AVS Forum banner

41 - 60 of 379 Posts

·
Registered
Joined
·
3,343 Posts
Try something more involved.



This is the ELS212 tapped horn.
 

·
Registered
Joined
·
1,771 Posts
Discussion Starter #42
Quote:
Originally Posted by soho54
Try something more involved.
Yikes... that one may be over my head a bit



This evening's effort may have produced the final folding diagram. I'm not unhappy with this. 78" x 43" x 13.5." Corner appears to be right at 15Hz. Horn turned out a little long again, but not nearly as bad. Attached is the Hornresp sim for the following fold diagram. Not sure it's entirely accurate, but it should be close.




300 watts in, excursion at about 21mm, half space:




Edit - hold the phone, I am unhappy with the phase response. I might tweak that baffle for driver one yet and forget about mounting it magnet down.

 

whsdxr3c.txt 0.408203125k . file
 

Attachments

·
Registered
Joined
·
1,771 Posts
Discussion Starter #43
The OCD in me just couldn't leave things alone. After thinking about it for a while, I just had to not only fix up the phase mess on the last revision, but eliminate all path length errors as well.


As a result - this is it. This is what's getting built. Revision 4 of the folding diagram shows a path length error of a whopping three centimeters. That's just over an inch that this one is short by. That does not need to be ironed out unless your name is Adrian Monk.


The final box will be 75" x 43" x 13.5." Length and width will be an inch longer than the TB horn, while the depth will be 4" less. This should pretty well fit the space I have ready for it. Dimensions will be smaller with 15mm wood, but I don't plan on using it. If desired, I'll cook up another fold diagram with 15mm.


Once I get closer to building day, I'll start up a build thread and share the fold diagram with dimensions added for anyone wishing to try this one too. Meantime, I have the final Hornresp simulation file attached - this is more or less what the box as drawn will actually do. I think I got pretty close to my original plan.




SPL, half space, 300W:


 

whsdxr3d.txt 0.4072265625k . file
 

Attachments

·
Registered
Joined
·
526 Posts
Looks good, OW! That's pretty darn thin for a sub; should be pretty easy if someone wanted to hide that behind a couch....or perfect for a riser. I'm jealous of your potential output!


Sigh....those dimensions are really close to my subwoofer 'cubbyhole'...I've got these annoying built-ins that restrict the width of my potential subs to 30", and go out 14". I need to learn to fold just so I can try and squeeze other people's designs in. I'm also getting closer and closer to just taking a drywall knife to the cieling in the corner and throwing an IB in the attic.
 

·
Registered
Joined
·
1,771 Posts
Discussion Starter #45
Thanks - I can't wait to build it



I checked my first folding idea, being the taller and narrower one - it still comes in at 7" too wide. Still, there's nothing wrong with a good IB - that's what I'd be doing if I weren't living in a basement.
 

·
Registered
Joined
·
69 Posts

Quote:
Originally Posted by Oklahoma Wolf /forum/post/19198051


I'm confused. Does Hornresp allow you calculate the result of two drivers at different locations? I thought 'TH - 2S' meant the drivers where side by side but electrically in series.
 

·
Registered
Joined
·
2,391 Posts
The drivers are well within a 1/4 wavelength at the frequencies we're working with. At 20 Hz, a single wavelength is on the order of 17 meters, so a quarter of that is 4.25 meters.


As designed, these drivers will nearly be touching. They'll effectively be a point source until the distance between their acoustic centers is over 1/4 of a wavelength. For easy math - let's say the drivers are 30 cm center to center. 30 cm X 4 = 120 cm or 1.2 M, which is the wavelength of 286 Hz, well outside our band of interest.


So - in a nutshell, it won't matter at all so long as the drivers receive equal power and are driven in phase.
 

·
Registered
Joined
·
69 Posts

Quote:
Originally Posted by lilmike /forum/post/19265933


The drivers are well within a 1/4 wavelength at the frequencies we're working with. At 20 Hz, a single wavelength is on the order of 17 meters, so a quarter of that is 4.25 meters.


As designed, these drivers will nearly be touching. They'll effectively be a point source until the distance between their acoustic centers is over 1/4 of a wavelength. For easy math - let's say the drivers are 30 cm center to center. 30 cm X 4 = 120 cm or 1.2 M, which is the wavelength of 286 Hz, well outside our band of interest.


So - in a nutshell, it won't matter at all so long as the drivers receive equal power and are driven in phase.

As a test, I recalculated the parameters by adding 30 cm (the center offset of two drivers close together) to L12 and L34, and subracting 60 from L34. Sure enough, the results were almost identical.


However, when I run the sim with only one driver (TH 1), the results are very different (not as smooth). Why is that?
 

·
Registered
Joined
·
2,391 Posts

Quote:
Originally Posted by tg3 /forum/post/19266225


As a test, I recalculated the parameters by adding 30 cm (the center offset of two drivers close together) to L12 and L34, and subracting 60 from L34. Sure enough, the results were almost identical.


However, when I run the sim with only one driver (TH 1), the results are very different (not as smooth). Why is that?

