Rerurn to Romy the Cat's Site

Horn-Loaded Speakers
Topic: Horses for courses

Page 1 of 3 (75 items) 1 2 3 »


Posted by Rewind on 09-06-2013
fiogf49gjkf0d
My work to make the Fane S 8m work, continues amd I have yet to hear my midbasshorn built perferctly according to Romy-specifications. I will continue because in early trials there was something I did like. I guess I will just end up with a measly €48, 000 midbass horn, but I fear it will pale in comparison to the Goto S-150. But even the Goto horn has shortcomings. It just goes to 200 Hz, and 300 Hz if I want to be safe in every song. Lowpassed at 600Hz. It leaves room for improvement, that is safe to say. Is it time to leave questions like "But where will you put it, you idiot, I mean - darling?" behind us and look forward? Compression drivers at 100Hz should be the goal.

Posted by noviygera on 09-07-2013
fiogf49gjkf0d
For me personally, midbass is the most important frequency range when it comes to sound reproduction. I believe you will be on the right track to focus on this range because so often it is overlooked and treated with superficial attention, as a "it's there somewhere" to be in the sound not too loud or boomy or obtrusive. But if we focus on this range and listen to it only, we will see that most of the emotion of sound resides there and around it, that is the tonality that makes the sound system special or vague.
I have been thinking about this region of sound and what if we start the system build with this step, what are the options? Has anyone done this and what were the observations of comparing different topologies. For example, here I see a few approaches, each specifically focusing on the midbass: 
1. high efficiency: some well made horn around a single specific paper cone driver. 2. high efficiency: some well made horn around a compression driver 3. low efficiency: array (rectangular) of direct radiators, 6" or smaller drivers, let's say 30 drivers, as a point source 4.  low efficiency: wrapped  array  of direct radiators forming a frame around the mid/high section (like a picture frame around a painting)
Has anyone compared the above midbass topologies? 
Gera

Posted by noviygera on 09-07-2013
fiogf49gjkf0d
lately I've come to theorize that the horn midbass may be a contadictive dead end approach. The reason is that proper horn midbass (lets say 80 to 400Hz) will be a big one and present a few inherent "brick walls". It will physically shade the midrange or high sections. It will throw a coverage pattern that is sourced from one point and be possibly localized at the upper end of it's frequency range, if using 1st order crossovers.
Let's picture another approach. An array of small, capable direct radiators, let's say 5" or 6" paper cones in sealed enclosures that do well 80 to 400Hz. Physically time aligned with the horn mid/high sections, hugging from all sides. A row below, above, right, left of the mid/high horn. A frame of midbass speakers. Efficiency will be low, but many advantages -- no shading, a stable sound field, good pattern control and low room interaction vertically and horizontally.

Posted by Jorge on 09-07-2013
fiogf49gjkf0d
Hi Gera

"Has anyone done this and what were the observations of comparing different topologies."

I have done just what you propose,  I tried the JBL2490 compresion driver with a 3" mouth in a 110 hz horn well made, MDF round tactrix horn,  it goes nicely down to 200 hz,  I run a couple of test tones with no xover and the most the horn would load it was to 200hz,  I was looking mainly for a response to match the compresion driver on top, crossing at 500-800 hz,  The sound in that frequency was clear and transparent,  not as dynamic as I would have expected, which could be a good thing.  The tone of the lowest frequencies it made was not too full or rich, it was actually pretty bland,  I had to cut the bass drivers a little higher to fill out the response.  If you play it without a woofer under it you will clearly miss part of the music, even a string quartet it will sound thin.
Certain female voices were phenomenal.
I had tried before that a lot of 8 inch to 10" paper cone drivers in a similar horn, round tactrix 110 hz MDF horn with recalculated throats for each driver.  The difference was obvious,  the compression driver was more detailed but would go no lower, the paper cone driver would go much lower,  down to 110-120 hz:  Now the difference between 120 hz and 200 hz is a big one (even 150hz)  there are so many thing happening at that range that most instruments and voices get "cut" in their response and the lower registers have to be played by the woofer.  This creates a step in which the horn is playing one part and a woofer solution, which ever you have, will play the rest of it ussualy with a very different tone and speed (if I may use the term).  You can tell.   Now if your horn makes it down to 120-110 hz this small diffference makes for most of the instruments and voices to be played more by the horn,  the step gets minimized  and the performance is much better.
Then I tried the Studio 8M, it is the best of both worlds, it plays clearer and more transparent than the compression driver (2490) and it goes even lower than any other paper cone driver I tried, with a great tone.
I tried an coupple of arrays.  One had 16 woofers, 4" aerogel small woofers per side in a sealed cabinet: The sound is indeed very strong and punchy,  the efficiency goes pretty high to the point that I was moving them with a 1.5 watt amplifier,  but it did better with more watts.  This array would go lower down to around 70 hz.  The quality of the sound, to non horn user was amazing (a close friend kept them for himself)  but when comapred to a "well made horn"  it became slowish, undefined with a darker tone.  When I changed to the 120 hz horns it was like turning the lights on.
I also tried a bigger array of 5 woofers per side each being 10".  It would not compare to any of the above at the 500 hz frequency.  I finally had it playing from 27 hz down in a Infra sub configuration with a Marchand active xover: WOW! that was very nice...
All the Best

