My TL/Bi-Amped Powered Speaker Project

[Nedrex]

Hey All! This being my first thread (and, come to think of it: first post) on DIY, I thought I'd make it a good one.

This project was spawned by a combination of quite a few different things. Primarily, the need for good audio in my bedroom (where the presently is none whatsoever), my desire to dabble in a transmission line enclosure, and my discovery of Dayton Audios MCA2258 Bi-Amp. Having recently setup a bi-amped system in my car I am very much "sold" on what an improvement bi-amplification makes.

So, after much brainstorming and a few revisions, I settled on the M6N woofer and TN28 tweeter from HiVi. I had heard some positive feedback on the M6N being used in a TL alignment and thought the TN28 would be a good match for it both aesthetically and sonically. I (now) know that the TN28 has met some mixed reviews, and plan on doing some listening myself before I go tearing the rear chamber off to pack it with acousta-stuff (as seems to be the "solution").

I'm working with the (slightly outdated, I know) 5th edition of the Loudspeaker Design Cookbook and took all my ques for the TL design from there. That being the case I tuned the line to 50hz, which is just above the M6Ns resonance (46hz). This dictated a 1/4 wave length of 68 inches, which I addressed with a once folded line, each half being 34". I settled on a taper ratio of 1.52, looking for a deep and slightly fat (as opposed to purely tight) bass response. The line starts off with a cross-sectional area of 32in^2 for the first 38 inches, then goes to 24in^2 for about the next 18 inches, then goes down to 21in^2 for the last 12 inches before reaching the rear facing terminus which is also 21in^2, compared to the woofers radiating area which is 20.6in^2.


Here you see the cabinet front on the left, back on the right. The front has some vinyl-based vibration dampening material already installed, as well as 8 pegs to keep the acousta-stuff from settling.


The equations in the LDC suggested an amount of the acousta-stuff of roughly 0.17lb/ft^3 (the volume of the line is roughly 1.1ft^3 btw). This sounded pretty low to me, so I conducted some listening tests once the cabinets were ready. I did a side by side comparison of 1/4lb vs 1/8lb, and sure enough: 1/8lb was much closer to the bass response I was looking for. 1/4lb sounded thin, whereas 1/8 sounded nice, fat and deep. Tho it did have a bit of a "honk" (or "cupped" tonality) in the upper bass range. I felt I could take care of this by going with an amount right in the middle. To that end I combined and then halved the 1/4lb and 1/8lb. A subsequent listening test proved that 3/16lb per box sounded just right.


1/4lb compared to 1/8lb.


First round of listening tests.


0.1875lb of acousta-stuff, evenly distributed throughout 75% of the line.


Second round of listening tests.



And that, my fellow DIYers, is essentially where I'm at for now.
Yesterday we got the front and rear cabinet panels glued in place. The sides and bottom have been primed (front top and back will have a cherry veneer), and the mounting parts for the spikes have been installed. Next weekend I hope to attack the veneer, get the final paint on and even (dare I say) finish 'em up. But first, a few questions for the forum:

1. I plan on aligning the acoustic centers of the tweeter and woofer. Naturally this will mean moving the tweeter back a bit, which leads me to worry about initial reflections off the hard top corner of the cabinet (the tweeter will be mounted in the center front top of the cabinet). I would like to address this by giving the top front corner of the cab a 3/4" radius, however I wonder how well (or otherwise) the veneer will take to this curve. Any thoughts on this (either the initial reflections issue or the veneer dealing with the 3/4" radius)?

2. At present I do not plan on using any additional crossover/output shaping electronics aside from the MCA2258's internal 3000hz 3rd order crossover. Maybe add a 8.8 microfarad cap inline to the tweeter as an added measure of security. Any thoughts on this? Does anyone with experience with the TN28 have any suggestions on some impedance/response shaping networks? Or likewise for the woofer? Any thoughts in this would be appreciated.

And certainly any thoughts or questions otherwise are welcome. Thanks for reading!!

More pics for the heck of it:


My father applying glue to the one of the cabinet rear panels. I have him to thanks for his woodworking expertise, lord knows I could not have made these cabinets half as well as he did.


Testing the amps for fitment, clearance and placement.


Rear panels glued and clamped.

Stay tuned for more soon!

[mayhem13]

I'm working on a project with those amps for a 2 way MTM nearfield listening. I'm interested to hear your opinions on the amps so far? My design is fairly efficient at around 92db.

[Nedrex]

My only opinions so far on the amps are that the build quality seems nice and the at the cutout dimensions are a little too large in the width dimension. We followed the cutout dimensions exactly and now we're a little worried about the mounting screws getting enough meat of the MDF on the sides. Once I actually hear them I'll be happy to give a full review thereof.

[planet10]

Originally posted by Nedrex

I'm working with the (slightly outdated, I know) 5th edition of the Loudspeaker Design Cookbook and took all my ques for the TL design from there. That being the case I tuned the line to 50hz, which is just above the M6Ns resonance (46hz). This dictated a 1/4 wave length of 68 inches, which I addressed with a once folded line, each half being 34". I settled on a taper ratio of 1.52, looking for a deep and slightly fat (as opposed to purely tight) bass response. The line starts off with a cross-sectional area of 32in^2 for the first 38 inches, then goes to 24in^2 for about the next 18 inches, then goes down to 21in^2 for the last 12 inches before reaching the rear facing terminus which is also 21in^2, compared to the woofers radiating area which is 20.6in^2.

