If I have the room and mains that go down to the low 30's, does it make sense to go with a 32ft folded tapered pipe with a 15" tapped in the middle. I can go with RDO and also have the ability to go isobarik, for tuning purposes. A 32ft pipe with an So of .5xSd and Sm 2xSd models decently as a normal TQWP with a centered driver with 85db at 2.83V. MJK's TSWP worksheet showed a nice even response from 20-65hz, which drooped between 20hz and 10hz. Hopefully the tapped alignment will flatten that out. I'm looking for 110db+ potential at 10hz in a pretty well sealed 2,200cuft room. This sub is for HT only duty. Does it make enough sense on paper to go for the big one right off the bat? It will be 20"x 96" x 36" and will require almost 4 sheets of plywood, so it's not really a quick and dirty test box, and I'd like to have some level of confidence going into it.
Well, if you're asking my opinion, I would say you are on a real slippery slope here.
You are tuning below the "acceptable" lower limit which is 18.8hz * .707 or 13.3hz.
Pushing your driver down to 10hz could be hazardous.
Have you tried simming it at 13.3hz?
I would certainly do a comparison.
You are in uncharted territory here, so there is no telling what the outcome will be.
As you tune lower and lower, the demands on your driver will increase geometrically.
If it were me and I had scads of time and money (I have neither), I would GO FOR IT (at 13.3hz, however).
If it works out it will be COOL 😎
You are tuning below the "acceptable" lower limit which is 18.8hz * .707 or 13.3hz.
Pushing your driver down to 10hz could be hazardous.
Have you tried simming it at 13.3hz?
I would certainly do a comparison.
You are in uncharted territory here, so there is no telling what the outcome will be.
As you tune lower and lower, the demands on your driver will increase geometrically.
If it were me and I had scads of time and money (I have neither), I would GO FOR IT (at 13.3hz, however).
If it works out it will be COOL 😎
Qi,
Fs is the lower tuning recommended limit for a traditional TL, so why should it apply here since there is no front radiation and my upper limit will be very low? Might it actually be better if I go longer so I'm tuning at half a wave length of 18.8hz ? I seem to recall that the Jensen Transflex used a 40hz driver on a 16ft pipe. Lastly, why apply the .707 ratio, is that much different than dividing by one of the "golden ratio's", because I don't understand how the critically damped Q applies here? The system Q isn't 1 tuning to Fs, is it?
Is there a more sure way to get to 110db at 10hz if I'm willing to work with 800+Liters? Dan Wiggins recommended the Tempests Iso loaded 6th order bandpass with a lower chamber of 340L and a high chamber of 57L and flared 4"dia x 17" long ports in each. I was worried about the vent on the big chamber sounding like a locomotive. Should I just go with that as a sure thing?
Fs is the lower tuning recommended limit for a traditional TL, so why should it apply here since there is no front radiation and my upper limit will be very low? Might it actually be better if I go longer so I'm tuning at half a wave length of 18.8hz ? I seem to recall that the Jensen Transflex used a 40hz driver on a 16ft pipe. Lastly, why apply the .707 ratio, is that much different than dividing by one of the "golden ratio's", because I don't understand how the critically damped Q applies here? The system Q isn't 1 tuning to Fs, is it?
Is there a more sure way to get to 110db at 10hz if I'm willing to work with 800+Liters? Dan Wiggins recommended the Tempests Iso loaded 6th order bandpass with a lower chamber of 340L and a high chamber of 57L and flared 4"dia x 17" long ports in each. I was worried about the vent on the big chamber sounding like a locomotive. Should I just go with that as a sure thing?
Traditional guidlines advise going no lower than 1/2 octave below Fs.
While I do not have the reference handy, I recall it to be
Fs / ( 2 ^ (1/2))
or
18.8 * .707
I believe your question hides your answer...
Here is the same basic design (not ISO loaded), but with 6" x 42" ports -- still tuned to 10hz - 40 hz +/- 3db
The chambers have been increased to account for the larger ports.
