DIY 225 Hz tractrix horns. Lots of pics!

[jwc]

jeremy

Way to go man. I got to the point of the mount and had to stop. All the parts are sitting in storage right now. You did it just like I was doing it. Same mouth...225Hz flare....pure tractrix.

Keep us updated.

jc

[klipschaholik]

JC,

Do you think this horn design would work on the Jamboree?

[D]

[jwc]

That would be a great match up. That is what I wanted to try. Or place them on top of Klipsch MWM's.

I would bet that the Pure tractrix 225Hz horn won't make to the high frequencies. You would have to do 3 way. Hopefully Jeremy will gets some curves up for us in the future. My build will be on hold for awhile.

If I end up ditching the project....I could send someone all the cutouts. I have enough to build two and possibly more.

jc

[DrWho]

I don't know what the frequency response is going to look like. I plan on investing in some measuring equipment. Maybe a USB microphone and some software for my laptop. I was thinking about going to the local Guitar Center tomarrow and see if they have anything. My JBL drivers are supposed to reproduce pretty well down to 300 Hz. I plan on crossing over some where in that area. I just wanted to make sure the horns would not cutoff before that. I may go with an active crossover at 1st to get things dialed in, and then see if I can figure out how to design a passive network for them.

You might find this helpful:

http://measuresound.googlepages.com

Basically I've got pics and links to all the equipment in my measurement rig.

I use the mixer for other things, but you can do the exact same thing as my rig, but replacing the USB interface and the mixer with something like this:

http://www.musiciansfriend.com/product/MAudio-MobilePre-USB-Portable-Audio-Interface?sku=701368

(It ends up costing about the same eitehr way)

Anyways, I'd love to hear about your measurement rig when you get it put together. Perhaps even add another page to the website (or more pages if anyone else wants to share their rigs too).

One quick question though...is there any reason you chose to only flare in the horizontal plane? The area expansion rate is all that matters for the loading of the driver (so you're fine there), but I would think you'd get better polars by flaring both ways?

[ZAKO]

For his first time around on horn building Dont be too pickey on the verticle flair....That is a great achievement...A job well done.

[Guest " "]

Looks super.  That bendable MDF looks easy to work with.

[capo72]

For his first time around on horn building Dont be too pickey on the verticle flair....That is a great achievement...A job well done.

Thanx Maron, that means alot coming from you. I admire you St. Louie horn guys (Dave, Carl, and yourself) for appreciating JBL and Altec along with Klipsch. I have not talked much about this project here because it is all JBL and I don't know how it will be accepted. I would love to get together with you guys some day.

Mike, the main reason for lack of vertical flare is ease of construction. I have given some thought as to how it could with flare in both directions and it would definately be more work, but I think I could do it. As Maron pointed out this is my 1st horn build and i'm pleased with the results. I learned the basic process behind this construction from others here and some on the Lansing Heritage board. I would not have known where to start without others laying the path. I learned alot along the way and feel I could expand from here we'll see where this takes me.

I'm not too concerned about the low end output of this horn driver combo. My woofers are plenty capable of playing up to them. I'm more concerned about the upper end of this horn combo not playing up high enough to crossover with my JBL 077 tweeters. I will probably have a large hole in the response there, and I may need to find a different tweeter. Maybe Bob's CT-125. I definately need to get some test equipment to see where I stand. I think it's more about the journey than reaching the destination with DIY speakers. Thanx for the kind words!

Jeremy

[Daddy Dee]

Jeremy,

wow, great looking work. Thanks for documenting with pics and posting.

i had no idea that sort of mdf material was available. That looks especially well suited for horn construction.

[capo72]

Jeremy,

wow, great looking work. Thanks for documenting with pics and posting.

i had no idea that sort of mdf material was available. That looks especially well suited for horn construction.

I purchased the MDF from Rockler Woodworking, though it is available from other sources. It was about $16 for a 2'x4' sheet. As I learned the hard way, you have to be very careful cutting it, so the ribs don't break. I would use it again though. I was going to try 1/8" Luan, but some of the scraps I had were just not flexible enough. I spend lots of money with Rockler, so it was a no brainer as I was order some other stuff from them. I think it was $10 to ship it and it showed up in 2 days, can't beat it!

Jeremy

[jwc]

Mike.

What do you mean no vertical flare? There is vertical flare.....it is just constant. Over the length of the horn about 17".....the vertical height goes from 2" to 12".

Mike this is a Edgar Horn. This was how he built them. The logic for the mouth shape was to make the width 2x the height. Therefore....not as much "flare" is needed in the vertcal expansion. This was optimal to his design to prevent impedance swings. I'm not making the vote that this is the way to go but this way works. This design is just like ALK's Trachorn except it is for a 2" driver and has an Fc of 225Hz.

jc

[Guest " "]

"This design is just like ALK's Trachorn except it is for a 2" driver and has an Fc of 225Hz."

I can see this working out perfectly.

[DrWho]

What do you mean no vertical flare? There is vertical flare.....it is just constant. Over the length of the horn about 17".....the vertical height goes from 2" to 12".

