Tapped Horn with Dayton SS460HO-4 18" driver - AVS Forum
Forum Jump: 
 
Thread Tools
post #1 of 21 Old 07-04-2013, 05:13 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I'm about to pull the trigger on a tapped horn build with the Dayton Reference 18" drivers. Any constructive criticism or advice is welcome and appreciated (especially before I start cutting wood). This is my first tapped horn build, and I'm by no means an expert at this stuff. I bought the drivers during their Memorial Day sale, and have been pondering what to do with them since then.

I think I have a tapped horn designed here that should work reasonably well. I am pretty lazy when it comes to building speakers, so I have made an attempt to limit the amount of cutting required. As such the box is 8' tall. Internal width is 18.75", with the remainder of that rip (~29.125") forming the sides.

Intended use will be 95% music and 5% movies. Power will be from a Crown XTI 2002 which is around 800W @ 4 ohms. There's some DSP filters available in the amp to smooth the response if needed. A 48db/octave high pass at 20 Hz will be used. I'll probably have the subharmonic synthesizer enabled with a fairly low setting around 2.5 dB below 45 Hz. The subs will be crossed over around 80-100 Hz to some mains with JBL 2226 woofers and SEOS 12 waveguides.













batdorf9 is offline  
Sponsored Links
Advertisement
 
post #2 of 21 Old 07-04-2013, 05:14 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
reserved for future stuff
batdorf9 is offline  
post #3 of 21 Old 07-04-2013, 05:14 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
reserved for even more future stuff
batdorf9 is offline  
post #4 of 21 Old 07-04-2013, 09:00 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,787
Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 528 Post(s)
Liked: 1147
this may be worth reading. it was a discussion on the viability of the 18ho
in a tapped horn of quite similar design to what you have there:

http://www.avsforum.com/t/1458644/gjallarhorn-vs-little-wrecker

btw, the 18ho makes a decent front loaded horn
if you want to go that route.

Listen. It's All Good.
LTD02 is online now  
post #5 of 21 Old 07-05-2013, 10:37 AM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by LTD02 View Post

this may be worth reading. it was a discussion on the viability of the 18ho
in a tapped horn of quite similar design to what you have there:

http://www.avsforum.com/t/1458644/gjallarhorn-vs-little-wrecker

Yes, I had read that thread before and certainly gave it some consideration. Regarding the compression ratio comments, I found that both the Othorn and the Gjallarhorn have compression ratios higher than 3.7. What I'm proposing to do here has a compression ratio of 3.19. Here's a summary of compression ratios I calc'd from the well-known existing designs out there:

SD, min S2+, Comp Ratio
wolfhorn sdx, 660, 275.29, 2.397
lil wrecker, 823.72, 315.77, 2.609
house wrecker, 1647.44, 558.14, 2.952
gjallarhorn, 1194.6, 308.14, 3.877
othorn, 1680, 400.79, 4.192

I did experiment in HornResp a bit with some other drivers that had higher ratios of BL^2/Re/Sd (LAB15, various Dayton Ref 12's and Ref 15's), but none worked out as well as the Ref18 in the HornResp simulations. Output per $ seemed pretty high with the 18. So is there something HornResp isn't modeling here, or some other limitation?
batdorf9 is offline  
post #6 of 21 Old 07-05-2013, 07:09 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,787
Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 528 Post(s)
Liked: 1147
"So is there something HornResp isn't modeling
here, or some other limitation?

ask lilmike. i'm pretty sure he did some comparisons
with drivers in tapped horns that had high and low
power motors bl^2/re/sd and found that he and others
could tell instantly that the higher powered motors
sounded much better, much cleaner.

that wouldn't show up in hornresp as far as i know
either, which is why i mention it. your model looks
good "by the book" and if lilmike hadn't mentioned
his experience in this regard, i'd give your design
the thumbs up.

on the other point, i don't see why 3:1 compression
would be a problem for the driver.