Rear chamber 48" x 30.5" x 15" (12.7 cu ft) inner dims
Front chamber 48" x 6.5" x 15" (2.7 cu ft) inner dims
Box size 49.5" x 39.25" x 16.5" outer dims
Here's a see-thru of what I am proposing...
Attachments
I used MJK's worksheets and modelled this stuff as normal TQWP's. It showed there's nothing to be gained going longer than 1/4 of Fs. Bigger volumes on the other hand got more extension. The best for what I need (10-30hz) was about 900L and 2.83V + one Tempest nets a quite flat 83db in that range, but excursion limits me to less than 100watts at 10hz.
It's looking more and more like sealed with a bunch of woofers and as much power as I can muster and flatten the response is the way to go, especially since I already have plenty of drivers and anything prodigous is going to be big anyway.
It's looking more and more like sealed with a bunch of woofers and as much power as I can muster and flatten the response is the way to go, especially since I already have plenty of drivers and anything prodigous is going to be big anyway.
John in Costa Rica
WOW -- that's a really expensive alternative!
Here's Dan Wiggins isobaric 6th order bandpass tuned to 10-40hz +/-3db, 105db (the two chambers still 2 cu ft, 12 cu ft)
It is 17" x 38" x 50".
I replaced the 42" x 6" pvc pipe ports wth rectangular ports (same CSA)
I would go with that...
WOW -- that's a really expensive alternative!
Here's Dan Wiggins isobaric 6th order bandpass tuned to 10-40hz +/-3db, 105db (the two chambers still 2 cu ft, 12 cu ft)
It is 17" x 38" x 50".
I replaced the 42" x 6" pvc pipe ports wth rectangular ports (same CSA)
I would go with that...
Attachments
any new info on the dts-20? has anyone figured out how it is designed? It wouldnt be that hard to take one appart and test the driver would it?
Well go ahead and buy a dts-20, take it apart and test the driver then - can't be that hard can it? 😉
JohninCR:
"It's looking more and more like sealed with a bunch of woofers and as much power as I can muster and flatten the response is the way to go, especially since I already have plenty of drivers and anything prodigous is going to be big anyway."
I've got 4 tempests sitting here, and I can't model a better way than sealed to load them below 15Hz.
I have 2 labhorns eq'd flat 20hz - 80Hz, and would like extension to the single digits. trouble is I tried the tempests below 30hZ (24Db XO) and they still mess with the sound quality of the labhorns. I've decided to build 2 more labs and eq flat to 15Hz. Will be selling the tempests on.
Cheers,
Rob.
JohninCR:
"It's looking more and more like sealed with a bunch of woofers and as much power as I can muster and flatten the response is the way to go, especially since I already have plenty of drivers and anything prodigous is going to be big anyway."
I've got 4 tempests sitting here, and I can't model a better way than sealed to load them below 15Hz.
I have 2 labhorns eq'd flat 20hz - 80Hz, and would like extension to the single digits. trouble is I tried the tempests below 30hZ (24Db XO) and they still mess with the sound quality of the labhorns. I've decided to build 2 more labs and eq flat to 15Hz. Will be selling the tempests on.
Cheers,
Rob.
RobWells said:
Hi guys! Long time no talk (I've been working (groan))
Current thinking on this forum is that below 50hz the ear is pretty darn forgiving so...
a BIGMUTHA TLINE
a 6th order bandpass (Dan Wiggins)
an EQ'd IB with four drivers in a manifold
a sealed array with scads of power
...tuned 10hz - 40hz should "do the trick" (pick any one of em)
Rob -- you're breaking new ground here.
I guess YOUR ear is not that forgiving?
Hi qi,
A slight correction - the tempests were crossed at 25Hz. Also I was using 3 tempests (I have 4 but never built the 4th box)
I think it's more that even though I used a 24dB/oct there were still higher frequencies (30 - 60Hz) from the tempests mixing with the labs. I was thinking about using a 48 dB slope on the tempests and cutting them in at 20Hz.
