Wide range horn confusion..

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Looking at the Oris horn projects, some questions jump to my mind from statements I have read on horn design, horn bandwidth, and the "3-octave rule".. here are some quotes:

Because horns have a limited bandwidth, they cannot do a "full" frequency range, but you can use a front horn to load a full range driver at its lower frequencies, making the unit wider range than just direct radiation alone would allow.

In theory, to achieve maximum efficiency, horn dimension should be 1 wave-length long, and 1 wave-length in mouth circumference, at the lowest frequency. At 3 octaves higher, wavelengths are 1/10 of horn size, and too small for the horn to direct them. These small wave-lengths bounce around inside the horn chaotically. It is essential to rescale a smaller horn for the next 3 octaves.

The Oris horns utilizes fullrange drivers such as Lowther or AER, and is supposed to cover the frequency band from 150Hz and up, wich will be equal to a bandwidth of aprox. 7 octaves...

Now, a quote from the Oris horn website;

The cabinet I developed for the Lowther was of course a horn, a front-horn with the throath in front of the driver and not like many other designs at the rear. Placed in front of the speaker makes the sound come from one spot, the interference from different sources (like with rear loaded horns) is not present and the sound-colour is the same over the complete frequency-range! An additional advantage is that the output at listening-position is minimal 6dB higher plus gaining more body in the low-mids!

The last statement in the quote inevitably suggests that the horn response not is configured to load only the lower rolloff frequency of the driver to extend bass response, but instead is used to load the entire band of 7 octaves.. :bigeyes:
How can this be?

ref:
Single Driver Website
Oris Horn Website
Lenard Audio

An externally hosted image should be here but it was not working when we last tested it.

The Oris Horn
 
Some more questions related to the same issue;

horn dimension should be 1 wave-length long, and 1 wave-length in mouth circumference, at the lowest frequency.

Does this apply to tractrix contours as well? My calculations points to a much shorter one.. actually <1/4 of the wavelength..

Does tractrix horns, paired with compression drivers, obey the 3-octave rule? I noticed Volvotreter's tractrix mid-horns operates from 650Hz to 10000Hz, being 4 octaves...

edit: V-treter's tactrix horn is actually made with cutoff @ 320Hz, giving a bandwidth of 5 octaves?

Anyone?
 
Horns

Hi Rocky

Most people shorten the horn (and reduce mouth circumference) to 1/4 wavelength for convenience sake, however there is usually a price to pay in irregularities in response. It is possible to smooth some of these irregularities with programs like McBean's Horn Response, but generally, 1/2 wavelength will yield better results, full wavelength better yet. The last is easily possible only at high frequencies, naturally, due to the small wavelengths = small horn. Not so easy in the lower registers without a crane or several friends.

The contour or flare doesn't really gaurantee the horn will be properly loaded over wider bandwidth. Depends on the design and the driver in use. As the frequency goes up, the portion of the horn that remains useful retreats inward toward the apex, narrowing dispersion. Constant directivity horns may correct for this to some degree, but inevitably the off axis response diminishes.

3 octaves is a lot on a horn, but possible. A decade is even better, and nearing the optimum. Beyond a decade one wonders if the horn is serving any useful purpose in directivity.

Tim
 
Thaks Tim.

With the bandwith limitations, I am not very keen on configuring the horn for any frequencies below the XO point.. say If I intend to cross at 500Hz, I configure my horn for 500Hz and not any lower to help it in the transition... How would this affect my desired 1st order crossovers...?

And regarding the tractrix shape... I use this formula to calculate it:
An externally hosted image should be here but it was not working when we last tested it.


How should I alter this shape to extend it's length to say 1/2 or 1/1 wavelength? Simply "stretching" it...?
 
Horn Design

Rocky,

Most of the older math models will give you a ratio of throat to mouth size, length, plus the front and rear chamber dimensions.
The throat to mouth size combined with length will establish the flare for a given cut-off frequency.
It would be wize to allow for a slightly lower cut-off than intended for use, as the response ordinarily becomes irregular as you approach cut-off freq.

Have a look at the following:

http://www.users.bigpond.com/dmcbean/

This program will allow you to pick all manner of horns and compare the results. It takes some getting used to, but is fairly accurate. High frequency response shown in the model will probably be quite a bit less extended than on the actual horn.

Also, some horn math calculators:

http://melhuish.org/audio/tractrixcalc.html

Might have a look at the "projects" portion of this site.

Also look at Till's posts concerning his horn fabrication here.

Hope all this will help, as a horn design is much more involved than a simple reflex enclosure. Not an easy task.

Tim
 
Horns

Rocky,

You will need to esablish the throat ,mouth, and length, then figure cross-sectional areas through the horn at given lengths along the expansion.