Listen. It's All Good.
LTD02 is online now  
post #7 of 21 Old 07-05-2013, 07:11 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,787
Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 528 Post(s)
Liked: 1147
one other minor bit, i think that deep notch just over
100hz comes from the length of horn behind the driver
into the very narrow corner. moving the rear "wall"
up to right behind the driver should push that cancellation
up much higher in frequency where it won't be a potential
problem.

Listen. It's All Good.
LTD02 is online now  
post #8 of 21 Old 07-06-2013, 11:55 AM
AVS Special Member
 
lilmike's Avatar
 
Join Date: Dec 2009
Location: Pacific Northwest
Posts: 2,193
Mentioned: 0 Post(s)
Tagged: 1 Thread(s)
Quoted: 18 Post(s)
Liked: 87
First - a disclaimer:

The following is presented based on my own empirical observations, by someone with no formal training in acoustics. I'm a hobbyist, I don't do this as a day job.

OK, that being said, let me see if I can help a bit here.

First off - it is great that you're taking panel utilization into mind. My experience suggests that it's a good idea to give yourself a bit more room to work though, that driver will be a tight fit. That being said - why are you wasting all that space inside? Make the horn bigger by increasing the rate of flare, there is no reason to block the space off.

Re the model:

Model with PAR flares, because that's what you'll be building.

A simulation at 2.00 volts is a fair 1 watt comparison, 2.83 is 2 watts with a 4-ohm driver.

Subtracting 3 dB from your plot, it looks like your modeled efficiency is somewhere near 94 dB at a watt, The driver is 89, so you're seeing some efficiency gains, that's good.

L12 is far too long. The impulse response can be cleaned up a lot.
Enter 0.00 for both APT1 and LPT. I don't think they matter in a tapped horn model, but it's a housekeeping thing.
Calculate your Vtc and Atc, then enter those values. The recent versions have a wizard that makes this easy to do.
Finally - look at your acoustic reactance plot when you calculate the model. Like all the other data presented in Hornresp, it matters. In a nutshell, big tall spikes are bad, we want the heights to be similar.

Re the driver (assuming I get the math right this time):
BL = 20.57
re = 3.41
Sd = 1090

BL^2/re/Sd = my "magic number"

In this case, my "magic number" is 0.1138. This means that based on my experience and observations, this driver falls on the under-motored side of the spectrum for use in a tapped horn.

That said, will it work? More than likely, but just cause you can doesn't always mean you should. To be totally honest, nearly anything can be made to make noise in a tapped horn, but there are reasons I choose the drivers I do.

Is it optimal? No, not quite, but it really isn't that far off. Qes is a touch high, and I'd like to see a BL nearer 25. Still - it will work, and if the horn is right for the driver, it should work quite closely to the model's predictions. Qes is not that high, so with a good amp, and the driver in a proper horn, things should be fine.

I've built horns using drivers that were undermotored, I've built several that were overmotored, and I have a few that used motor to cone ratios that fall in the sweet spot. I much prefer the sound of the horns with enough (or even too much) motor, and I much prefer the sound of a proper amp. I will definitely explore motor force further, I don't have a lot of spare cash for driver purchases at the moment, nor do I have any shop time for the builds and testing.

I really look for the magic number to be higher than 0.15, but no higher than 0.25. A lot depends on what I am trying to do with the horn though, and there is a lot more that needs to be considered.

Also - keep in mind that the small signal parameters we're modeling with really aren't valid at anything other than small signals. I know we have to start somewhere, but do keep in mind that several of the parameters will shift significantly as power is applied.

Regarding compression ratio:
It's a grossly oversimplified way of looking at things, because the aspect ratio of the horn matters a lot too. The losses in the horn are something that Hornresp can't account for. Losses result from friction, and with more side wall relative to the sectional area of the horn, losses increase. The more squished the aspect ratio is, the lossier the horn becomes. These losses are non-linear, and will also vary with the power applied.

