multi-driver sub question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I wondered how they got such a high sensitivity. I will use a more sensitive driver since the Kef is all but impossible to find. Too bad they discountinued it. In order to use the same cabinet design I decided to use some software to help me make some better predictions on performance. I downloaded a shareware program called TLBoxModel and played with it awhile to get the gist of its operation. To check how well it would predict response and as a starting point for my modifications, I input the Kef 139s T/S specs in the driver fields and the Perfectionist sub specs (one side) in the enclosure fields. To my amazement, the program predicted the same response as claimed: fairly flat down to 17hz w/ a spike at 68hz. My next step was to change the driver specs from the Kef to the Titan 10. The response took a huge dive but after playing with enclose volumn and tapering ratios I was able to almost match the Kefs response. Actually, the Titan shows flat below 17hz! Whether this is really attainable I dont know but it sure looks impressive. The modifications were a reduction in taper from 3.0 to 2.5 and the S0?(the area of the baffle or the large end of the line) was decreased from 1200cm2 to 880cm2. This all resulted in a footprint reduction from a 1' x 2' to 18" x 13" exterior dimensions using 1" MDF. The length of the line, 3.65m dropped to 3.4ms. I added accousical stuffing to the design parameters although from everthing Ive read, this can only really be done by trial and error, using ones ear.

Have I missed anything? The Dayton is 90db efficient and has decreased the cabinet size due to a lower VAS but why it needs a larger port I dont know- xmax maybe?. The taper is still larger (smaller port) than most of the designs I have seen but I calculates that way. Should I build a test enclosure and buy a driver or do I need to do some further engineering on this.

Thanks Taylor
 
Taylor,
As a general rule, you start the transmission line at something like 1.25 to 1.5 times the area of the cone, then taper to an opening equivalent to the area of the cone. Yes, lots of people do other tapers--this is the traditional way.
The volume, per se, isn't relevant except for determining how much of your living room it will take up. A transmission line, by definition, isn't a resonant enclosure such as a reflex cabinet; you're not trying to give it a resonance of its own to offset the falling output of the driver.

Grey
 
diyAudio Moderator Emeritus
Joined 2001
Taylor:

Should you build a test box? Well, even a test speaker will take time to build when it is this large.

So far, all you know about "The Perfectionist" is that it is a website. And their sensitivity specs are kind of deceptive. For all you know, the reason your Transmission Line program fits the response curve of "The Perfectionist" so well is that "The Perfectionist" people are simply publishing the predicted output from their own Transmission Line computer program as measured results!!

I suspect you need a confidence boost in this large project. Might I suggest a solution? Take a 4" woofer, and design and build a couple of transmission line systems with it for test purposes. There is test equipment available for free on the web to test frequency response, etc. Instead of MDF material, cheap pine shelving will suffice for these little test enclosures. You can slap them together really quickly and easily, and measure the response. If they both check, you know the Transmission Line formulas work in the real world.

For some nice looking inexpensive 4" woofers, try the Versatronics DOB 100 R PP/8
http://www.speakercity.com/specs/versatron_ts.shtml , from Speaker City, http://www.speakercity.com .

For some test equipment on the web, (I have not tried this software), try Steve Ekblad's site:
http://www.wssh.net/~wattsup/audio/ I am sure there are other places to download test equipment as well.

Nothing like designing a speaker, even a small one, and seeing it work as predicted to build confidence in what you are doing.
 
diyAudio Moderator Emeritus
Joined 2001
Taylor:

As long as you have your Transmission Line software mastered, I submit the following woofers to your consideration.

Peerless is a Danish company that frequently gives engineering advantages found on more expensive woofers. I have used Peerless woofers and can vouch for the fact that when they go into the box, you get the predicted response. If anything, Peerless has more of a "high-end pedigree" than the Dayton Titanic, good as that is.

Both of these cast frame woofers cost about the same as the Dayton Titanic, and have the same long excursion: 2 inches peak-to peak. They have lower free air frequencies, and smaller Vas. They also have extremely low Qts.

In a reflex box that low Qts would make them suitable for a smaller box with a higher cutoff, but who knows how they would work in a Transmission Line? Try plugging in the numbers to your software and see what you get. Maybe you can get that footprint still slimmer than it is!


SWR 269 Ten Inch

Thiele Small parameters: Free air
Nominal impedance Zn (ohm) 8
DC resistance Re (ohm) 3.4
Resonance Frequency fs (Hz) 18.9
Mechanical Q factor Qms 2.63
Electrical Q factor Qes 0.18
Total Q factor Qts 0.17
Equivalent volume VAS (ltrs) 89.7
Reference voltage sensitivity Re 2.83V 1m at 115 Hz (Measured) (dB) 88.4




SWR 308 Twelve Inch

Thiele Small parameters: Free air
Nominal impedance Zn (ohm) 8
DC resistance Re (ohm) 3.5
Resonance Frequency fs (Hz) 18.1
Mechanical Q factor Qms 3.70
Electrical Q factor Qes 0.21
Total Q factor Qts 0.20
Equivalent volume VAS (ltrs) 139.2
Reference voltage sensitivity Re 2.83V 1m at 106 Hz (Measured) (dB) 90.6

More specs can be found at : http://www.peerless.dk/
On the left side of the page, click "XLS series", then click on the numbers "830514", (10 inch), and "830515", (12 inch).

Good luck!

[Edited by kelticwizard on 10-11-2001 at 11:23 PM]
 
I'll check out the Peerless drivers. They have a good reputation although I havent any experience will them. So far, I have experimented with several drivers, mostly cheap ones, just to see how the software works. It appears that the traditional taper, 1 > 1.5 produces a smoother upper bass/lower mid response that is probably easier to eliminate with stuffing etc. But in most cases, using a larger taper, like 2 to 3 and a larger SO area(closed end/baffle) the spikes get way worst but the fs drops considerably. What I surmize is that since almost all TLs are full range enclosures and the effects on the upper freq. are important, the design dictates a smaller taper. But if one is only concerned with 100hz and below, the extra LF is chosen over flattened upper bass. I can only assume this is the reason that the Perfectionist sub is designed this way. By the way, I believe I read that it was originally a Fried design. Enclosure volumn changes quite a bit with the changes in taper and SO. When I talk about enclosure volumn, I really am talking about the size of the box. If the SO is 1000cm2 with a 1 taper(none) the folded TL will have a 2000cm2 footprint(baffle and port of the same dimension). But if the taper is increased to 3, it now becomes 1000cm2 + 333.3cm2 = 1334cm2. The footprint is now quite a bit small. I also found that many drivers SD is actually much smaller than their overall diameter. A taper of 1.2 using the SD as the port size doesnt produce an area large enough to mount the driver. Most of the full range cabinets are wider and shallower with a labyrinth within to accomidate the driver.(many use several smaller drivers to make it work)

I think my next step is to purchase several publications on the subject and see if I really have a clue as to what Im doing. Building smaller experimental boxes w/ cheap drivers is a great idea and I happen to have several laying around and alot of damaged MDF to use for disposable projects.

Thanks again for all the support and guidance.

Taylor
 
diyAudio Moderator Emeritus
Joined 2001
diyAudio Moderator Emeritus
Joined 2001
I should point out that, based on Mr. Risch's message board posts and their replies, his designs result in a transmission line where the port output is NOT significant. Essentially, it is very similar to a sealed box. The project pictured is based on his ideas, so we can expect the cone to do all the work in this particular project.
http://www.audioasylum.com/audio/speakers/messages/11476.html
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.