I thought that as the labs were sealed boxes below their horns cutoff then why not give them as much vd as the tempests and lose the crossover problem.
Tom Danley thinks it's feasible.
Cheers,
Rob.
A slight correction - the tempests were crossed at 25Hz. Also I was using 3 tempests (I have 4 but never built the 4th box)
I think it's more that even though I used a 24dB/oct there were still higher frequencies (30 - 60Hz) from the tempests mixing with the labs. I was thinking about using a 48 dB slope on the tempests and cutting them in at 20Hz.
I thought that as the labs were sealed boxes below their horns cutoff then why not give them as much vd as the tempests and lose the crossover problem.
Tom Danley thinks it's feasible.
Cheers,
Rob.
Hi Rob
Good luck with it.
That is going to be some system.
I wish I could come over and audition and tip a few with you.
Good luck with it.
That is going to be some system.
I wish I could come over and audition and tip a few with you.
QI, Did you see that the mystery of the "tapped horn" has been solved? Someone posted a cross section of it on here a few weeks ago. That may have been you for all I know! Do you know what the driver is? I believe that I know what Tom is using, but I don't want to post it on here, since the design is still unpatented. I don't know Tom or anything, but I have a hunch what it is, based on some information that I found.
It sounds like both of us have spent a lot of time modeling front and back loaded horns in Mathcad.
Take a look at some of my posts if you're interested in what I've come up with. I'm publishing a design on the carsound forums that's something like the tapped horn. Over there I use the moniker "majestik6".
It sounds like both of us have spent a lot of time modeling front and back loaded horns in Mathcad.
Take a look at some of my posts if you're interested in what I've come up with. I'm publishing a design on the carsound forums that's something like the tapped horn. Over there I use the moniker "majestik6".
Hey John...
Love your posts and that PDF!
My guess is that it is a beefed labhorn driver.
I'll bet you could model a BIB with two labhorn drivers mounted isobarik and get similar performance results to the TOP.
Have you tried it?
Tom's design apparently smooths out the mid-dip in the "camel hump" response curve, however.
If anyone can "push the envelope" he is THE ONE ...
Love your posts and that PDF!
My guess is that it is a beefed labhorn driver.
I'll bet you could model a BIB with two labhorn drivers mounted isobarik and get similar performance results to the TOP.
Have you tried it?
Tom's design apparently smooths out the mid-dip in the "camel hump" response curve, however.
If anyone can "push the envelope" he is THE ONE ...
qi said:
Look at the response for the DTS-20; it's 6db LESS efficient than a lab sub. I don't think that the DTS-20 is using anything similar to the labsub driver.
Based on what I've read, the best driver for a 'tapped horn' would be quite different than what you would use in a labhorn. IMHO, the best driver for a tapped horn would have a lower FS and a lower qts, possibly a lower VAS. Basically the labhorn driver has higher efficiency, at the expense of low bass extension.
So to make a long story short, if you're trying to make something like the DTS-20, try modeling it with a woofer that matches those characteristics.
If you do a little looking online, there's someone else that is using a tapped horn in a DIY format, and I think they have a VERY good match for the cabinet. And they're using a vendor I'd never considered.
Details? Links? Please?
In my (very) different version of this sub I got help modelling various drivers in a traditional style tl watching the low frequency response in particular. The lowest frequencies should be about the same in either traditional style and this box. I don't think the driver selection is as critical as you might think. There are many 12 and even 15 inch drivers that would probably work very well.
With respect to the Lab12, I think there is a very good chance that this is the driver in the DTS-20, slightly modified or not. First off, Mr Danley probably had lots of them lying around after his Labhorn projects. Second, take a look at the earlier PB12 frequency response chart and compare with the new DTS-20 chart. I'm betting they both used the same driver. I don't have the appropriate software, but I'm also betting that if you model the PB12 with a Lab12 and a 2:1 expansion ratio (in a traditional tl of course) and then model the DTS-20 assuming the same driver but with a straight (not tapered) line and compare these graphs to Danleys subs, the low end response will be very similar. The tapped horn magic should "fix" the top end response in your traditionally modelled simulations.
