| bigwill |
| Why are transmission lines usually tapered? Are there any sonic benefits of this over just having a line that's the same area all the way through? |
|
|
| MJK |
For a given length, a tapered TL will have a lower tuning frequency then a straight TL. So if you want to tune your TL to say 30 Hz, a tapered TL will be shorter.
As a second benefit, the tapered TL's next harmonic will be higher in frequency so there will be a bigger frequency span between the fundamental (1/4 wavelength standing wave) and the next resonance (3/4 wavelength standing wave). Since the 3/4 wavelength resonance occurs at a higher frequency, the fiber stuffing is more effective at attenuating the resonance and reducing the ripple.
You can see some of these trade-offs in my TL alignment tables for classic TL designs. |
|
|
| jackinnj |
| isn't it just as effective to offset the placement of the driver by 1/3 to 1/4 of the TL length? |
|
|
| Tweeker |
| Offseting the driver doesnt change the frequency at which the 3/4 pipe wavelength resonance occurs, though it does mitigate its effects. |
|
|
| bigwill |
One more thing... the line length of my current design is about 31Hz WITHOUT stuffing. The Fs of my driver is 26Hz. Is the line too long? I don't want to to be tuned below Fs when I've stuffed it
Edit: The line is tapered, so it's probably LOWER than 31Hz :xeye: |
|
|
| MJK |
| quote: | | The line is tapered, so it's probably LOWER than 31Hz |
The first thing you need to do is determine the resonant frequency of your TL geometry. From the statement above it would appear that you don't really know. If you go to my alignmnet tables, you should be able to detemine this frequency.
| quote: | | the line length of my current design is about 31Hz WITHOUT stuffing. The Fs of my driver is 26Hz. Is the line too long? I don't want to to be tuned below Fs when I've stuffed it |
Stuffing is not going to lower the line's tuning frequency significantly. Stuffing will have less than a 1% impact on the frequency. Stuffing will attenuate the resonance and smooth the SPL response. |
|
|
| bigwill |
| I'm sorry to say I don't understand these tables :) Another problem is that I don't have a full set of TS parameters for my drivers. What would happen if I just experimented? |
|
|
| MJK |
| Sounds like you are just going to have to try and make some good choices. It might work out OK or you might learn something that you could apply to the next attempt. |
|
|
| bigwill |
| What a carefully worded version of 'There's no helping you now!' :D |
|
|
| MJK |
| Unfortunately, that is just about right. In my opinion, if you really want to do your own designs you need to perform lots of measurments, a fair number of simulations, and really engineer the enclosure. I don't know of any other way. A few people can do it by intuitive feel from trial and error building of a few prototypes and get great results, they are much smarter then I am. |
|
|
| bigwill |
| I don't have any measuring equipment, experience or knowledge :D I've never even seen a transmission line speaker. I'm just going to dive in and make one and see what happens! It can't hurt to try! |
|
|
| MJK |
| quote: | | I'm just going to dive in and make one and see what happens! It can't hurt to try! |
It definitely will not hurt, and it WILL be fun!
If you dont mind a straight constant cross-sectional area TL, assume a tuning frequency close to your drivers resonance, and calculate the length using :
L = c / (4 x f)
where c = 344 m/sec. f is the tuning frequency in Hz, and L is in meters. This should be an appropriate length. If you use a cross-sectional area that is between 2 and 3 time the driver's cone area and then mount the driver 1/5 to 1/3 of the way along the line from the closed end you will improve your chances of success. Now just stuff with fiber to get the best sound to your ears. Anything outside those general guidelines can be a bit more risky.
Bob Brines has some good guidelines on his site (linked from mine) that might also help. If you think the enclosure is way too big, then you are probably on the right track.
Good luck, |
|
|
| bigwill |
| If I make a non-tapered TL, does it matter if my line area is LESS than that of the cone? |
|
|
| MJK |
I recommend an area of between 2 and 3 times the cone area. This is a guess since I don't know the driver parameters.
Occasionally you see a driver that can get by with an area approaching the cone area of the driver, this is not very common. I personally have not seen a good design with an area less then the driver's cone area.
In general, the smaller the cross-sectional area the less bass you get from the TL. A small line chokes the bass and produces disppointing results.