Obviously, if the two sides are fixed at a given flare rate (no curvature), or are parallel surfaces with no flare, the design becomes easier.

Conical is a fixed flare angle with no curvature, and is easier to calculate and experiment with. Also, because of the fast flare rate there is usually less distortion in the throat.

Tractrix are commonly used for the midrange, and your 500 Hz cut-off will be around 70cm long full size.

Tim
 
Thanks for the replys, Tim.

I must say I find the concept of horns especially interresting, but while a midrange horn can easily be fitted into my listening space, Bass horns are an entirely different matter. In fact out of the question for the time being. :( If I were to start on a horn system, I would cross from bass to mid at around 500Hz using a 2 inch compression driver, hopefully something like the TAD 4002 or a JBL 375 or something similar.. My concerns are two major ones;

1) 500Hz is well into the human voice frequency area. Taken into account that a huge bass horn is out of the question, I see no other alternate for the lower end than a coaxial lower extension. It frightens me to pair two potential very different sounding sources at this frequency.. :( Can anyone provide some feedback on this matter based on experience? As said, my midrange candidate being a 2 inch compression driver..

2) Crossovers.. My initial interrest for horn designs is founded in a newly aquired 300B SET delivering 9 sweet watts, well, at least 2-3 of the watts are very sweet.. My understanding is that SETs have "hate"-relationship to crossovers.. Now, Assuming the horn system is properly built (i'll get spherical horns milled to mathematical specs), and assuming the crossovers are well tuned and thought out, and the bass driver chosen with care; will a horn system outperform or underperform a simple well-thought full-range setup like for instance a Fertin field-coil dipole..? (as a companion for SET amps)

I should add that my goal is not very loud sound, but rather using as few watts as possible to get the sound I want.. For obvious reasons.. *distortion* :rolleyes:
 
Rocky said:
is configured to load only the lower rolloff frequency of the driver to extend bass response, but instead is used to load the entire band of 7 octaves......
How can this be?

I didn't follow the links, so I may be contradicted here _grin_ but the Oris and Azuras essentially are out of horn loading above 1500 hz. They depend on rising response of the driver in the upper bandwidth to offset the horn loading below that point. The Measurements I've seen of Azuras with DX4's drop off steeply above 10K. John K who posts as Kloss and has run several Oris configurations with two Oris and a Fostex tweeter is now configuring a three Oris horns per side with a super tweeter in the center.

I am using Azurahorns with DX4's and am now starting to work with a DDS ENG 90-1 horn (waveguide) and Radian 475PB compression driver with the intent of crossing at around 1250 hz. The Azura loads up to about 1500 hz, the eng 90-1 loads down to about 1000 hz.

If needed, I can add a horn loaded supertweeter on top for three way horn loaded system above 120 hz - at 120hz I am biamping with a digital crossover to a pair of NHT1259 sealed subs.

I would eventually like to have a 4-way fully horn loaded system, with bass horns- however, size, cost, complexity, WAF enter into the timing of that. The sealed NHT1259's pushed by a Crown amp sound so good that bass horns is my lowest priority at this point.

Will a 4 way horn setup underperform or outperform a well thought out full range or dipole setup?

At the risk of being contradicted here _grin_ while I personally lean to horns beating anything _bigger grin_ - I just don't think anyone can ever answer that with any certainty either way -

They're different, we not only have different priorities as listeners, but due to physiology we actually _hear_ things differently. I would easily believe that two seasoned listeners might hear both a really good open baffle setup and a really good horn setup while one listener preferred one and another listener preferred the other.

Every setup, no matter how _ultimate_ has tradeoffs. So it just depends on what tradeoffs an individual prefers over another different set of compromises.

FWIW, I don't think there is such a thing as a "full range driver" . I believe they would be better described as "wide band augmented" drivers. What bandwith(s) and how you augment is what the question is.

It usually takes individuals _years_ of tweaking, trial and effort, trying and discarding various efforts to get a horn system how they want it. I'm not aware of many that reach that point with horns that leave horns for open baffles. Nor am I aware of many that leave open baffles for horns. It does seem that open baffles proponents are growing, but they appear to be widely out numbered by horn enthusiasts.

What you really want to do is ensure that you have plenty of fun on the Journey - at times the destination becomes more and more elusive _big grin_

Regards

Ken L

PS the two best systems I have personally ever heard are low powered SET systems that have horns and crossovers - and both system owners were not still completely happy with the crossovers. But they were _magical_ and a true joy to listen to.

It's all about the _magic_ And _magic_ is elusive, poorly defined, and oh so hard to achieve.
 
Re: Re: Wide range horn confusion..