Jorge






Array and horn.jpg

Posted by Jorge on 09-07-2013
fiogf49gjkf0d
Array 10.jpg

Posted by noviygera on 09-07-2013
fiogf49gjkf0d
Jorge,
interesting findings. However, the tall arrays with small woofers is actually worth exploring further based on your findings. I am theorizing, but perhaps the "smearing" effect you were finding in comparison to the midbass horn was due to the height of the array and differing distances it creates to ear and also room reflections. This is exactly what I was trying to avoid by creating a "frame" array, by bringing the source closer together and around the other channels.


Posted by scooter on 09-08-2013
fiogf49gjkf0d
This is good issue to raise when employing lots of drivers. Maybe four years ago there was an interesting thread here noting the challenges this brings. Statistically it is just a mess although that is not the whole story. In this context, those posts are worth revisiting. 

Posted by Rewind on 09-08-2013
fiogf49gjkf0d
Crossing the woofer so high is not really a fair comparison. The compression drivers needed to be in a bigger horn, and possibly be replaced with a driver that does 100Hz for certain.

Posted by Jorge on 09-08-2013
fiogf49gjkf0d
There are a few at least theoretical backdraws for the Array in my view.

Of course having a lot of different drivers playing at the same frequency implies that all of them should work in perfect unison,  if one starts lagging a little bit it will mess everything up.  I dont know how perfect drivers can be made or quality control issues and what parameters affect exactly what they can play, but it seems to me there is a lot of room for error.  Soundwise, they are not too bad actually.

Be careful about time alignment, I once thought it would look cool to have a slanted array of drivers,  what a mistake, it sounded just awful all drivers perfectly NOT time aligned.  I am afraid this will happen with your "square" configuration if you try to toe in your speakers:  But it is always good to make it a see for yourself.

Rewind, The compression driver that would go the lowest that I could find was the JBL 2490, the specs have the FS at 250 hz.  I placed it inside a huge 36" across Horn and I made it go down even lower to 200 hz  the horn should have loaded it down to 120 hz.  Maybe not the best sound from it but the idea was to see how low it could go and how it would sound.  The only drivers I know off that can go lower are the Goto SG146 LD  specified from 20 hz to 500 hz. It is a monster and that would require as you say a much bigger horn.  There are also the GOTO 500 series specified from 100 hz up to 8 khz which should be wonderful for this application,  but I have heard from owners that it does not go as low as the JBL2490 does.   Unluckily I dont have those Goto drivers around to play with to confirm this,  but the SG 146 LD is on my very short list.