Mike this is a Edgar Horn. This was how he built them. The logic for the mouth shape was to make the width 2x the height. Therefore....not as much "flare" is needed in the vertcal expansion. This was optimal to his design to prevent impedance swings. I'm not making the vote that this is the way to go but this way works. This design is just like ALK's Trachorn except it is for a 2" driver and has an Fc of 225Hz.

FIrst off, I hope I didn't come across extra critical or anything like that. I was just curious if it was an intentional part of the design, or more an ease of build thing.

So what is the advantage of having the width be 2x the height? Is it more an aesthetic thing, or are you saying it directly relates to the impedance response? There are many approaches to horn design, but I think the PWK way starts with efficiency, then polars, distortion, and finally bandwidth? Would impedance come 5th in the compromise chain? I don't mean to belittle other approaches (so if I come off that way I don't mean to). I'm just trying to understand the goals of I guess the "Edgar approach". I had an interesting chat with Roy at the pilgrimage and he was talking about some of the horn design philosophies...he made a comment that it was more important to know why something worked well than to come up with a design that happened to work well. Basically the point being that knowing the approach makes it possible to engineer similar performance for different applications. I know it's a first project and this is just a hobby and all, but I think it would be interesting to learn more about how the horn performs compared to the expected outcome.

[jwc]

Doc...no I didn't sense any bitterness.

This technique that Jeremy used is how Bruce Edgar designed the DIY Tractrix flare horn. You can download the original article from the volvoteer site mentioned above.

The constant vertical flare is definately there for ease of build for the woodwooker. Since the vertcal flare is constant...the horizontal flare will need to be "curved" as such to keep the "area" appropriate to meet the tractrix calculation. In Edgars article....he did the area at every one inch. But...like jeremy did...you can calculate that "area" at every 1mm if you want and then make the adjustments on the horizontal flare to keep the right "area".

It is a great project that you ought to try.

The article explains Edgars reasoning for the 2x width to height mouth ratio. You would think it would have something to do with polars or coverage but it was a means to "control the driver".......less impedance swings.

This same technique is used on his midbass horns too. Either the vertical or horizontal flare was kept at a constant expansion and the opposite flare was "curved" to the extent to keep the area expansion correct to maintain tractrix. I have a 90Hz tractrix design for two 12" K31 woofers straight horn with the same technique. I have eight of those woofers here....don't know when I will get around to it.

Now of course......I'm sure with modern testing and fiberglass designs.....there is a way to improve.

I'm not giving the above information as something I think is correct or the right approach...but all these horns of the DIY Tractrix you seen posted on the forum....are all Edgar horns.

jc

[jwc]

Here it is. Whether this is REAL important or not.........

[jwc]

Damn Doc....I just realized....you pound all kinds of threads with the request of data......man...I actually presented some....I momentarily have a big head......soon to be squashed I'm sure.

jc

[capo72]

Thanx JC for providing all of that info, you understand it better than I.

I just got back from Guitar Center and I looked at that M-Power USB interface that you have Mike. I think it was $99, then they had one with Phantom power and more input and output options for $149. I think that may be the way to go for me. I looked at microphones too but i'm a little confused there as to what I need. Maybe you could give me some suggestions. If I understand right, I can get that USB interface w/phantom power, a mic, and some freeware and i'll be able to get some measurements. This measuring response stuff is all new to me so I may need some direction. I'm more of a "figure out how to build stuff" kind of guy. Thanx for all the help and info and positive feedback so far. Keep it coming!

Jeremy

[DrWho]

Damn Doc....I just realized....you pound all

kinds of threads with the request of data......man...I actually

presented some....I momentarily have a big head......soon to be

squashed I'm sure.

Do I really pound threads for data? Now you've got me all self-concious. lol

I need to finish reading up on all the horn literature I've come

across. I've got that long 100 page thesis on Tractrix horns, all of

Keele's PDF's from his website, and then some other random articles and

now apparently more from Edgar. I'm forcing myself to understand

everything as I go, so it's taking me a while to move through them and

I'm usually too tired in the evening for thinking hard (especially with

classes and other projects wearing me out too).

I've only read the snippet provided in that image you posted

and I've gotta wonder if most of what he's measuring isn't due more to

the shape of the throat.

A while ago (around the time Al was doing his Tractrix horn) I went through and created my

own spreadsheet to calculate horn flares, starting with this same style

horn to make sure I was applying the math correctly. I then expanded to

create horns of all sorts of shapes that followed the area expansion

rate. One of the things I noticed is that you often run into sections

where the horn becomes narrower in a certain plane because the other

plane is expanding too fast. The narrowing has gotta totally disrupt

the flow of air along that surface, introducing a short segment of

increased air velocity - which to an extent introduces "chuffing" (or

random motion of air) that will result in distortion. I would imagine

that this would also change the load the driver sees and thus the

measured impedance will change.