Ken L said:
I am biamping with a digital crossover to a pair of NHT1259 sealed subs.
I was thinking the same thing, that is if i can get horns to cover the area down to at least 150-200Hz.. I may build a tube based active LR, just for the phun of it.. or 1/2 tube.. using tubes for high pass and op-amps for low pass.. :xeye: just one of many thoughts running through my head...

It usually takes individuals _years_ of tweaking, trial and effort, trying and discarding various efforts to get a horn system how they want it.

For some reason, that doesn't scare me.. :eek: better to mill a new horn for my system than building yet another speaker, that I have nowhere to place.. :nod:

What you really want to do is ensure that you have plenty of fun on the Journey

I think my journey has just started. :D I'll get the midrange horns milled very soon, when I get them I will try to get a decent compression driver for them off ebay, and measure them to see if they can be mated directly to a Fostex bullet horn tweeter or something similar.. Or I might mill tweeter horns for a 1 inch compression thingy.. I'll find out when I get my first measurements.

It should be quite an interresting journey, and I would not be suprised if it takes years.. :) when the midhorns is successfully mated to a high-frequency partner, I'll look more into what options I have on the lower end, and the feasability of a midbass horn using some sort of cone driver.. perhaps a Fostex..? :scratch:

You mention the use of the DX4 for your midbass :rolleyes:, I suspect I will put my money elsewhere in the system :scratch: , as the midbass range to be covered will only be from around 150Hz upto perhaps 500Hz-600Hz.. Any thoughts here will be appreciated although it's a long way to go before I make any midbass horns..

It's all about the _magic_ And _magic_ is elusive, poorly defined, and oh so hard to achieve.

That is exactly what I am looking for... magic... :nod: With the 300B SET blocks, delivering something very special.. perhaps out of this world.. I want them to have a descent way of transmitting this *magic* to the human ear... I suspect horns can create some of this magic I am looking for.. :)
 
Hi Rocky, all

Nice thread , this ;)

Rocky, horns are addictive so watch out :D
Ken is right,
if you start fooling around with horns and triodes
chance are you will never be finished.
That's part of the charm , tho, isn't it

Here's some small tips on things I've learned through the years.
All very subjective, YMMV and all that.......

If you can, go active, preferebly with a digital crossover.
Yes, I know, digital is evil, but still.....

Use a horn that is big enough for the intended frequency range.
A 500Hz tractrix horn will probably only load to 650Hz or so.
IMHO it also usually helps a lot to put the horn in a baffle,
sounds much better, even if it doesn't look so good.

have to run

cheers ;)
 
Horn system

Rocky,

Embarking on this pathway does have its drawbacks. The drivers are not cheap, the bandwidth is small, the horns large, and the bass units needed to keep up with horns are limited and expensive. You will need good crossovers and at least two, possibly three good amps, and the amount of gear and boxes takes up most of the house. So, consider this as a warning.
How's that for encouragement?

On the other hand, the realism and dynamics from a well done full horn system can be stunning - unmatched by any conventional direct radiator system that I have heard, and I have heard a lot. Much lower distortion and lots of headroom may account for it.

The 2" JBL 375 and later counterpart 2440/41 are a good mid drivers. I have the 2440, which is a very large and heavy driver that sounds very good on the right horn. Bandwidth is about 500 Hz to 8k Hz. The TAD 4001 is even more extended, with HF out to 16K perhaps, but very costly.

Get a good 107 - 109 dB super tweeter or 1" driver/horn to go with the JBL, and it's off to the races.

Once commited to a full blown 2" compression driver and horn system can lead to complications.
The main problem then becomes reproducing the midbass and bass portions of the spectrum properly. Even with dispersion issues managed at the crossover, conventional midbass boxes loaded with direct radiators sound different...less good. They can be made to work though, and most people choose this direct radiator midbass/bass path as the fastest way to get a system up and running.
TAD makes good stuff and their 12" midbass is no exception.
JBL's 2226 15" woofers are excellent drivers, and commonly used for just such a purpose.
But an 80 -100 Hz horn with a good driver will sound better, if properly done.

Just a few thoughts...

Tim
 
Hi,

haven't dared to venture into horns (yet?), I'm on the dipole/open baffle side. I always thought dipole bass units should be a good match to horns. Why, because it's the only way to achieve a more directional bass with a direct radiator. It's not likely as directional as a horn, and it is on the opposite side of the efficiency/power requirement side, BUT it will have a dispersion closer to the matching upper register horns.

Just a thought.

(strangely I have never seen anybody trying this; commercial and DIY projects invariably use closed box woofers below the horn).
 
MBK said:
Hi,

haven't dared to venture into horns (yet?), I'm on the dipole/open baffle side. I always thought dipole bass units should be a good match to horns. Why, because it's the only way to achieve a more directional bass with a direct radiator. It's not likely as directional as a horn, and it is on the opposite side of the efficiency/power requirement side, BUT it will have a dispersion closer to the matching upper register horns.