Posted by Scott L on 09-09-2013
fiogf49gjkf0d
Gentlemen:

I've thought about this mid-bass challenge for quite some time. I have tried many approaches.
For the sake of my example, let's assume mid-bass to cover from 60-600Hz. That just happens to
be the classic "decade". I have to fully agree with a previous poster who said a single unit horn large
enough to do that, is just too big to physically fit in with the rest of a multi-diver, multi-horn set up.
My idea is to treat the mid-bass as a large 2way, yet still to be driven by the output of one section of
an active crossover. The output is to be fed into two different amplifiers. In order to match acoustically
with a mid-range horn, I intend to use an Altec/VOTT type short horn. I'll be lucky to get horn loading
down to 150Hz, and that's fine as I intend to also use augmenting direct radiator drivers below this.
At least 2 more 15's per side, but probably 4. In order to keep the audio signal away from these drivers
that the Altec already covers, I'll use EQ. These additional 15" drivers are NOT subs, they are augmenting
mid-bass. The subs are already exisiting in an infinite baffle manifold arrangement. In order to play back
the sound of a Grand Piano, ya just gotta move some air. This should do it. Happy listening to all.

Posted by Rewind on 09-09-2013
fiogf49gjkf0d
You said are lucky to get hornloading down to 150Hz. Then I guess you suggest you will end up at maybe 175Hz. Hornloading a compression driver down to 175Hz is not too difficult and the horn will not be larger than the Avantgarde Trio midbass horn, at least not in diameter. Horns and compression drivers go together, cone drivers and horns, not so much. For bass, yes, maybe. Bass is not so detailed for our ears anyway. In the area our ears are detailed, why not use a driver for our horns that is built for this purpose. One can pretend a cone speaker is compression drivers but it will not be the same. If you are dead set on cone drivers at ~200hz, then maybe do direct radiators everywhere. The Avantgarde Duo had a famously poor bass because of the mix of direct radiators and horns at this area.
Compression drivers under 200Hz is still a theory for me, and I will have to verify it after a few weeks in the garage.

Posted by Romy the Cat on 09-10-2013
fiogf49gjkf0d
 Rewind wrote:
You said are lucky to get hornloading down to 150Hz. Then I guess you suggest you will end up at maybe 175Hz. Hornloading a compression driver down to 175Hz is not too difficult and the horn will not be larger than the Avantgarde Trio midbass horn, at least not in diameter. Horns and compression drivers go together, cone drivers and horns, not so much. For bass, yes, maybe. Bass is not so detailed for our ears anyway. In the area our ears are detailed, why not use a driver for our horns that is built for this purpose. One can pretend a cone speaker is compression drivers but it will not be the same. If you are dead set on cone drivers at ~200hz, then maybe do direct radiators everywhere. The Avantgarde Duo had a famously poor bass because of the mix of direct radiators and horns at this area.
Compression drivers under 200Hz is still a theory for me, and I will have to verify it after a few weeks in the garage.
Holy foolishness!

Posted by Rewind on 09-10-2013
fiogf49gjkf0d
I will check it all soon and come back. Smile Then I will know how foolish it was.

I wanted to see if anyone had any new ideas.

Posted by Rewind on 09-10-2013
fiogf49gjkf0d
Turns out we were in agreement in 2009: http://www.romyhome.com/Forums/ShowPost.aspx?PageIndex=2&postID=11289#11289

Have you changed your mind about such a horn that you propose, or do you have any more hints before I build it?

Here is a nice example made by Josh Stippich who has a cool shop in Saltlake City. Scroll down to the red horns: http://www.audioheritage.org/vbulletin/showthread.php?8382-Horn-system-pictures/page6

