Not sure if already discussed, but how significant is the contribution of a non-parallel wall in a tapered TL construction towards fighting the internal box reflected waves? As far as I know, the parasitic harmonics need as much as bouncing as possible to get caught the most into the stuffing, before reaching back to the driver or port. Is a straight TL considered evil in this aspect and if yes, how can it be fought without tapering? Adding some asymmetric construction inside, like braces?
Last edited:
Its mass loading and heat or friction. But its created a long LP filter and compared to a straight or expanding shaped pipe it stands out in its affective ‘qw Fb for its length(shorter). at some point thats bass reflex, just diverge from the blended shapes into distinct ones (more than a Dalene TL). The MLTL is pretty neat, it would be interesting to stick a mic up inside or drag it with an exterior magnet(fish tank glass cleaning tool?) and find the pressure nulls and put them at harmonics by design(fold at the those points)? But also in the ‘straight csa TL, to compare . And answer parts of your question, cause its a good one!!
Last edited:
After specific frequencies, yes. But what about the lower midrange, 200 to 400Hz? The closest parasitic harmonic is removed with driver offset.
One tip I know is making the depth (distance from the driver to its rear wall as small as possible, to avoid big wavelengths (low frequencies) developing.
One tip I know is making the depth (distance from the driver to its rear wall as small as possible, to avoid big wavelengths (low frequencies) developing.
Makes sense! But thinking pragmatically, if you find out where these hot spots are, what follows? Some kind of local Helmholtz resonators?
A tapered line vrs a straight one will be a shorter but they usually end up the same volume.
The taper has little or no effect on reducing the (usually) close to non-existent side-to-side waveform because they are at such a high frequency that the absorbion used to kill the unwanted quarter-wave harmonics is more than sufficient to kill higher frequencies.
I have found that getting any interior walls too close to the driver leads to significant mid-range/voice colouration.
dave
The taper has little or no effect on reducing the (usually) close to non-existent side-to-side waveform because they are at such a high frequency that the absorbion used to kill the unwanted quarter-wave harmonics is more than sufficient to kill higher frequencies.
I have found that getting any interior walls too close to the driver leads to significant mid-range/voice colouration.
dave
I've found out during the latest years, when listening to bigger box projects of a same constructor, the larger ones having sounded like having more box coloration than the smaller ones. My only two explanations for this phenomena are the possible lack of bracing and probably lower frequency standing waves developing inside. Some bigger box speakers have even sounded colored from the drivers themselves (possible leaking) and not from the port.
So I'm a bit confused. One of our local speaker designers from whom I heard one of the most transparent ML-TL project so far, said, the best way to treat rear waves is to make the box as little shallow to avoid wave development or to make them as deep as possible, like a Nautilus. But he also puts in a lot of bracing.
The other extremely transparent TL I've ever heard was made like an Ariel. There wasn't even driver offset. A lot of folding downwards into small sections, which I presume, gives included bracing. This TL sounded like an OB. Here's a video of the project, you can see construction photos: Stoy40 TL - YouTube
So I'm a bit confused. One of our local speaker designers from whom I heard one of the most transparent ML-TL project so far, said, the best way to treat rear waves is to make the box as little shallow to avoid wave development or to make them as deep as possible, like a Nautilus. But he also puts in a lot of bracing.
The other extremely transparent TL I've ever heard was made like an Ariel. There wasn't even driver offset. A lot of folding downwards into small sections, which I presume, gives included bracing. This TL sounded like an OB. Here's a video of the project, you can see construction photos: Stoy40 TL - YouTube
Last edited:
Yes, by sending it in each direction and it coming back in a way that aids cancellation. At this point it's all seen the damping twice.
Proper use of Zd can almost completely null the 1st harmonic regardless of damping.
By doing this one can use less volume fill and retain more of the desired harmonic.
A modern understanding of how TLs work also adds a restricted terminus (leadng to much confusion with BR towers that are not BR), and internal helmholtz absorbers.
dave
dave
By doing this one can use less volume fill and retain more of the desired harmonic.
A modern understanding of how TLs work also adds a restricted terminus (leadng to much confusion with BR towers that are not BR), and internal helmholtz absorbers.
dave
dave
I think a fold that puts a panel thickness between a hi pressure and a null and a ‘orifice size to equalize as needed? Size the ‘hole’ per response needed.
i think the ,taper, idea gets thrown out the window and its just segments of csa increase to support the wAve shape while the folding acts as a LP filter each time? But a taped driver entry is needed? It allows a rearwave to be stronger or more useful? it seems to help at a predicted(in sim 1/4-3/4 split. Fold at 1/4 (Fb) points (2 folds). The tapped driver at that point and add 1/4(Fb)after to exit.
Remove that last segment and youre back to offset driver at 1/3... win win either way. So remove the ‘offset driver position. Now its 1/4-3/4...
Last edited:
Whether it behaves this way will depend on the wavelengths involved. It is possible to decide what frequencies are needed or inevitable, and which will be eliminated. This way you can decide the resolution/length of your taper sections. On the whole though, the expansion should probably conform to the plan.
In the absence of any real expansion on the theme, I assume it was an arbitrarily selected value intended purely to ensure the small dimensions he would ideally favour in ensuring an air load that was c. 'pure compliance' and without pipe resonance in the 'low hundreds of Hz region' he was concerned about.
I have used 200/400/800/1600cm2 and a driver entry at 0 and 200 of 400 cm long. this if folded, seemed to be needing that as a ‘LP filter’ at halfway in each?
Or at 120/120/120/120cm it made sense with two driver entries. i never built that(yet) only a 120/80 tap and 120/80 is left to try or 120/120 twice version) ive got a 120/120 tap then 120 after built. and an offset (120/120 parallel to an 80cm) which exit together... folding was a bit off... in splitting each by the needed in between to link them...
If anyone can suggest better details in this to improve it im looking for those ideas. I cant sim a fold, i use horn response...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.