CSS TRIO12 Front Loaded subwoofer 16hz F3

Hello Gents here are a few pics of a true horn loaded sub that I put together over the last few days. I took it to the DIY Ottawa show over last weekend. All were impressed.

The sub was measured using a Beringher EQ at 12feet away. The room was 16 by 30 feet. The bumps are 3 db. So this sub is essentially flat in room. The Fs is at 18 hz and the F3 is 16 hz.

It is totally non compromised. It is an exact representation of the simulated box from Hornresp. Took quite a bit of work to fold it optimally. But the proof was in some very nice pudding.

So to scare every body away I will post the size!

16.5" x 32" x 92" THe mouth can and will be put on the long side to allow greater placement opportunity.

And yes that is 7/16" OSB. Used intelligently you can use almost any sheeting if you know how to think about proper bracing. I spent more money on PL Premium than I did on sheeting!

Mark

No Problem!

See the attachment.

Compresion ratio is 4.73 to 1

path length is approximately 23'

Mark

Pics available at:

http://s407.photobucket.com/albums/p...n%20Subwoofer/

A bit oh comparison.

Hornresp gives a 2Pi output right around 94 to 96 db/watt.

Measured in my shop in a bad location that is 3 feet off of the floor and 7 feet and 4 feet respectively from any boundary I get 94db/watt. I call that location about 1 and 1/2 Pi. So works as per design. What I will actually get to measure when it finally stops raining long enough will be ground plane measurements. That will get me closer to maximum SPL.

I can max SPL inthe shop to. But it will be a dirty measurement due to the great many things a bouncing at those levels. Might be better to do a ground plane and bump up the numbers to show inside measurements. A little safer and a lot cleaner sound.

Hornresp gives peaking at 200 watts and 125 db. With decent boundary reinforcement you could get 130 db with this power input. THis would be right where X-max is approached. So this is really a killer sub.

If the results are as accurate in the simfor Max SPL as they are for 1 watt this should be very interesting to measure!

I do agree that on face value 130 db is a very high number.

But take this into account. The measurements are done outside on the ground plane. D.B. Keele among others has shown how to use such a measurement,and some math to get you SPL numbers that correspond to interior sound pressure levels. So outside you will not hit 130 db at the distance of 10 meters. But that is the optimal distance to measure a horn. You can for instance measure at 10 meters using 100 watts and corelate the SPL level to that of 1 meter and 1 watt. Add in 12 db for boundary gain and you can have a quite accurate internal SPL level.

Now throw this into the mix. I've been doing loudspeaker and driver design work both professionally and diy since 1989. I have decent test equipment ( Beringher ECM 8000 ) and an old calibrated Ratshack SPL meter. I know what each of these tools is good for and not good for.

I built this test bed to see how far I could push a design I have been modeling for the past six months. It worked perfectly the first time. The only setback was some additional bracing required due to my own impatience!

Mark

Well rain every day this week!

Kind of hard to do electrical measurements on wet grass. So the forecast says clearing towards the end of the week. I'm holding my breathe!

Doing SPL tests on this box inside is not my idea of fun. So wait until the weather co-operates.

Mark

Thanks BIGmouthinDC

It is so gratifying when a first physical prototype works as modeled.

Mark

Still working this weekend. Sucks I know but thems the breaks.

But good news is that it is raining and I couldn't make the tests any way. I will be back to more normal hours starting this Monday. Gives me some free time to do some proper measurements.

Any requests? Every body wants SPL over frequency. That's a given. Most want max SPL. THat will be hard with this horn as I'm not doing that inside! To many things to break. I'll do it outside and show the tests and we can strap on the required db's for each boundary we would see indoors.

I have listened extensively indoors. It really does plumb the depths with gusto. I have quite a few CD's that I have collected over the years that I know well and have measured using a spectrum analyzer. So the parts that should be played loud and proud I know exatly what's down there. And I'm still grinning. 32 foot stops on pipe organ s are a whole new thing with this sub. SImple jaw dropping. And I have been building speakers for over 22 years. The closest I have done was a 5pcs 15 inch monkey coffin driven by an early class D amp from a company I was doing work for back in the mid 90's. But this sub more than competes and can exceed almost any sub with 100 watts.

Mark

Well a while back I actually got to measure this box. In between rain! The tractor did the lifting and we placed the box under the patio out back. THe peak in the response is an earlt reflection due to mike location and a big earthen burm close by.

But all in all other than the wierd impedance response that I finally figured out to be a leak in the seal around the driver access the box measures as specified in Hornresp.