Posted by Romy the Cat on 09-10-2013
fiogf49gjkf0d
 noviygera wrote:
For me personally, midbass is the most important frequency range when it comes to sound reproduction. I believe you will be on the right track to focus on this range because so often it is overlooked and treated with superficial attention, as a "it's there somewhere" to be in the sound not too loud or boomy or obtrusive. But if we focus on this range and listen to it only, we will see that most of the emotion of sound resides there and around it, that is the tonality that makes the sound system special or vague.
I have been thinking about this region of sound and what if we start the system build with this step, what are the options? Has anyone done this and what were the observations of comparing different topologies. For example, here I see a few approaches, each specifically focusing on the midbass: 
1. high efficiency: some well made horn around a single specific paper cone driver.2. high efficiency: some well made horn around a compression driver3. low efficiency: array (rectangular) of direct radiators, 6" or smaller drivers, let's say 30 drivers, as a point source4.  low efficiency: wrapped  array  of direct radiators forming a frame around the mid/high section (like a picture frame around a painting)
Has anyone compared the above midbass topologies? 
Gera
Gera, mid and upper bass is one of few elements that distincts high-end from HF-FI. You do a mistake however in my view to attenuate good mid and upper bass to a specific topology wrather then to specific implementation. I have seen interesting mid and upper bass made by horns and not by horns and I do not think that the topology matters. What matters is the head of system owner and his ability to put it all together aming the proper objective. It happens that at this site people mostly deal with horns. I am sure good mid and upper bass migh be made with direct radiators but you need to look somewheres also as I think this site has no constituencies for direct radiators. The paragon of Fane in 4" throat and 125hz horn is quite uniq. It was given as is as many people feel that this is another outcome of many DIY-audio craziness. That is not correct and it requered a lot of practical expireance to understand what it as and why there are very few good topological moves if horn topology was chosen.

Posted by skushino on 09-27-2013
fiogf49gjkf0d
I'm also of the opinion that the bass is one of the most critical and most challenging parts of playback. Not talking about quantity, but rather the quality of bass reproduction. My approach has been to split the bass in two parts - Low and mid bass covered by Danley's SPUD tapped horn (< 90 Hz), and upper bass / lower mid served by a 140 Hz tractrix horn (between 100 Hz - 800 Hz). This approach is satisfactory, but I believe there's room for improvement by adding a dedicated mid bass channel to cover ~ 50 Hz - 250 Hz.

I'm finally starting on a pair of mid bass horns. Why tackle the effort of building a huge pair of horns rather than a different solution? I chose horn loading to match the sensitivity of the other channels. But most significantly, bass reproduced by an 'adequately large' horn is unique compared to other solutions. Listen to cellos, standup bass, tenor / barry sax, or piano on a big horn like Tannoy Autographs or Klipschorns horns and you might agree. One of many challenges is making a horn work in an average-sized residential listening room. A straight horn is not an option due to size considerations (short of heroic built-in solutions like Romy's). BLHs can sound fine but have freq response irregularities. I've decided to build a folded horn based on the University Dean. The drivers are C-15W dual-impedance woofers with huge alnico motors and paper cones. I hope to have them finished in time for the holidays.

 


Posted by Romy the Cat on 09-27-2013
fiogf49gjkf0d
Can you put your thought in sketch? There was image with your post that was not viewable, I deleted it. what you propose is doable but it needs to be visualized

Posted by skushino on 09-27-2013
fiogf49gjkf0d
Sure, Romy.

The commercial product has both midrange and tweeter horns inside the mouth.  My plan is to use the bass horn on it's own, as a dedicated bass channel.  The plan appeals to me because there are no folds, among other reasons.  The expansion is a gentle 'C' path.

Here you go....

(no image)

Posted by skushino on 09-27-2013
fiogf49gjkf0d
I was rushing when I made the last post.  Wanted to share a little more about my thought process and how I arrived at doing another large-scale masochistic audio project.  Making a bass horn is a huge risk - the results can vary from exceeding expectations to ending up with something that doesn't work at all, or just partial success.  One won't know the outcome before building and listening.

I'm currently using two channels for bass - tapped horns from the bottom up to around 100 hz, and a 142 hz tractrix horn w/ 4" throat covering around 115 hz - 800 hz.  It's well known that tapped horns have benefits in form factor and packaging, but they sound progressively worse the higher they go.  I can hear the benefits when I listen with the crossover set lower and lower.  Naturally I can only set the crossover point so low before a hole opens up between the THs and the upper bass horn.  I would like to end up with the THs limited to a narrow bandwidth, around 20 - 50 hz.  In this range they work fine, pressurizing the room per expectations.

