Anyone familiar with this excellent piece of work? I am interested in learning how to design without software, at least for early planning of transmission line subwoofers.
There are 4 tables. I understand the first table and the fourth table, regarding length and driver offset. I do not understand table 2 and 3 concerning dz and dr. I know they are used to calculate the shape but I don't even understand what shape function and resistance factor mean.
I know I'm new here but please don't tell me to read the article again, as I've studied it for hours and don't get it.
There are 4 tables. I understand the first table and the fourth table, regarding length and driver offset. I do not understand table 2 and 3 concerning dz and dr. I know they are used to calculate the shape but I don't even understand what shape function and resistance factor mean.
I know I'm new here but please don't tell me to read the article again, as I've studied it for hours and don't get it.
Dan, thanks very much, but you are a bit above my level, I'm pretty new at this. Let's see if I understand what you are saying.
Choosing any value in table 1 shows you the tuning frequency and SL/SO for that particular length.
The corresponding spot on table two defines a value for dz. After this point it gets a bit fuzzy for me, probably mostly because I do not understand what shape function is or how it relates to what I am trying to do.
You said "Once you have Dz you can solve for Dr using the last function on page 3." I thought we got dr from table 3, defined by the driver's qts. Did you maybe mean take the values for dr and dz from tables two and three and use them to solve the last equation on page 3 to determine area of the closed end?
That would seem to make sense to me, because once you know SO you can calculate SL defined by the ratio chosen in table 1, and finally enough information to draw a design on paper.
If I am keeping up so far, that means there is only one formula to solve, correct?
There lies another problem, I only have grade 10 basic math. I could probably solve the equation, with the values of dr and dz from the table, the value of Re from driver specs (I assume), but what are the values of B, L and that symbol for acoustic impedance? And aside from the values, what are those 3 things?
That's a lot, so I'll just shut up now and thank you in advance for your help.
Choosing any value in table 1 shows you the tuning frequency and SL/SO for that particular length.
The corresponding spot on table two defines a value for dz. After this point it gets a bit fuzzy for me, probably mostly because I do not understand what shape function is or how it relates to what I am trying to do.
You said "Once you have Dz you can solve for Dr using the last function on page 3." I thought we got dr from table 3, defined by the driver's qts. Did you maybe mean take the values for dr and dz from tables two and three and use them to solve the last equation on page 3 to determine area of the closed end?
That would seem to make sense to me, because once you know SO you can calculate SL defined by the ratio chosen in table 1, and finally enough information to draw a design on paper.
If I am keeping up so far, that means there is only one formula to solve, correct?
There lies another problem, I only have grade 10 basic math. I could probably solve the equation, with the values of dr and dz from the table, the value of Re from driver specs (I assume), but what are the values of B, L and that symbol for acoustic impedance? And aside from the values, what are those 3 things?
That's a lot, so I'll just shut up now and thank you in advance for your help.
just a guy
Take a look at this thread...
http://www.diyaudio.com/forums/showthread.php?postid=678731#post678731
There are two downloads in those posts.
Planet10's download of an AlphaTL article by Rick Schultz (post 9)
Give it a read
GM's xls spreadsheet from the same article (post 11)
Experiment with it (using the article).
This is a "simpler" way of doing TLs
It may help you get jump-started.
You will see that you only need three driver TS parameters.
Qts
Vas
Fs
Also, be sure to review the following...
http://www.t-linespeakers.org/design/MJK-for-dummies/index.html
http://www.diysubwoofers.org/
Sure hope this helps...
Take a look at this thread...
http://www.diyaudio.com/forums/showthread.php?postid=678731#post678731
There are two downloads in those posts.
Planet10's download of an AlphaTL article by Rick Schultz (post 9)
Give it a read
GM's xls spreadsheet from the same article (post 11)
Experiment with it (using the article).
This is a "simpler" way of doing TLs
It may help you get jump-started.
You will see that you only need three driver TS parameters.