You've also got termination issues

between the driver and the throat of the horn (round diaphragm versus

square throat opening). I suppose there are ways of improving this

transition, but it can't possibly be perfect. You've also got different

frequencies wanting to diffract at different angles, so if the shape of

the flare isn't fast enough, then the sound is going to reflect off (I

think), which is probably part of what causes the harshness with horns

(comb-filtering from the reflections). Eventually the waves have gotta

be so big that the reflections don't matter (not enough phase

rotation). Horn designers are always talking about the directivity of

the HF being determined more towards the throat, whereas the low-end is

controlled closer to the mouth. Conical expansions are always

considered to be the highest distortion, and the vertical part of the

flare is essentially conical in this situation.

Anyways, my point being that fixing the mouth ratio and setting the

vertical expansion is changing all these other aspects. In other words,

I would propose that you might not notice similar trends by comparing

mouth ratios with other methods of flaring the horn.

For instance, one of the horn designs I "forced" started from the

original approach for the tractrix horn - which was to find an

expansion rate such that the wave was always perpendicular to the walls

of the horn. The original equations assume a spherical horn and thus

are perfect in that regard. I went ahead and created a spread sheet

that forces either the vertical or horizontal to follow the same shape

as the spherical horn (at least in the middle). The horizontal

expansion is then determined by the area expansion required. I need to

pull out the spreadsheet, but I think the ratio ended up being close to

2:1 (width:height). I think it might have been closer to 1.8 or

something like that. Anyways, the idea behind this approach would be to

find at least one plane where the sound was expanding the same as the

spherical horn. I drew up another design that modified it a bit so that

the flare happened twice on the vertical flare (basically 1/3 of the

way in) in an attempt to "average it out". And then I'm working on

another one that forces the corners of the square horn to follow the

expansion, but I don't think it's going to work (constriction up near

the throat).

I dunno if it matters or not, but I've been putting the

project off until I can read more about the math to see if it makes any

sense to even try. Everytime I talk to someone about horns I always

come away with other things to try and having absolutely no background

to know what works or what is even the common problems to overcome

(kinda hard to find solutions if you don't know what the problems are).

Well I've probably talked too much again, but such is the way of forum life.

[DrWho]

I just got back from Guitar Center and I looked at

that M-Power USB interface that you have Mike. I think it was $99, then

they had one with Phantom power and more input and output options for

$149. I think that may be the way to go for me. I looked at microphones

too but i'm a little confused there as to what I need. Maybe you could

give me some suggestions. If I understand right, I can get that USB

interface w/phantom power, a mic, and some freeware and i'll be able to

get some measurements. This measuring response stuff is all new to me

so I may need some direction. I'm more of a "figure out how to build

stuff" kind of guy. Thanx for all the help and info and positive

feedback so far. Keep it coming!

If I had to do it again, I would go with the $150 option. I needed the

mixer for other reasons which is why I went with the cheaper $100 one.

As far as microphones, you need a "calibrated mic" - which is to say an

omnidirectional mic with a flat frequency response. The Behringer

ECM8000 is the cheapest available ($50) and my measurements came out

within 1/2dB of the Klipsch anechoic chamber for most of the audible

spectrum. Information above 10kHz isn't very reliable, but there's

rarely any reason to be analyzing that region anyway. I don't trust the

low-end below 20-25Hz, but you're still going to be within about 2dB.

Again, there's really not much you're going to do down that low that

will matter with the mic. There are better options out there if you

want to spend a little more money, but I really see no reason to unless

the cost difference doesn't matter. I believe ETF Acoustic sells their

mic for $300?

http://www.etfacoustic.com/

(scroll down to the bottom of the main page - at the time of writing this it seems their server is down)

You will have no problem spending upwards of a few thousand if you wanna get insane.

Anyways, it's really that simple...get the Behringer Mic, the M-Audio

USB with the phantom power, and then all you need is a computer/laptop

with the free software installed. The hardest part will be finding the cables to connect everything.

Btw, if you want to do impedance and distortion measurements, you migth check out this free software:

http://www.audua.com/

I'm not sure yet how you do microphone compensation (I have correction

values for the Behringer), but you'll also need a fancy box to do the

impedance measuring. This is currently my next project to put together

so if you'd like I could keep you updated (or even build a second box

at the same time if you'd like).

[George Roland]

Hi Jeremy,

Thank you so much for posting these pictures and description of your construction process on these beautiful horns. You did a great job on them! I just posted a follow-up on my Altec 290--16K drivers which were mated with an Altec 311-60 horn for a new mid-range setup for my k-horns.

Your work makes the whole process of calculating and plotting the curves through building the horns more comprehensible, and I am inspired by it. I am sure many others will benefit from the detail and photos you have provided. 

Keep us posted as your research continues!

George Roland

[ZAKO]

DOC....Chuffing for a HF horn??? come on its a compression horn design...You read alot but what do you absorb Grasshopper??? Horn designs have been around for centurys...For makeing music,,and reproducing music...Try makeing a trumpet..a good compression design...Try actually making a reproducer audio horn Tractrix or Smith DSH...Its all in the sweat equity to learn the why and how of whats going on. Ive done some horn designs to learn whats going on...Do,nt forget to collect some good professional Not the cheap crap. testing equipment along the way....and dont throw away the mistakes.