Remember, air can be compressed and rarefied. If the losses in the horn you build are high enough to where it becomes easier for the driver to compress and rarefy the air in the throat rather than resonate the entire air column, the horn ceases to function as a horn. This will not be evident in the model, but it will be painfully obvious in the results. I've built several of these, I don't mention them much other than referring to them as learning experiences. This is not just a result of too much compression at the throat - the entire horn has to be "big enough" for the driver. That's where the rules of thumb I mention in my tutorial came from.

Don't think that I am trying to sink your project, quite the contrary. If I were doing it, the model would look something more like the attached. The fold could be similar too, just use the space inside a bit differently.

rss460.txt 0k .txt file

Considering the excursion limits Josh identified, the driver should be fine at power levels up to the driver's RMS limit, provided you use a ~15 Hz highpass. Sure - it takes things past xmax, but the moving parts shouldn't slam into the stationary ones.

@ LTD02 I believe you're correct, that notch appears to be a result of the long L12.

edit: Completed my thought...oops.

edit 2: Added equation for clarity
Attached Files
File Type: txt rss460.txt (431 Bytes, 25 views)
lilmike is offline  
post #9 of 21 Old 07-07-2013, 08:11 AM
AVS Special Member
 
Martycool007's Avatar
 
Join Date: Jun 2011
Posts: 3,692
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 280 Post(s)
Liked: 172
Two questions for you, (lilmike or who ever else wants to answer)

First, how do you know if a driver is under-motored, over-motored, or just right? Does this matter as much in a sealed or vented design?

Second, how do you figure compression ratios of both sealed, vented, and horn based subs?
Martycool007 is offline  
post #10 of 21 Old 07-07-2013, 08:35 AM
Advanced Member
 
diaz's Avatar
 
Join Date: Feb 2011
Posts: 614
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
Quote:
Originally Posted by lilmike View Post

Don't think that I am trying to sink your project, quite the contrary. If I were doing it, the model would look something more like the attached. The fold could be similar too, just use the space inside a bit differently.

rss460.txt 0k .txt file

Do you think a front loaded horn would be better given the lower xmax variable? Or is the underpowered motor ok with tapped horn in this situation?
diaz is offline  
post #11 of 21 Old 07-07-2013, 09:46 AM
AVS Special Member
 
lilmike's Avatar
 
Join Date: Dec 2009
Location: Pacific Northwest
Posts: 2,193
Mentioned: 0 Post(s)
Tagged: 1 Thread(s)
Quoted: 18 Post(s)
Liked: 87
Quote:
Originally Posted by diaz View Post

Do you think a front loaded horn would be better given the lower xmax variable? Or is the underpowered motor ok with tapped horn in this situation?

The OP is talking about a tapped horn, so I addressed that, and provided some information to adjust the design presented.

As I look at it in choosing drivers, motor strength is not directly related to throw, it is the force that the motor can exert on the area of the cone, and is a product of the magnetic field strength and the amount of coil in the gap.

This driver is not "that" undermotored, and actually, has quite a bit of throw.

Define "better". All speaker cabinets are compromised, cause there are no free lunches.

You can design anything for any driver, but as I said, just because you can doesn't mean you should.
I would not have posted the tapped horn simulation I did if I didn't think it was sawdust-worthy.
lilmike is offline  
post #12 of 21 Old 07-07-2013, 09:55 AM
AVS Special Member
 
lilmike's Avatar
 
Join Date: Dec 2009
Location: Pacific Northwest
Posts: 2,193
Mentioned: 0 Post(s)
Tagged: 1 Thread(s)
Quoted: 18 Post(s)
Liked: 87
Quote:
Originally Posted by Martycool007 View Post

Two questions for you, (lilmike or who ever else wants to answer)

First, how do you know if a driver is under-motored, over-motored, or just right? Does this matter as much in a sealed or vented design?