Another goal is to raise the 'center of gravity' of sound presentation.  The horns are presently set-up with the upper bass horn physically lowest, with the mouth resting on the floor.  The mid and treble horns are above the larger upper bass horn, around 36" high.  This works ok, but I think there's an opportunity to improve the sound presentation by raising the overall height of the speakers, so that the apparent center of gravity (center of acoustic mass?) is taller.  Re-arranging the horns so that the folded bass horn serves as a base for the other horns, located closest to the floor, and raising the 142 hz horn up high, closest to the ceiling, should move the entire sound stage up, for more life-like sense of scale and size.

Another consideration is how to power everything.  I played around with different amps dedicated to single or double channels.  I'm generally happier with one amp powering all.  My amps are ML-2s.  Operating in the SET world, the bass channels need to be 'SET friendly' - benign impedance and sensitivity congruent with an 18W tube amp.   I considered building a bass tower with direct radiators in sealed or ported cabinets.  I think these could be made to work acceptably well.  But the horn topology seems more congruent with the other horn channels and the limitation of using a single channel SET power amp.

So a folded horn seemed to be an acceptable compromise for my application in my 16' x 18' listening room.  The Dean bass horn is elegant in it's simplicity, and has an adequately sized mouth, throat, and path length for the freqs I wanted covered.  There are no sharp bends in the path.  The numbers for length and area are acceptable (but they too are compromises).  And it passes the 'intuitive' test - the cabinet measurements are way too big for my standard sized listening room, which means it could be large enough to work as a proper horn over most of it's range.  So the cabinet dimensions are adequate - how about the driver?  The C-15W makes my Altec 515s look like toys. The magnet is massive.  The frame is massive and strong.  And the cone is all paper, including the stiffly-suspended surround.  The TS numbers seem reasonable (although I haven't measured my pair, relying on second hand measurements).

I hired someone with wood and metal working skills to do the actual fabrication per my guidance.  He's a quirky character, artsy and sort of a loner.  He has a strong sense of aesthetics and seems to be as excited about the project as I am.  I've positioned the project as a combination of wood and steel media, aesthetic proportions, textures, and colors, and sound and music combined in to a grand sound sculpture.

The pieces are in place to move forward.  I'm committed to writing the check, and having a significant interruption in my life for 2 - 3 months, all without any guarantee of the results.  It feels like that first step jumping out of an airplane with a parachute in to the void, or boosting a jump down a mountain on skis, deep out of bounds.

Here's to soft landings.  



Posted by Romy the Cat on 09-28-2013
fiogf49gjkf0d
Scott, you have to find another way to post the pictures, the site has multiple options to do so but you still insist to load a binary sting – sorry it does not work. It would be nice to see the idea behind the  Dean bass horn and your sketch how your installation will look like after it will be complete. If you can’t load image to the site then send me the image and I will post it for you. I find your observations about the tapped horns are interesting. Your fight for the gravity center is interesting but there are other ways to deal with it, particularly if the crossover points and the locations of the channels are known. I would not worry at this point about amplification.

A folded horn is a compromise but no one said that it might not sound interesting. I have seen better and worse result and I personality do not know what make it better or worse.  I do not know the C-15W drivers. The massive magnet might not be massive as it might be just a faux cover, not to mention that it shall not be “massive”. The massive magnet is good for high exertion but you do not need for horn loading any horn loading any exertion. Anyhow, only god knows hot any driver will sound being loaded into a horn, this is a pure lottery.

I know that you were working in IT project management word and if you remember the people with brain in there begin to test the systems when user requirements and functional specifications are collected.  I would encourage you do the same with your horn project. Test your planning and your objectives before rendering them.

The Cat

Posted by Romy the Cat on 09-28-2013
fiogf49gjkf0d
II would like to pass a word of warning about the C-15W driver. I do not know this driver but it looks like it has dual voice coil. This might be a hell or blessing. On a paper the idea sound very interesting but unfortunately when I seen it it never worked. University had M4 driver with dual voice coil and it truly very bad sounding driver. You need to make you own subjective judgment about the C-15W driver. Put it in 3x5 or 2 x4 open baffle fame and listen it for a while.  Use it if you like it. Still, make a provision to be able to put in your back chamber another driver if C-15W will not work out for you.

