It seems that the overall length of the transmission line behind the driver directly effects the low frequency kick to the speaker. Many sources talk about the benefit of tapering the line as it extends away from the driver. Here's my question:
Thanks.
.DS.
if you were to design a TL application that had a fork in the transmission line - say at one of the taper points - would the two lines add together to form a "longer" total line?
(or)
would the two lines average, thereby (just guessing here) giving you more sound pressure but at a higher frequency than the longer line?
(or)
would you effectively have your speaker tuned to two different frequencies?
(or)
would the two lines average, thereby (just guessing here) giving you more sound pressure but at a higher frequency than the longer line?
(or)
would you effectively have your speaker tuned to two different frequencies?
Thanks.
.DS.
Download this program, sim it yourself and see. It's super easy to use. Download | Leonard Audio
Not sure about "kick" but a longer line equals a lower tuning, all else being equal.
Tapering the line affects whether the tl will act more like a horn or a ported box or somewhere in between. It also can make it a lot easier to get a flatter frequency response, along with offsetting the driver along the line length.
Two (or more) divergent parallel paths will NOT add together to form a longer total line.
They do effectively average, not tune to separate frequencies. How they average and what response you will see depends on the size, shape and length of the paths as well as the size shape and length of the beginning of the line and the driver details including location.
If one of the parallel paths is open and the other is closed, the closed ended stub will form an acoustic trap defined by the shape, size and length of the closed path and it's location in the line.
Not sure about "kick" but a longer line equals a lower tuning, all else being equal.
Tapering the line affects whether the tl will act more like a horn or a ported box or somewhere in between. It also can make it a lot easier to get a flatter frequency response, along with offsetting the driver along the line length.
Two (or more) divergent parallel paths will NOT add together to form a longer total line.
They do effectively average, not tune to separate frequencies. How they average and what response you will see depends on the size, shape and length of the paths as well as the size shape and length of the beginning of the line and the driver details including location.
If one of the parallel paths is open and the other is closed, the closed ended stub will form an acoustic trap defined by the shape, size and length of the closed path and it's location in the line.
Tapering allows for a shorter line.
With a bifurcation in the line, the cross-sections add and the net length is the average.
Driver offset, initial pre-volume, mass loading can all be used to improve the terminus low pass function (you can then use less damping and get more bass quantity).
Do model the line before building.
dave
With a bifurcation in the line, the cross-sections add and the net length is the average.
Driver offset, initial pre-volume, mass loading can all be used to improve the terminus low pass function (you can then use less damping and get more bass quantity).
Do model the line before building.
dave
Are you thinking of something similar to this? (below) it was just a crazy thought I had, no promises, facts, conclusions or evidence, just a thought, that's all, and I'm painfully aware that it is a fair bit "out there" 
Martinsson's Blog - DSTHAM - Dual Summation Tapped Horn - just a thought
Martinsson's Blog - DSTHAM - Dual Summation Tapped Horn - just a thought
Thank you.
About 10 minutes after posting I realized there is no way the two lengths add. Then I thought they averaged, so thanks for your reply. At the same time, a fork in the line will effectively thwart whatever tapering effect is in the line's cross section if there are two smaller cross sections at the fork.
In conclusion, thought I was onto something with a fork, but given that tapering itself help shapes the sound quality, forking seems to me will have no real benefit.
Thanks again for your answer and not calling me an idiot in it.
DS
About 10 minutes after posting I realized there is no way the two lengths add. Then I thought they averaged, so thanks for your reply. At the same time, a fork in the line will effectively thwart whatever tapering effect is in the line's cross section if there are two smaller cross sections at the fork.
In conclusion, thought I was onto something with a fork, but given that tapering itself help shapes the sound quality, forking seems to me will have no real benefit.
Thanks again for your answer and not calling me an idiot in it.
DS
Any words you might have on the behavior of tapering I would enjoy reading. How does tapering work? Does a tapered line go lower than a line of the same size with a fixed cross section?
Though I will search for "driver offset", any summary you might have about that I'd enjoy also. My application involves a forward firing driver with the TL behind it and the baffle.
Thanks for not calling my "add the two lengths" hypothesis stupid - I thought I was onto something.
Ds
Though I will search for "driver offset", any summary you might have about that I'd enjoy also. My application involves a forward firing driver with the TL behind it and the baffle.
Thanks for not calling my "add the two lengths" hypothesis stupid - I thought I was onto something.
Ds
(Does your design have a driver that is firing inside the cabinet {rather than outside into free space/"the room"}?)