Second, how do you figure compression ratios of both sealed, vented, and horn based subs?

My approach is empirically derived, the math I use is explained a couple posts earlier. I built lots of tapped horns, then measured them. I looked at what worked, and what didn't.
Then, I figured out what was different in the drivers I used and found a way to differentiate things based on simple math using the Thiele Small Parameters.

Motor force matters, period. The motor strength doesn't really come in to play in the selection of drivers for simpler cabinets, because the designs are far more forgiving, accepting a wider range of options with good results. In cases where the cabinets are more complex, the criteria range tightens. The more esoteric alignments like horns, bandpass and such can vary the loading on both sides of the driver's cone through horn or bandpass tuning tweaks, which isn't exactly possible with a sealed or ported cabinet. When you load the cone with an additional acoustic load like a horn or a bandpass, you've got to have enough motor to drive it.

Compression ratio is simply the area of the cone divided by the area that the cone is radiating into. Sealed and ported cabinets don't have a compression ratio per se, as the cone is exposed to the room, there is no compression. In horns and tapped horns, we look at the area of the throat at the driver midpoint. As I alluded to in my earlier post, the concept is a gross oversimplification, but it provides a starting point for design.
lilmike is offline  
post #13 of 21 Old 07-07-2013, 03:54 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 16,787
Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 528 Post(s)
Liked: 1147
i've been trying to figure out why you use
par to model a straight horn.

i just worked through the math and you are
right. even though par creates a parabola,
which is a curve, the cross-sectional area
is a linear function of the horn length.

son of a gun.

Listen. It's All Good.
LTD02 is online now  
post #14 of 21 Old 07-08-2013, 09:09 AM
AVS Special Member
 
lilmike's Avatar
 
Join Date: Dec 2009
Location: Pacific Northwest
Posts: 2,193
Mentioned: 0 Post(s)
Tagged: 1 Thread(s)
Quoted: 18 Post(s)
Liked: 87
Quote:
Originally Posted by LTD02 View Post

i've been trying to figure out why you use
par to model a straight horn.

i just worked through the math and you are
right. even though par creates a parabola,
which is a curve, the cross-sectional area
is a linear function of the horn length.

son of a gun.

I did the math a LONG time ago, and I was very pleased whern HornResp added the PAR flare.
I use PAR for most of my models, because most of the time, what I build expands in one plane. When the taper expands in two planes it is a conic section, so modeling with CON is appropriate.
The differences are subtle, but there are differences.
lilmike is offline  
post #15 of 21 Old 07-08-2013, 10:03 AM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
lilmike - I was away for the weekend, and am just looking back here this morning. Thanks for all your comments and suggestions. I'll dig into this a bit later tonight at home. Thanks again! smile.gif
batdorf9 is offline  
post #16 of 21 Old 07-11-2013, 02:41 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
OK, I've done a bit more homework. I can't get lilmike's hornresp config to package exactly, but this is reasonably close I think.

Some background... Originally I was hoping I could do a simple 2.5' x 8' "coffin" with a single divider down the middle. However with 2 "legs" this gives a path length a tad less than 14 feet, and this can't get me down to 25 Hz, much less 20. Maybe 30Hz. So this brings about the 3rd leg/path. But with 3 legs folded up in there, you can't have a nice constant or expanding angle. So I'm starting with a tiny chunk of 1.5deg expansion in the throat, then a big 5deg expansion, then a smaller 3.5 deg expansion. I realize it'd be better to go 1.5 to 3.5 to 5, but that doesn't package in here. If someone could show how to do that in this "coffin" shaped box, that'd make my day.

I'm seeing when using the "loudspeaker wizard" that shortening up L45 could help reduce the dip between 25 and 55 Hz, however I've already got L45 as short as I can make it here. One benefit I'm seeing to this is that it will be much easier to access/install the driver.