Qts
Vas
Fs
Also, be sure to review the following...
http://www.t-linespeakers.org/design/MJK-for-dummies/index.html
http://www.diysubwoofers.org/
Sure hope this helps...
Hi, qi, the article "Pearls from..." is an excellent resource and was the basis for trying to figure out the alignment tables. There are two reasons why I am not interested in MJK's full works at this time.
First, if I can't manipulate the alignment tables I don't like my chances with the real deal. Second, my current project can't be modelled with his spreadsheets, but I'm still interested, even though my current project is past the planning stage I am interested in continuing to work with tl's.
Looks like I have a lot of studying to do with the alpha article and playing with GM's spreadsheet. Thanks for the info.
First, if I can't manipulate the alignment tables I don't like my chances with the real deal. Second, my current project can't be modelled with his spreadsheets, but I'm still interested, even though my current project is past the planning stage I am interested in continuing to work with tl's.
Looks like I have a lot of studying to do with the alpha article and playing with GM's spreadsheet. Thanks for the info.
Have you seen the Sample Design Problem in pg13? It shows a solved example with the values of
ρ(pronounced 'rho',density of air = 1.21 kg/m3),
c(speed of sound in air = 342 m\s).
The values of Bl(force factor),Sd(effective cone dia) and Re(DC resistance of voice coil) should be mentioned in your driver specs.
It's then a straightforward task of solving for So where
So=(ρ x c x (Sd)^2 x Dr x Dz x Re) / (Bl)^2
Remember Sd is in meter-square.
Thank you so much, that's exactly what I was looking for. I studied the sample for awhile and thought that equation might amount to the one referred to on page 3 but could not be sure at all. That is some math that I can actually do, and finally I can use the alignment tables to their full advantage.
Qi, I read the alpha article and it shocked me to no end. It specifically contradicts just about everything I have ever read about tl's, it goes against the "Pearls from..." article and no surprise it completely undermines the MJK alignment tables and probably the full spreadsheets as well.
The only explanation I can think of is that it is the fastest and easiest way to make a small FULLRANGE tl. If FO cannot be lower than FS it does not give much consideration for real sub tl's, certainly not of the magnitude that are talked about around here. At this point in my studies I firmly believe that the line can be tuned below fs if the driver qts is high, and I also believe that the geometry of the line is integral to the tuning and overall enclosure size, two major points that are argued against in the article but clearly upheld in MJK's alignment tables.
For subwoofers at least I think the alignment tables are a much better resource, and it would seem most people think so. I've never seen anyone give advice to SHORTEN their design to go lower in a subwoofer.
Anyway, thanks everyone for all the help and Qi, it is a Transflex I am making, fully adjustable during the testing phase. I am not trying in any way to copy Danleys sub but I will be experimenting with line geometry by moving the divider to create expanding, straight and decreasing lines as well as playing with driver offset. Function determines form so it may end up looking like Danleys if a similar configuration tests best.
I have been contemplating the Transflex for weeks now on a different forum with a very knowledgeable gentleman who is also a member here and during the conversation the Danley sub was discovered, but I was set on making a Transflex before I ever saw the dts.
Qi, I read the alpha article and it shocked me to no end. It specifically contradicts just about everything I have ever read about tl's, it goes against the "Pearls from..." article and no surprise it completely undermines the MJK alignment tables and probably the full spreadsheets as well.
The only explanation I can think of is that it is the fastest and easiest way to make a small FULLRANGE tl. If FO cannot be lower than FS it does not give much consideration for real sub tl's, certainly not of the magnitude that are talked about around here. At this point in my studies I firmly believe that the line can be tuned below fs if the driver qts is high, and I also believe that the geometry of the line is integral to the tuning and overall enclosure size, two major points that are argued against in the article but clearly upheld in MJK's alignment tables.
For subwoofers at least I think the alignment tables are a much better resource, and it would seem most people think so. I've never seen anyone give advice to SHORTEN their design to go lower in a subwoofer.