YMMV, |
|
|
| bigwill |
| The 10 inch driver I have has an Fs of 26Hz. This will be a ridiculously huge design. There must be some trick to make it smaller? The likelyhood of me making a TL is getting less and less :bawling: |
|
|
| GuyPanico |
| You can do what i'm about to attempt, and fold the TL as another member has. ( clicky ) The downside is you cant do a smooth taper, and there will be a turn in your TL. But you could almost cut the TL in half doing this. |
|
|
| GuyPanico |
| quote: | Originally posted by MJK
If you use a cross-sectional area that is between 2 and 3 time the driver's cone area and then mount the driver 1/5 to 1/3 of the way along the line from the closed end you will improve your chances of success.
|
If your sections sheet gives a similiar curve from 0-70hz with the driver at the closed end of the TL, and 1/4 of the way up the TL, is there any advantage to one or the other? |
|
|
| MJK |
| If you cross-over at 70 Hz with a steep slope, then putting the driver at the closed end will produce more gain and probably be better. If you cross-over higher where the 3/4 wave can effect the response then mounting the driver closer to 1/3 of the length would probably be better. |
|
|
| bigwill |
| I'm crossing over at 100Hz actively |
|
|
| navin |
MJK,
does this mean that the more one tapers the line the smaller the TL will be (in length) but also one will get less output?
refering to the table on page 10.
say my TL is 100" long. the fs of my woofer is 20Hz. i would have to have SL/SO = 0.1 that means a 10:1 taper. it also means Dz of 36.482!
also i dont see anywhere where there is a relation between Sd and area of the line. this area i assume is the start of the taper. |
|
|
| MJK |
Hi Navin,
| quote: | | does this mean that the more one tapers the line the smaller the TL will be (in length) but also one will get less output? |
Yes, the more tapered the TL the shorter the length. Why do you think that the output would drop? I have not reached any conclusions on the output, Attachment C shows that the tapered TL's output is sufficient to match the driver's efficiency.
| quote: | | say my TL is 100" long. the fs of my woofer is 20Hz. i would have to have SL/SO = 0.1 that means a 10:1 taper. it also means Dz of 36.482! |
Those values look about right.
| quote: | | also i dont see anywhere where there is a relation between Sd and area of the line. this area i assume is the start of the taper. |
If you follow the sample problem on page 13, it demonstrates how to calculate the line areas as a ratio with Sd. You need the drivers T/S parameters and the values from reading the tables to perform this calculation. Be very careful to make sure you are using the correct units in the calculations.
Hope that helps, |
|
|
| navin |
MJK, I get lost where So/Sd is calculated as per page 13.
i get a number that is 2236.618 and you get a number that is 2.872. |
|
|
| navin |
oops. i punched the numbers wrong. Bl is in the denominator.
ok i get it.
my woofer's specs.
Fs=17Hz, Qts = 0.44, Sd= 0.044m2, BL= 7.8 Tm, Vas=380 liters.
If I tune the TL to 20Hz and have a SL/SO=0.1 I need 103". based on the above specs the So/Sd=3.
Now how do I detmine offset and stuffing? |
|
|
| planet10 |
| quote: | Originally posted by navin
Now how do I detmine offset and stuffing? |
Offset is given as a ratio from the closed end... from memory 0.34 something on a 3:1 line... and i believe i asked Martin the question about stuffing and it was 0.25 lb/cu ft or about 4 g/liter (if my calc is right -- it is time to go to bed) but you could easily play with up to double that.
dave |
|
|
| navin |
| So/Sd=3. Sd=440cm2=70 sq. in.. so So=210 sq. in. ouch! that is huge! what happens if we retain length as 103" but reduce to So=Sd. |
|
|
| MJK |
| quote: | | Now how do I detmine offset and stuffing? |
For the alignment tables stuffing really needs to be done by ear. The other alternative is to put the alignment into the mathCad worksheest and further optimize the geometry and the stuffing.
| quote: | | So/Sd=3. Sd=440cm2=70 sq. in.. so So=210 sq. in. ouch! that is huge! what happens if we retain length as 103" but reduce to So=Sd. |
Decreasing area reduces bass output, increasing area provides more bass output. If you really want to know how much smaller the enclosure could get, you need to model it in the worksheets. I do not predict an optimim design with the alignment tables, I just try and provide a good estimate for an enclosure that will produce TL type bass. Unfortunately, TL's tend to be bigger then was previously thought which is the conclusion I came to years ago with my Focal design. |
|
|
| navin |
u can that again! TL's tend to be much bigger than thought.
presently the woofers i have are in 65liter sealed boxes. there is some boom. partly due to the high Qtc and partly due to the room.
my hope is to reduce the boom. |
|
|
|