I've done some tweaks to get acoustic impedance spikes down a bit, though this has been more guess-and-check than anything else. Shortening L12 seemed to help the most -- I'm at the limit for shortening that now too. I have no intuitive feel for what acoustic impedance is or how to change it. Can anyone enlighten me here?

Final call for some additional suggestions here... is this thing ready for sawdust? Anything to tweak further to improve this config?










HornResp Input: TH8_6_RSS460-HO.txt 0k .txt file


ID=29.00
Ang=2.0 x Pi
Eg=2.00
Rg=0.00
Fta=2.10
S1=376.26
S2=402.97
Par=21.91
F12=0.00
S2=402.97
S3=1525.93
Par=305.54
F23=0.00
S3=1525.93
S4=2003.22
Par=164.41
F34=0.00
S4=2003.22
S5=2206.45
Par=33.35
F45=0.00
Sd=1079.30
Bl=20.54
Cms=1.82E-04
Rms=15.93
Mmd=414.30
Le=1.38
Re=3.41
TH=1
Vrc=0.00
Lrc=0.00
Ap1=0.00
Lpt=0.00
Vtc=8908.00
Atc=1338.00
Pmax=800
Xmax=20.0
Comment=TH8_6 RSS460-HO Tapped Horn
Attached Files
File Type: txt TH8_6_RSS460-HO.txt (438 Bytes, 7 views)
batdorf9 is offline  
post #17 of 21 Old 07-11-2013, 05:14 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Replying to my own thread again... Now have something much simpler to build. Very slight loss of low frequency extension, but improvements in SPL for most of the range. And much easier to build. I'm not wild about the driver firing vertically, though I'm not sure that's a show stopper. Thoughts on this config?










TH10_3_RSS460-HO.txt 0k .txt file
Attached Files
File Type: txt TH10_3_RSS460-HO.txt (439 Bytes, 8 views)
batdorf9 is offline  
post #18 of 21 Old 07-11-2013, 05:56 PM
AVS Special Member
 
lilmike's Avatar
 
Join Date: Dec 2009
Location: Pacific Northwest
Posts: 2,193
Mentioned: 0 Post(s)
Tagged: 1 Thread(s)
Quoted: 18 Post(s)
Liked: 87
Simple is good. Seems to be a reasonable layout too. Areas and expansion looks OK to me at a glance.

Be sure to brace it well.
lilmike is offline  
post #19 of 21 Old 08-01-2013, 02:33 PM - Thread Starter
Newbie
 
batdorf9's Avatar
 
Join Date: Jan 2006
Location: San Diego, CA
Posts: 10
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Progress report on the build here. 3 sheets of 4x8 baltic birch per box. Cut list was very, very simple as planned. 3 hours with a table saw got all the main pieces cut for 2 boxes. Bracing was done a different day. This was my first time building anything with the Kreg pocket screws, and that's gone really well. I'm finding the "coffin" size of the thing to be a bit of an advantage, as I'm able to crawl inside the speaker to install/tighten pocket screws, install bracing, and apply a caulk seal at the corners (though it appears the wood glue and pocket screws closed everything up well). This is making things look very clean on the outside of the box. With any luck, one of these will be finished up by the end of the weekend.

Making sure I could fit inside to tighten pocket screws and install bracing once the top was closed out.


Almost done.


Moments before closing out the top.


CAD showing internal bracing intent.
batdorf9 is offline  
post #20 of 21 Old 08-01-2013, 09:18 PM
Member
 
boostick4's Avatar
 
Join Date: Sep 2006
Posts: 43
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 10
Awesome! Cant wait to see the response.
boostick4 is offline  
post #21 of 21 Old 02-20-2014, 12:32 AM
Senior Member
 
ozziedog's Avatar
 
Join Date: Oct 2011
Location: Gravette, Ar.
Posts: 394
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 22
Did this ever get finished?
ozziedog is online now  
Reply DIY Speakers and Subs

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off