I will soon be changing mine to be much more similar to Danley's, but I have a cheap 15 inch driver in mine and I fully expect results similar to the PB12.
In my (very) different version of this sub I got help modelling various drivers in a traditional style tl watching the low frequency response in particular. The lowest frequencies should be about the same in either traditional style and this box. I don't think the driver selection is as critical as you might think. There are many 12 and even 15 inch drivers that would probably work very well.
With respect to the Lab12, I think there is a very good chance that this is the driver in the DTS-20, slightly modified or not. First off, Mr Danley probably had lots of them lying around after his Labhorn projects. Second, take a look at the earlier PB12 frequency response chart and compare with the new DTS-20 chart. I'm betting they both used the same driver. I don't have the appropriate software, but I'm also betting that if you model the PB12 with a Lab12 and a 2:1 expansion ratio (in a traditional tl of course) and then model the DTS-20 assuming the same driver but with a straight (not tapered) line and compare these graphs to Danleys subs, the low end response will be very similar. The tapped horn magic should "fix" the top end response in your traditionally modelled simulations.
I will soon be changing mine to be much more similar to Danley's, but I have a cheap 15 inch driver in mine and I fully expect results similar to the PB12.
Not what I have seen in all the threads from TD. What I have gleaned is that the classic camel hump "middle" (only) is lifted -- giving a much flatter response.
While this means you get much more SPL over a broader range, there is absolutely NO benefit to the bottom or top end -- only the middle.
There is a really good explanation of how flattening the response curve yields significant improvement in dynamic range on the Whyse Profunder web-site (which uses VERY similar technology to TD)
Actually, the TAPPED HORN technology "robs" bottem-end performace for a flatter response curve.
Qi, that's what I was getting at. The low end should operate in a similar fashion whether the driver is inside the box, or at the same place on the outside (traditional). So a driver that models well in the same box with traditional tl driver loading should work in a tapped box.
The top end response I was referring to is everything above the first harmonic up to about 100 hz, which is smoothed out nicely on the tapped horn frequency response charts.
The top end response I was referring to is everything above the first harmonic up to about 100 hz, which is smoothed out nicely on the tapped horn frequency response charts.
That is exactly what I proposed to JohninCR months ago.
John, are you here?
What dimensions did you finally settle upon -- especially the port?
I know your MJK sims were getting you down into the low teens...
John, are you here?
What dimensions did you finally settle upon -- especially the port?
I know your MJK sims were getting you down into the low teens...
I'm still up in the air about what to build and keep getting back to going sealed and using a number of sub drivers flattened by EQ and Linkwitz transform. I was looking at a big box anyway and spreading them in a line across my front wall, then I get the advantages of line array behavior, reducing reflections and increased efficiency.
Plus there's a new monkey wrench in the mix. RobertG the designer at Gemme Audio is claiming over at the fullrangedriver forum that with a 1.5ft3 box using a 4" driver (Fostex FE108EZ) to have measured anechoic results of 90db/1w/1m at 25hz with reasonable distortion. It's some kind of multichambered horn, TL, BR. If that can be done with a 4", then imagine what a subwoofer driver could do in a similar, but larger, alignment.
Plus there's a new monkey wrench in the mix. RobertG the designer at Gemme Audio is claiming over at the fullrangedriver forum that with a 1.5ft3 box using a 4" driver (Fostex FE108EZ) to have measured anechoic results of 90db/1w/1m at 25hz with reasonable distortion. It's some kind of multichambered horn, TL, BR. If that can be done with a 4", then imagine what a subwoofer driver could do in a similar, but larger, alignment.
- Status
- Not open for further replies.