boxes.

I hope you understand what I want to do here.

Now for your questions;

Question 1

Have you thought the space in front of the driver as front chamber, acting as mechanical low pass filter, described in horn design papers (was it Dinsdale papers)

- Yes I have, but also, because of the principle of a horn. A horn improves the coupling efficiency between the speaker driver and the air. The horn can be thought of as an "acoustic transformer" that provides impedance matching between the relatively dense diaphragm material and the less-dense air. It "transforms" high pressure / low air displacement (by the driver) at the throat into low pressure / HIGH air displacement at the mouth, hence high SPL. The result is greater acoustic output power from a given driver. Since this is a basic horn principle, I don't want to ignore that, so I added a pressure chamber with a ratio of 2:1, so the effective surface of the horn throat is halve the size of the effective surface of the speaker. I don't know if 2:1 is the best, but did not want the push it to hard, since I want to be able to produce a big bang with a small box and NOT end up in distortion at higher levels.

May be , we could also have a full size opening and consider the first segment of the tapped horn as the pressure chamber, so the volume of air in that segment is half the volume of the displaced air and loads the rest of the tapped horn. Doing this could mean that S0 and S1 are not part of this first segment of the horn, but can be designed as trapezium shaped chamber, where the surface that starts the horn is half the surface of the speaker. So for a 12 inch speaker (effective surface around 490 cm2), the surface at the start of the first bend will be 245 cm2). For a THAM 12, which is 32 cm wide (internal) the hight of this surface would be 7,6 cm. The pressure chamber could even start as small 0 cm and expand to 7,6 cm with a full size speaker opening.

Question 2

Danley has wrote about the subject, when they tried all rounded corner version of servodrive subwoofer, result was less lowest notes (less air space contributing in air movement) and no noticeable difference at higher end. Tham is very small in measures, so standing wave resonances are quite high up in frequency...

- As stated above, yes I know they have. Something however does not fit here. If it was less air because of the corners adjusted, than there is a mismatch with the design and the practical build. The corners ARE NOT IN THE PRINCIPLE DESIGN, so why is there less air when you fold the design and correct the mismatch in the corners to come close to your design.? That makes no sense. On the other hand, IF the extra volume IS needed, why is it NOT in Hornresp design results?
- Standing waves is also correct, but since each frequency in the music we make also has harmonics, they can become harmonic standing waves. Even if they don't cause a big problem, I still want to try the panels for reasons stated above.
- When we do NOT measure a real advantage, I found that measurements don't tell the whole story of the quality of sound. I mean, when you would put a number of speakers next to each other in a hearing test and ALL these speakers have (almost) identical measured characteristics, SOME will sound "better" than others. On paper they are the same, audible can make a hell of a difference.
- I have done some listening tests and found the horn with the corrections panels to sound more natural (less boxy / trapped sound impression) compared to the horn without. At low voles, there is no audible difference, the loader it gets the more noticible it gets. As an "independent" proof, I asked other people which box sounded better (they don't know the difference inside the box, and see 2 identical boxes. ALL individuals stated the box WITH the correction panels to have a "more pleasant" low end, "rounder" and such qualifications when played loud. And THAT is what I am looking for. They also stated that the THAM 12 WITHOUT the correction panels sounded a bit LOUDER, so that matches what Danley found.
Having said that, it is common practice that boxes that produce VERY clean sound, tend to come across as LESS loud, same goed for distorted. Our ears seem to be master in fooling us......

To me that is what it is all about, if it works, but we can't explain it or measure it, its still OK, because it produces positive results.

Now if these panels were not needed, than WHY does Tom Danley actually uses them, like can be seen in the TH-115 wire frame drawing?

Andy (and others); Feel free to comment or add / contribute to help improve the end result.

MAYBE, WE CAN SET A BEST PRACTICE SET OF RULES TO USE IN TAPPED HORN DESIGN.


Measurements will follow if a have a complete test set available. I have build a standard THAM12 and a "corrected" THAM12, currently I am building a TH-Mini clone, inside, it is designed with a combination of corrections and additions form the forums and the patents and my "gut feeling".

greetings,

Frans

Frans, thank you for your lenghty reply. My intention is not slaughter Your methods as I've taken same path as You when building loudspeakers.

Small agitation got your tongue loose, in a good way.

I tried the corner reflectors when I built my large tapped horns, called Miss piggy. You can found few pages from speakerplans.com.


I had some limitations to my design.

- 18" driver
- Must be usable in singles
- minimum single box sensitivity 100db / 1w / 1m at 30Hz.
- bandwidth must extend to 120Hz as usable crossover frequency, as I didn't want separate kick bass cabs,
but continue with straight mid bass horns playing roughly 100 - 1200Hz
- most efficient use of 2440 x 1220 plywood. This lead to 1200x1200 side panels and 600 inner width.
(and bloody heavy and big cabinet, dolly boards+muscle are must)

I started the prototype build without corner reflectors and attached one side panel with weatherstripping and screws, so I could try modifications to the hornpath.

At the end of prototyping the largest (last) corner reflector was deleted, because that sounded better, more accurate and gave bit more to 30-40 Hz area.



I also tried different size front chambers by attaching round disks in front of the driver. This is easily possible because the next surface is parallel to the driver mounting plane.

Throat area was maintained in round slot shape.

When the throat chamber got too small, efficiency and accuracy got lost. For my surprise the too small throat chamber influenced to the lowest notes also, for some reason I thought that would only act as mechanical lowpass filter.


For my background I do commercial sound, live sound, studio work and I have few guitars and 5 string bass...


Keep up, Andy

Hi Andy,

Thanks for the reply. By the last panel, do you mean the panel that is closest to the mouth of the horn, as the last if you follow the path from the speaker to the mouth opening?

Interesting, would you have any clou as to why this is. It is the biggest "mismatch" from the model, but as by your testing has the least effect.

I will take a look at miss piggy, must keep you out of the gym

Since i need to earn a living, my building activities are a bit on hold. As soon as i find some time, i will update this topic.

I also want to do an experiment; build a tapped horn WITHOUT hornresp, just the following parameters;
- Volume of the horn (depends on lowest freq.)
- Throat surface size (half the effective speaker surface)
- Mouth surface (depends on lowest freq. for the horn)

Then use a large peace of paper and fit the folding scheme into the volume. The width is 2 inches more than the outside diameter of the speaker. I know the mouth surface, so i know the height of the mouth. Total height of the box is 2x height of the mouth. I know the box volume, i know height and width so this gives depth size. Then start folding, where every corner is an exact square and add reflection panel later.
This would give me a horn drawing within an hour that works.

Keep you informed

kind regards,

Frans.