Driver for a DTS-20

[JayDee]

Hello my name is Jack and im new to these forums.

I have been wanting to build a TL sub for a long time, but havent gotten around to it because of work.

But ever since i heard of the DTS-20 over at avsforum.com, i've been wanting one of those bad boys badly!

So after alot of snooping around in varous forums (here, prosoundweb and others) i've gathered alot of info on Tom Danley's "tapped" design, although most of it is speculative since the technology is so new and practically imposible to model (in affordable programs anyway).

I have made a litle spreadsheet with the DTS-20 driver spec's and two possible alternatives. (All in all I think the Infinity Kappa Perfect 12d VQ would be the ideal choice because of the variable Q feature).

Regarding the cabinet construction itself, there will have to be a bit of tweaking to the new driver since you can't model the "tapped" alignment yet.
I will post my bid on the precise cabinet dimensions modelled to the Infinity driver, when i get a program i can draw in (other than paint!)

The main design ideas for the cabinet can be seen in the next post's attached picture
(The picture is from someone elses post here but i cant find the post so i hope it is ok with him that i just post the image)

[JayDee]

Here is the cabinet design idea since you are only allowed to upload one file at the time:

(and it takes hours to get your first post approved)

And just to be clear, since i dident actually post a question:

In the "tapped-horn whitepaper" Tom Danley says that the driver used in the DTS-20 is nearly 200 grams in order to impedance match the driver to the air resistance in the TH design.

Using Thielesmall.com and various other sites, i have looked at T/S parameters of over 100 12" drivers from varous manufactures with the:
Mms = "...Almost 200 grams..." as starting point and to me the Infinity seems to be practically the only possible driver?

(The LAB-12 weighs in at only Mms 146 grams but has a stronger moter)

[pinkmouse]

From the little research I've done, I suspect the driver is a modded Lab12, they may well have just added mass to the cone to make it suitable for the new system. It's certainly something you could look at, you could test with modelling clay, then replace with coins superglued to the cone.

I too, would love to have a go at this sort of design, if only more info was available, so I look forward to your results.

[GM]

Originally posted by pinkmouse
From the little research I've done, I suspect the driver is a modded Lab12, they may well have just added mass to the cone to make it suitable for the new system.

Greets!

Yeah, I'm thinking he doped a LAB 12 with the same compound (or very similar) to what's used on the Contrabass and BT-7 drivers since it also makes the diaphragm considerably more rigid/damped. Anyway, the extra mass only flattens its response and lowers its HF corner a bit, so having less for a HIFI app shouldn't be a big deal if there's any room gain.

GM

[Circlomanen]

How about using two drivers in a pushpull or also called, isobaric arangement. That way you get twise the motorstrengt and twise the cone mass. The cancelation of even order harmonics is a good bonus too.

This is something I tried but in a very small scale. 2 4½ inch car speakers in a 240 cm long folded pipe with an slowly expanding crosssection area. 78 cm2 to 200 cm2

[just a guy]

I made something similar to this very recently (very close external measurements), but mine has a 2:1 decreasing taper ratio instead of a mass loaded expanding taper. With the 14 foot pipe I have strong response down to about 16 hz, but without the strong bump at the tuning frequency as indicated in the DTS-20 FR charts.

I chose my driver by a very different method. I asked for help simulating different drivers in my concept tl with MJK's tl spreadsheets. I chose the Dayton Quatro 15 because it was cheap and the Jensen Transflex used a 15. It showed good response as low as I wanted to go.

More recently I have been thinking about a smaller driver with lower q to really bring out the most of the bottom end (6db or so peak at tuning like the DTS-20). The Lab12 popped to mind but I had no idea that other people thought it might work too. I was also thinking about a Peerless 12.

I am very interested to find out how your project progresses.

[johninCR]

TD mentioned at some point that the driver output on the side closer to the terminus has little impact on response. If this is true then we do have a way to model this tapped horn. My approach has been to use MJK's mathcap spreadsheet for a TQWT, but focus only on the predicted port output.

My goal is extension not extreme output, so I've moddled my plan using a 16ft line and a larger CSA using 2 15's (push/pull but not isobarik). The extra CSA and Sd shows very flat port output all the way down to 10hz.

Also, if it doesn't work I can always mount the drivers outside of the line and EQ down the response above 20hz.

[JayDee]

Originally posted by just a guy:
More recently I have been thinking about a smaller driver with lower q to really bring out the most of the bottom end (6db or so peak at tuning like the DTS-20). The Lab12 popped to mind but I had no idea that other people thought it might work too. I was also thinking about a Peerless 12.

I've also checked out the peerless XXLS drivers but ended up with the Infinity bacause of the low Q's, high Mms and Xmax of 3.35 cm peak-peak wich gives it a VD of 822,4cc which translates into ALOT of headroom : ) Not to mention the variable Q feature and the dual 4 Ohm voice-coil for 2 Ohm or 8 Ohm operation.

The only "bad" thing i read about it was on caraudiomag.com where it did'nt score to high in a subjective listening test (80%)
- I don't take subjective tests in magazines too seriously anyways.

Originally posted by Circlomanen:
How about using two drivers in a pushpull or also called, isobaric arangement. That way you get twise the motorstrengt and twise the cone mass. The cancelation of even order harmonics is a good bonus too.

Actually that was my first idea. I thought about two 9" Scanspeak "revalator" 23W/4557T02 in push pull, mounted in a folded QWTH. But their low Qms/Qts and the fact that they cost alot more than the 12" Infinity driver made me look at the DTS-20 design, with extremely low distortion in the first place.

Anyway i think you loose alot of the benefit of push-pull (even-order distortion reduction) in open cabinets since the drivers does'nt "see" each other that much compared to sealed cabinets.
-I have'nt tried it, so i would'nt know for sure.

To GM and pinkmouse:

Even if i modded a LAB-12 to make it heavier (>Mms <Fs <Qes <Qts) dosent the Infinity driver seem to surpass it anyway?
-Or is it just me? ; >

[pinkmouse]

Originally posted by JayDee
To GM and pinkmouse:

Even if i modded a LAB-12 to make it heavier (>Mms <Fs <Qes <Qts) dosent the Infinity driver seem to surpass it anyway?

Without knowing more about the maths used to derive the enclosure, it's pretty much guesswork as to which would work better, so I'd just go for the cheapest for now, and see how you get along.

[just a guy]

I think John is right. From my experiences (unfortunately without an SPL meter) it looks like you can predict to an extent what will happen by modelling the driver in a tl enclosure program.

In my case the sensitivity seems close to the model, the harmonic peaks are in exactly the same places and low frequency extension is very close to the chart. I wouldn't try a new driver in my box without modelling it, which unfortunately I can't yet do myself. I was surprised when I got the Dayton Series II and the Dayton Quatro simmed. Both driver's specs are fairly typical, but the Quatro modelled about 6 db louder with the same power in the same tl.

Also, if you model it, I think you will find that the drawing you selected, if you stay with the DTS external measurements, will tune the box closer to 30 hz than the DTS's 20 hz. But very, very sensitive, very loud.