The driver needs to be loaded by the horn - you've got one driver in a horn sized for two, so it is severely underdamped.
 

·
Registered
Joined
·
69 Posts

Quote:
Originally Posted by lilmike /forum/post/19266316


The driver needs to be loaded by the horn - you've got one driver in a horn sized for two, so it is severely underdamped.

Got it. And how can you tell when a driver is properly damped (aside from the FR)? Surely the Sd/S2 compression ratio is important.


Many thanks!
 

·
Registered
Joined
·
1,771 Posts
Discussion Starter #52
I just go by the FR. The compression ratio is important, though... that's why I went with the SDX woofers instead of the TC Epic 10. They both model equally well, but the TC wants a higher (but not unreasonably so) compression ratio.


Well, that and the fact that I can get the SDX's easier and cheaper.
 

·
Registered
Joined
·
2,391 Posts

Quote:
Originally Posted by tg3 /forum/post/0



Got it. And how can you tell when a driver is properly damped (aside from the FR)? Surely the Sd/S2 compression ratio is important.


Many thanks!

In addition to compression ratio and frequency response, I look at the impulse response. The cleaner the impulse, the better the match between the horn and the woofer, all else being equal, and grossly oversimplified.
 

·
Registered
Joined
·
21,339 Posts
ha, i was experimenting with a mockup for penn for a horn and arrived at the exact same fold. different dimensions, but same design.


this design definitely has simplicity going for it.


lilmike i wonder how important the impulse response is as a practical matter. i've listened to the impulse response of some really good pro audio drivers that sound almost like a chirp, while others such as the tht sound more like a "whoomp" with lots of after resonances. in practice though, the tht is generally reported to be a very clean sounding device.


oklahoma, it is looking good from over here.


when you say, "I just go by the FR." what specifically are you looking for/at?
 

·
Registered
Joined
·
2,391 Posts
I'm posting mobile today, a bit of a lull in the work going on onsite.


I said grossly oversimplified
.


There is a lot more than I fully understand, but a good design can eliminate a lot of the resonances, leading to a much more controlled impulse with minimal ringing.


This leads to better reproduction of the input signal, because less garbage is introduced as a result of the cabinet. To be fair, consider the band of interest, which removes a lot of the hash. Akabak can do this, soho54 showed us this a while back. Hornresp kind of can, with the spectrum option.


Anyhow, having made cabinets with bad to worse impulse responses, as well as good ones, I can say most assuredly that I prefer the sound of the cabinets with cleaner impulse responses.
 

·
Registered
Joined
·
3,343 Posts

Quote:
Originally Posted by LTD02 /forum/post/19269821


i wonder how important the impulse response is as a practical matter. i've listened to the impulse response of some really good pro audio drivers that sound almost like a chirp, while others such as the tht sound more like a "whoomp" with lots of after resonances. in practice though, the tht is generally reported to be a very clean sounding device.

The IR is hard for most to look at, and glean anything from. HR's more so as it doesn't band limit it. The trick is to remember the large wavelengths involved with sub frequencies, and the large time gaps they would have. This is why the Spectrogram option is so nice. It allows a better presentation for people not hip to squiggly lines. Funkiness far above the passband just doesn't really matter.


The lower and flatter the FR the more whoomp that is present. The upper rolloff of the sub/driver also shapes this. If the sub is on it's way out in the 100s there isn't going to be a lot of chirp there compared to a sub that has a driver which could meet a tweeter.
 

·
Registered
Joined
·
1,771 Posts
Discussion Starter #57

Quote:
Originally Posted by LTD02 /forum/post/19269821


when you say, "I just go by the FR." what specifically are you looking for/at?

Same thing I look for in the ladies - good looking curves



Really, what I like to see is a mostly even slope inside the intended bandwidth with maybe a small hump in the middle. That first harmonic spike (about 62Hz in my latest SPL screenshot) I usually mentally suppress a little bit to coincide a little better with reality. I also like to see a smooth set of ever decreasing dips and peaks above the bandwidth (I call it the boing-boing-boing effect - I know, I'm a loon)... this is how I know when I'm starting to get close to something that will look decent in the impulse response and phase screens.


I would say about 85% of my Hornresp modeling is done first making the FR graph look good in full, half, and quarter space; 10% checking the schematic for size; and then 5% checking the other factors to see if I got something halfway decent.
 

·
Registered
Joined
·
1,074 Posts

Quote:
Originally Posted by Oklahoma Wolf /forum/post/19270068


I call it the boing-boing-boing effect

You guys are going to have to cut down on the high tech terminology, you are losing us amateurs with all this techno-speak.
 

·
Registered
Joined
·
1,771 Posts
Discussion Starter #60

Quote:
Originally Posted by BeerParty /forum/post/19270473


You guys are going to have to cut down on the high tech terminology, you are losing us amateurs with all this techno-speak.

Go ahead - ask me how I came up with "vomit factor," or how capable a tapped horn is of making the listener lose his lunch after some bad pizza



My Tang Band horn gets a rating of 4/10 on that scale. I'm hoping for a solid 6 or 7 from this one... not that I plan on eating bad pizza again.
 
41 - 60 of 379 Posts
Top