Posted by Paul S on 09-28-2013
fiogf49gjkf0d
Scott, it sounds like you have properly tempered your expectations and steeled yourself against the very real possibility of "problems" over the course of developing this channel.  Is the Dean originally designed +/- as a corner horn, and do you plan to use it +/- this way?

I would not say the 515s are toys, rather people seem to insist to make them play ULF, which they are not suited for.  In the A7 bins, they dropped off like a rock below 50 Hz, and to push them below this will at least take exponential power, along with the "special" X/O, and all that comes with that.  I don't know the other driver you mention, or how it would suit that cabinet. No doubt you have done your homework, and you have built in ways to accommodate it over your desired frequency range.  Anyway, from what you say, it sounds to me like you will be making this channel the "pivot" of your speaker, and if this is the case, then you can probably make it work, one way or another, once you get it voiced to suit yourself.

Best regards,
Paul S

Posted by skushino on 09-28-2013
fiogf49gjkf0d
Another very quick post - I have guests arriving from Japan in a couple of hours.  Will respond more thoughtfully when I have some free time.

University Dean plans: 

Mouth - 1080 in^2

Throat - 78 in ^2

S(d) - 132.7 in^2

Compression ratio - 1.70

Path length ~ > 1.3 m

If my math is correct, the profile should be valid for horn loading down to around 55 Hz


university dean plans-2.jpg

C-15W Dual Impedance Woofer:

This woofer does in fact have dual coils.  Agreed this is one of many unknowns affecting the sound.  I like the idea of having the ability to vary the impedance as another tool in shaping the final sound.  In addition to these drivers, I have others to audition in this horn, including Altec 515G and EVM-15.  The back chamber in the horn appears adequately large to optimize for various drivers.

University C-15W.jpg

C-15W Nude:
University C-15W Nude.jpg

Dean Horn Exposed:
Dean Horn Exposed.jpg




Posted by Jorge on 09-28-2013
fiogf49gjkf0d
Skushino, I found the same thing, TH can only go so high before they start sounding :funny, of course due to the principle involved, if you want to raise the response further from the design the same Tapped Horn begins to work against itself.  So one option could be to redesign a tapped horn that would go higher.
Now you state that your TH can go up to 90hz with good sound.  But your UBH only goes down to 140 hz,  maybe a little lower due to the fact that it is coupled to the floor.
If you make your UBH bigger say 110 hz- 100 hz it would go lower with the same sound you already like.

I fnd it hard to go into unchartered territory trying to make, get to sound good and then integrate a horn that you have no idea how it will sound.
I had for a couple of years a pair of Hartsfield top loaders,  I hated them...

Posted by skushino on 09-28-2013
fiogf49gjkf0d
Guest napping now from 16 hrs jet lag, so I have a moment...

Jorge, you raise an interesting point about THs.  I may be incorrect in assuming THs sound worse as they play higher in freq.  Perhaps instead the limitation is one of narrow bandwidth rather than frequency.  In other words, having my TH cover 20 - 100 hz, more than 2 octaves, may be a bridge too far.  20 hz is a very different operating point compared with 100 hz.  Sure the system works, but to my ears definitely sounds better when i fix the high pass around 18hz and lower the crossover from 90 to 70 hz.  I have a hunch one could design a TH suitable for playing higher frequencies, but in a narrower BW.  But I can't think of any practical reason to do it.

No doubt a 100 hz UBH would be fantastic compared with my 142 hz horn.  We always want our bass horns to be a little bigger, right?  But using the horn to play lower comes at a price - I would want to lower the upper crossover point too, meaning my MR horn playback needs to go lower.  And unless I made an even larger folded horn to go lower than 50 - 55hz, the larger UBH just squeezes the operating bandwidth of the folded horn narrower, in this case 50 to around 80 or 90 hz (with boundary and room gain applied to UBH).  In that case the effort of building a dedicated mid bass channel is hardly worth operating in a bandwidth less than an octave.  So I think if one is using a dedicated mid bass channel under the UBH, maybe 140hz or similar is a good decision for this application.

bty, thank you for sharing the results of your mid bass experiments -  your experience is educational for me, and you're one of the few with an opinion informed by real world experience.



Page 1 of 3 (75 items) 1 2 3 »