Sorry, but my Winnie The Pooh brain doesn't understand your diagram. I sincerely thank you for the effort, clarity and detail that is in it, though.
All the best ...
.Ds.
Sorry, but my Winnie The Pooh brain doesn't understand your diagram. I sincerely thank you for the effort, clarity and detail that is in it, though.
All the best ...
.Ds.
Are you thinking of something similar to this? (below) it was just a crazy thought I had, no promises, facts, conclusions or evidence, just a thought, that's all, and I'm painfully aware that it is a fair bit "out there"
Martinsson's Blog - DSTHAM - Dual Summation Tapped Horn - just a thought
Are you thinking of something similar to this? (below) it was just a crazy thought I had, no promises, facts, conclusions or evidence, just a thought, that's all, and I'm painfully aware that it is a fair bit "out there"
Martinsson's Blog - DSTHAM - Dual Summation Tapped Horn - just a thought
Hi martinsson,
I read your very nice Blog and got a DejaVu: See the picture.
Your intention is to improve a Tapped Box by Dual summation..
Can you provide an Math. Electro Acoustic Calculation or a Lumped Circuit Model where one can follow your thinking?( I mean No 'Tombola-ing' of Simulation Programs). IMO:AkAbak will not do as providing too simplified Lumped Circuits if not expanded with elaborate custom Math. modules.
How are the Losses distributed(eventual a Box Damping Scheme) together with the Bend folding effects and what obviously different distributed Complex Wave-numbers are you using in your Calculation( 2-D,ev. 3-D.)
Do you count the evanescent Waves too? i.e. eventually a H.O.M. Calc.?:
You're talking about an improved Design for a Full-range..Interesting!..
b
Attachments
I have no idea about the mathematics or models of calculation required to describe my thoughts in this case beyond that the "key" seems to be to equalize the two path's contributions (fork pressure difference) once they have been tuned for two lightly overlapping pass bands not to out of phase from each other in order to get a least amount of negative or destructive interference at the first and second summations, if that is that is at all possible...
Lot's of ramblings I guess... in picture form I see it kind of like this :
The above serves only as an example if the Idea, the path length differences in this case might end up being totally destructive...
A minor path length difference might be more beneficial (below) if the idea holds water that is (totally unproven or backed up by any form of facts or evidence).
I do not want this to get off topic, if that is the case please let me know and I'll stop
Lot's of ramblings I guess... in picture form I see it kind of like this :
The above serves only as an example if the Idea, the path length differences in this case might end up being totally destructive...
A minor path length difference might be more beneficial (below) if the idea holds water that is (totally unproven or backed up by any form of facts or evidence).
I do not want this to get off topic, if that is the case please let me know and I'll stop
Last edited:
Hello ...
Your first drawing here is what I thought you were doing, but wasn't sure. Thank you for supplying it. My application had one TL emerging from the rear of the cone, and then forking off about 70% down the line. My thinking was, similar to the way electricity "finds" the lowest R path in a circuit "before" it actually encounters the circuit, that the total sound output (SP) spectrum would be more efficiently tuned by the dual path. The higher frequencies of SP would "find" the shorter path, the lower frequencies would "find" the longer path, and greater dampening efficiency would be achieved.
Really great drawing. Could see everything that was going on it. Thank you for supplying it.
.Ds.
Your first drawing here is what I thought you were doing, but wasn't sure. Thank you for supplying it. My application had one TL emerging from the rear of the cone, and then forking off about 70% down the line. My thinking was, similar to the way electricity "finds" the lowest R path in a circuit "before" it actually encounters the circuit, that the total sound output (SP) spectrum would be more efficiently tuned by the dual path. The higher frequencies of SP would "find" the shorter path, the lower frequencies would "find" the longer path, and greater dampening efficiency would be achieved.
Really great drawing. Could see everything that was going on it. Thank you for supplying it.
.Ds.
I have no idea about the mathematics or models of calculation required to describe my thoughts in this case beyond that the "key" seems to be to equalize the two path's contributions (fork pressure difference) once they have been tuned for two lightly overlapping pass bands not to out of phase from each other in order to get a least amount of negative or destructive interference at the first and second summations, if that is that is at all possible...
Lot's of ramblings I guess... in picture form I see it kind of like this :
The above serves only as an example if the Idea, the path length differences in this case might end up being totally destructive...
A minor path length difference might be more beneficial (below) if the idea holds water that is (totally unproven or backed up by any form of facts or evidence).