None of your pictures show the line itself... i am having trouble visualizing the description of the line folding you have.

I've not seen Dickason's TL coverage... does he cover off that a line that tapers needs to be shorter than a straight xSection line?

dave

[Nedrex]

Sorry if I didn't make it clear: if you look at the first picture, the cabinet on the left shows the first half of the line (M6N mounted at the top opening on the bottom leads to the rear half), and the cabinet on the right shows the rear half of the line (terminus at the top).

[planet10]

OK, so the TL is formed by a single sloping panel? The most common assembly method has one of the sides going on last which makes it easy to see the folding...

What criteria did you use to choose the line cross-section?

dave

[Nedrex]

Originally posted by planet10
OK, so the TL is formed by a single sloping panel? The most common assembly method has one of the sides going on last which makes it easy to see the folding...

The TL is formed by a single panel, but it is vertical, not sloping (in retrospect, sloping would have been nice, since it would have allowed a constant gradual taper... but hey: live and learn). The box overall internal dimensions are 34"h x 8"w x 7.75"d. The vertical internal partition starts at the top and divides the cabinet into a front half (34 x 8 x 4) and a rear half. The 180 degree turn happens at the bottom. The cross-sectional area of the opening at the bottom is consistent with the rest of the first half of the line: 8" x 4", but then after the turn is made the rear portion of the line is 8" x 3" for about 18" (don't have the plans or speakers in front of me, so these dimensions are from memory). After the 8" x 3" x 18" section it tapers slightly to 8" x 2.625" for the last 12" of the line. This final taper also roughly corresponds to the end of the acousa-stuff. Then the line ends with the terminus at the top back being 7" x 3".

Originally posted by planet10

What criteria did you use to choose the line cross-section?

Basically, I started with the LDCs instructions to have the line start off with a cross-sectional area at least 25% greater than the radiating area of the driver and have it taper to being equal to the drivers radiating area for the last bit of the line. The LDC also talks about taper ratios in the area of 1.25 having a "tighter" bass quality whereas ratios of 1.50 or greater have a deeper bass quality. Since I wanted these to be a stand-alone pair (no sub), I went with a taper ratio of just over 1.5, which keeping the terminus close to equal to the radiating area of the M6N dictated a starting cross sectional area of 32 square inches.

Originally posted by planet10

I've not seen Dickason's TL coverage... does he cover off that a line that tapers needs to be shorter than a straight xSection line?

Hmmm, while I don't recall him saying that in so many words, he does suggest that you tune the line slightly above the drivers resonant frequency.... maybe that is why?

[planet10]

Originally posted by Nedrex
Basically, I started with the LDCs instructions to have the line start off with a cross-sectional area at least 25% greater than the radiating area of the driver and have it taper to being equal to the drivers radiating area for the last bit of the line. The LDC also talks about taper ratios in the area of 1.25 having a "tighter" bass quality whereas ratios of 1.50 or greater have a deeper bass quality. Since I wanted these to be a stand-alone pair (no sub), I went with a taper ratio of just over 1.5, which keeping the terminus close to equal to the radiating area of the M6N dictated a starting cross sectional area of 32 square inches.

So he is mostly regurgitating Classical design techniques... while these can produce quite musical loudspeakers, they don't (usually) come near getting the most possible out of a driver... with modern TL models like Martin King's & George Augspurger's, it is possible to do better. Martin King's alignment tables are a good place to start.

I don't mean to rain on your parade, you will likely end up with a perfectly decent speaker -- you have clearly put a lot of effort into these. You'll have to excuse me, i've been waging a campaign for the last 8 years to help bring people the word about modern TL design (i'm guilty of spreading much of the classical line (before Martin) -- i wrote an in-house paper in the late 70s that likely isn't far from Dickason's treatment of the subject -- King's work & augspurger's were released pretty much simultaneously in late 1999). Lines designed the way yours is, and stuffed as yours are, are properly termed classic aperiodic TLs...

dave

[Nedrex]

I don't mean to rain on your parade, you will likely end up with a perfectly decent speaker -- you have clearly put a lot of effort into these....Lines designed the way yours is, and stuffed as yours are, are properly termed classic aperiodic TLs...

Oh hey, don't give it a second thought! I figured from the outset that (since I was working with an out-dated edition of the LDC) I wouldn't be "in the know" with all the latest and greatest trends in TL design. But in a way, that wasn't the point... like most DIYers (at least I imagine) I love the process. And if I can finish this thing and have a set of speakers that I can be proud of, that can be pleasing sonically and aestheically for years and years, then I'm all set.

And besides, knowing that there are better ways to execute a TL alignment just gives me an excuse to build another set!!

[planet10]

Originally posted by Nedrex
And besides, knowing that there are better ways to execute a TL alignment just gives me an excuse to build another set!!

that's what i like ... a guy who just knows he is going to become an addict

dave