When I finally get a chance where there is no one in the house I will do some in room SPL tests and distortion. As this weekend is quite nice I may try a true ground plane measurement. The one done as shown is not a true ground plane and there is significant error above 60hz. Can't beat the reflections but below that where we are actually most interested it is indeed a rather flat subwoofer. I repost the measured in room RTA of the same box. We did this at the Ottawa DIY show. It reflects the outside measurements quite well. This is a real killer box!

Big yes. But it performs. As simulated it should peak out just above 120db in room with 200 watts. 118db with 100 watts. So this is what you get with some good engineering and lots of box volume. Effortless bass. Not for the faint of heart.

Mark

Rusty! Hey that paint job is one year old!

Won't fit in the NRC anechoic chamber anyway!

Will do some high power ground plane measurements this weekend I hope. THey are calling for decent weather and we get a long weekend up here.

Mark

Thanks for the kind words gents. (gpmbc and noah katz )

Noah

I have been trying to do max SPL tests in room but I am not able to pull this off for the amount of noise generated by doors vibrating and opening and closing even when latched. I can't get any clean measurements. I tried in the garage. The door vibrates so violently I swear it is louder than the horn!

So heck if I know at this point.

As a pretty good answer I'll post this. It is a 2Pi and a 0ne Pi SPL simulation done in Hornresp. Now before you get all up in arms with this note that the box has tracked the simulation very accurately even down to the baseleine SPL. I went through 6 folds to get the horn to match the data from Hornresp. THen I back simulated the as folded version to confirm the efficacy of the fold. Yes I'm nut's. But I have been doing horns for many years and I have come to know two things. Hornresp is accurate. And if you make the horn as simulated it will perform very closely. The things that wash out in the mix are the narrow peaks that Hornresp gives you.

So the graphs are posted to give you an idea of what a proper corner loading will get you. Just about 120 db at or near 16 hz with 200watts and at the edge of X-mech. THe baseline 1 watt graphs are there for confidence purposes. They reflect almost perfectly what I am measuring.

I have to say I have never had to measure something as difficult as this horn. But I have never had to measure something with the SPL capabilities of this horn so I'm on a learning curve.

The thoughts from here on out will be high power testing outside on a ground plane. If the results still jive with Hornresp we can extrapolate the figures much more accurately.

And note I'm not posting .5 Pi space. That is always a bit on the optimistic and unrealistic side of things. From experience I can say that what you simulate for 1 Pi is what you will get from a corner loading in a real room. But at the low end in a basement corner you will get at least another 6 db on the very bottom end and the F3 will drop a bit more. The roll off as measured generally reflects decent room gain. This was the design goal in the first place. The horn in room in a good reflective corner should be esentially flat down to 18hz and minus 6db at 16hz.

Mark

Hi fredfish

Your eyes work well. Me thinks you misread the text. In a corner those missing 2 hz will benefit from room gain and a different loading. of the horn mouth. You will get about 6 db boost near the bottom end. Hence the claimed approximate 120 db at or around 16hz.

I have listened to pipe organ recordings that go down low loud and proud with this sub. It performs well in the netherworld of ultra low bass.

It's actually kind of spooky. Very low distortion and enormous dynamic range. First speaker I have built in a long time that reproduces concert bass drum in a way I remember. I used to play in an orchestra (french horn).

The only thing close was a five 15'' cabinet that could double as a phone booth tuned down to 9hz. I built that back in 97. It needed a bit more power. And it sounded a bit on the muddy side.

I just read third floor walkup. Almost sounds like Montreal?
Could get you out of a lease in a big hurry!

Mark

pics available:

http://s407.photobucket.com/albums/p...n%20Subwoofer/

Mark

I've been building horns professionally since 1997. Yes I have modeled tapped horns in Hornresp. But there have been only two posted by anyone I have watched that I know of that were actually true tapped horns. The vast majority are tapered quarterwavelength resonators.

I did some wierd ones here:

http://www.diyaudio.com/forums/subwo...-horn-car.html

Mark

Yep this is definitely the build in place type of sub. But the performance is just nuts.

Mark

Hi Noah.

I just crunched it.

I pushed the efficiency a bit higher and raised the low end up a bit. The Fs is now 14hz and the F3 is 13 hz

I also post a fairly realistic corner loading as a 1 Pi space graph.

But this box is the size of a small room! 113 cubic feet. Yes you can shrink it some about 25% at the expense of ripple in the response. But this kind of performance is not for the faint of heart. Makes the box in this thread look small!

Mark