I do not want this to get off topic, if that is the case please let me know and I'll stop
... even knowing what terms to google is part of the learning process. Thanks for the input.
.Ds.
.Ds.
MJK's alignment tables article summarizes both tapering and driver offset nicely and offers suggestions on what to build if you don't want to simulate.
http://www.quarter-wave.com/TLs/Alignment_Tables.pdf
This article covers up to 1:10 and 10:1 taper ratio. If you go beyond these limits the tl will work increasingly as a ported box (negative taper) or as a horn (positive taper).
The article also covers stuffing.
A tapered line would go lower than a constant cross sectional area line of the same size, although these two lines would never be the same size as only one of them could have a decent frequency response. If you sized it so the tapered version had decent FR the other would not, and vice versa.
I have no idea about the mathematics or models of calculation required to describe my thoughts in this case beyond that the "key" seems to be to equalize the two path's contributions (fork pressure difference) once they have been tuned for two lightly overlapping pass bands not to out of phase from each other in order to get a least amount of negative or destructive interference at the first and second summations, if that is that is at all possible...
Lot's of ramblings I guess... in picture form I see it kind of like this :
The above serves only as an example if the Idea, the path length differences in this case might end up being totally destructive...
A minor path length difference might be more beneficial (below) if the idea holds water that is (totally unproven or backed up by any form of facts or evidence).
I do not want this to get off topic, if that is the case please let me know and I'll stop
The software I linked to cannot simulate this. It can sim two divergent paths but not two paths that converge again later in the line. Akabak can simulate this just fine. This is really not much different from a bandpass in that chambers (and drivers) port into each other. This is what Akabak is made for.
Last edited:
... I looked over your tables paper. I did not see anything that spoke to a closed TL. The paper only deals with the open ended case. If you have any general behavioral descriptions for a closed TL (driver at one end of the TL, firing out into space, the far end from the driver closed), I would appreciate it.
1) the longer you can make the TL, the deeper the predominant frequency of your system will be - true?
2) tapering the closed system makes the TL "appear" or behave longer than it actually is - true?
Thanks.
.Ds.
1) the longer you can make the TL, the deeper the predominant frequency of your system will be - true?
2) tapering the closed system makes the TL "appear" or behave longer than it actually is - true?
Thanks.
.Ds.
( does the moderator automatically strip out formatting? If so, why offer it? )
You say "needs to be twice as long". Twice as long as what? Please clarify.
(also, great picture).
.Ds.
You say "needs to be twice as long". Twice as long as what? Please clarify.
(also, great picture).
.Ds.
A closed line is a ½ wl resonator. Acts like a sealed box. Needs to be twice as long. The goal to completely absorb the back wave before it returns to the back of the driver.
dave
For emphasis, not for global use. Please leave your posts to the default (which every viewer can set to their preference)
dave
ha ... didn't even see you were a moderator!
For emphasis, not for global use. Please leave your posts to the default (which every viewer can set to their preference)
dave
Yes, the paper is only concerned with open ended transmission lines. There's no real need for a closed tl paper since a closed box is just a closed box.
Is there a reason you don't want to do a quick and simple sim to find these answers? It took me about 5 minutes to do these sims, it takes longer to create and host the picture than to do the sims.
For each of these graphs the enclosure size is 400 liters and the enclosure is sealed. The only thing that changes between sims is shape.
first graph - 4 meters long, beginning csa 1000 cm, end csa 1000 cm (long, straight, constant csa tl)
second - 1 meter long, beginning csa 4000 cm, end csa 4000 cm (short, straight, constant csa closed box too short to be tl)
third - 4 meters long, beginning csa 182 cm, end csa 1820 cm (long, 1:10 ratio positive taper tl)
fourth - 4 meters long, beginning csa 1820 cm, end csa 182 cm (long, 10:1 ratio negative taper tl)
An externally hosted image should be here but it was not working when we last tested it.
Notes -
- in each graph the blue line is the IB response for the driver, in each case regardless of enclosure shape the response tracks the IB response pretty well
- none of these graphs have losses or stuffing applied so the spiky responses shown will not be nearly as bad in practice when measured in real life
- there's no secrets hidden in unshown graphs - impedance is what you would expect by looking at the FR, displacement is ~= for each shape, there's really not much advantage to any shape when it comes to sealed boxes; the horn shaped tl has the worst response of the lot but it's not really that deviant compared to the others, but a good guideline would seem to be that if you want to make it very long you might also want to incorporate a negative taper as well
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.