Just a thought.


Hi MBK

I've been using open baffle bass with horns for years now
works fine, I'll probably never go back to ordinary boxes.
You'll have to go active, tho

cheers ;)
 
Tim Moorman said:
Rocky,
Embarking on this pathway does have its drawbacks. The drivers are not cheap, the bandwidth is small, the horns large, and the bass units needed to keep up with horns are limited and expensive. You will need good crossovers and at least two, possibly three good amps, and the amount of gear and boxes takes up most of the house. So, consider this as a warning.
How's that for encouragement?

Encouragement? not really.. However, I would like to approach this without filling the house with equipment. The ultimate setup for me would be 3-way horns from 150Hz and up, all powered by a single 300B monoblock per side.. Then actively cross it down to a dedicated subwoofer with built-in amplifying such as a gainclone..

To get away with smaller midbass horn, I would probably go for a corner placement, creating a rack supporting the 3 horns.. This will ofcourse take up more space than most conventional speakers, but I would like my setup not to be too dominating..

By getting the horns milled from wood, I guess I could also keep an esthetic appereance (my personal taste ofcourse), even with a 60-70cm diameter midbass horn. If i impregnate the horns with clear laquer with the wooden colours shining throuh, putting some green plants around the setup and lighting it up with a spotlight, I can even imagine a suitable WAF :D

BTW it would be one heck of a job to get a 70 cm diameter horn milled... I wonder if anyone in NORWAY has a lathe that is big enough and powerful enough for such a task....

______________

edit: drikk ølet ditt før det blir varmt...
 
Re: Re: Re: Wide range horn confusion..

You mention the use of the DX4 for your midbass

I would call that the lower mid range - and much of the fundamentals of the human voice occur beneath 1500hz - a cone seems very natural and smooth through that range to me - I am attempting to keep crossover point as high as I can within the horn loading range of the Azuras while shifting to the DDS and compression driver at the top of that range - I'm actually crossing at 134 hz right now - so 1250 puts me under a decade with a little leeway both directions

Rocky said:

as the midbass range to be covered will only be from around 150Hz upto perhaps 500Hz-600Hz.. Any thoughts here will be appreciated ..........

This is where you need to spend the most thought and be very careful of your design.

I have no answers for you here - but Tim is correct

Tim Moorman said:


3 octaves is a lot on a horn, but possible. A decade is even better, and nearing the optimum. Beyond a decade one wonders if the horn is serving any useful purpose in directivity.

Tim

If crossing at 150 3 octaves is 1200 and a decade 1500 - What I am saying use the full range or much of it or you have another horn, another crossover, another driver to match -

I do suggest that you select the drivers and horn combinations that will cover the range of approx 150hz to 3k as a pair and how they will work together with each other, crossover points, level matching, crossover slopes types, etc.

This will be the heart of the system - and _must_ be right.

You must consider, dispersion characteristics (particularly where crossover slopes overlap), sensitivity, chararacteristics of those slopes, lobing, phasing issues, alignment for phase and/or timing, etc. , not to mention whether you want to go with cone drivers ala Oris/Azura or crossover around 500 with a compression driver

The Oris/Azura are relatively new concepts, we're still learning to maximize them - you lose the benefit of resolution of the compression driver 500 hz up and yet gain point source coherence above that - I am embarking on an attempt to add the 1" waveguide - I am unaware of that having been achieved successfully or even having been tried previously. which is why I have been learning a lot about this _grin_

Tim has pointed out the largest problem and compromize area of 2" compression drivers where it is difficult to do the 500hz down part

Rocky said:
I suspect horns can create some of this magic I am looking for.. :)

I and lot of other people think so also. _big grin_

Tim has fooled around with a good bit more of this than I have - read his comments carefully, he knows a good bit about horns.

Actually, there are a number of posters here that do.

regards

Ken L
 
Re: Re: Re: Re: Wide range horn confusion..

Ken L said:

I do suggest that you select the drivers and horn combinations that will cover the range of approx 150hz to 3k as a pair and how they will work together with each other, crossover points, level matching, crossover slopes types, etc.

Ken L also says:
Tim has pointed out the largest problem and compromize area of 2" compression drivers where it is difficult to do the 500hz down part

Ok, Tim. I would be grateful to hear your experienced opinion of the best companion below a 2 inch compression driver most likely being a JBL. I must set the limit of 150Hz where no horns are allowed below.. XO point preferrably between 500Hz and 600Hz

Tim said:

But an 80 -100 Hz horn with a good driver will sound better, if properly done.

What sort of driver/arrangement have you used in such a horn setup and with what results?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.