Anyway, thanks everyone for all the help and Qi, it is a Transflex I am making, fully adjustable during the testing phase. I am not trying in any way to copy Danleys sub but I will be experimenting with line geometry by moving the divider to create expanding, straight and decreasing lines as well as playing with driver offset. Function determines form so it may end up looking like Danleys if a similar configuration tests best.
I have been contemplating the Transflex for weeks now on a different forum with a very knowledgeable gentleman who is also a member here and during the conversation the Danley sub was discovered, but I was set on making a Transflex before I ever saw the dts.
Greets!
Alpha TLs are classic end loaded stopped pipes (QWP), so the theoretical ideal is to make Fp (F0) = Fs since it damps the impedance peak to mate properly with an impedance matching amp they were originally voiced with, but in today's high DF amp 'world' you can tune it to any Fp you want.
It doesn't 'undermine' any design routine, just merely provides a relatively simple way to design a pipe with a response that mimics an IB in a more compact size, ergo has no practical BW limits.
Right, the pipe's geometry is key to getting the max gain/BW for a given bulk.
GM
FYI, I made an Excel SS in Imperial Units for MJK's, but it does scaling for odd values of Fp and Qts, so requires more inputting/paying attention to get the right results.
Alpha TLs are classic end loaded stopped pipes (QWP), so the theoretical ideal is to make Fp (F0) = Fs since it damps the impedance peak to mate properly with an impedance matching amp they were originally voiced with, but in today's high DF amp 'world' you can tune it to any Fp you want.
It doesn't 'undermine' any design routine, just merely provides a relatively simple way to design a pipe with a response that mimics an IB in a more compact size, ergo has no practical BW limits.
Right, the pipe's geometry is key to getting the max gain/BW for a given bulk.
GM
FYI, I made an Excel SS in Imperial Units for MJK's, but it does scaling for odd values of Fp and Qts, so requires more inputting/paying attention to get the right results.
Thanks for all the great info. GM, I probably should have mentioned that if I were to make a small fullrange tl, the alpha is a great model, in my opinion. (Based only on the frequency response graphs in the article)
I had a feeling that I would get my answers if I just waited a couple of days. Thanks guys.
I had a feeling that I would get my answers if I just waited a couple of days. Thanks guys.
Greets!
I'm higher math 'challenged', so don't waste my time with trying to understand all the underlying math, instead concentrate on deciphering any examples, which in this case begins on pg. 13. Anyway, if you 'run the numbers' using the formula on pg. 3, the TL's cross sectional area (CSA) would be larger than it needs to be in most (all?) alignments, so a greater stuffing density would be required, though of course a 'newbie' to TL design isn't expected to know this.
Regardless, the formula is wrong IMO and is further grossly misleading to the extent that it assumes you understand what the formula is actually calculating, which isn't the SO's CSA, but a factor to multiply Sd by to get it. Apparently he assumes that the reader will realize this when either studying his referenced Focal design or the example on pg. 13, etc., or inputting a design in the WS.
This points out another apparent 'error' IMO in that the SO/Sd's '/' notation to define this multiplication factor indicates to me that I should be dividing rather than multiplying with it. Better IMO to label it Sdx or somesuch even though his WS shows that it's a multiplier.
FYI, I sent MJK a 'heads up' to this post, so maybe he'll take time to comment.
GM
I'm higher math 'challenged', so don't waste my time with trying to understand all the underlying math, instead concentrate on deciphering any examples, which in this case begins on pg. 13. Anyway, if you 'run the numbers' using the formula on pg. 3, the TL's cross sectional area (CSA) would be larger than it needs to be in most (all?) alignments, so a greater stuffing density would be required, though of course a 'newbie' to TL design isn't expected to know this.
Regardless, the formula is wrong IMO and is further grossly misleading to the extent that it assumes you understand what the formula is actually calculating, which isn't the SO's CSA, but a factor to multiply Sd by to get it. Apparently he assumes that the reader will realize this when either studying his referenced Focal design or the example on pg. 13, etc., or inputting a design in the WS.
This points out another apparent 'error' IMO in that the SO/Sd's '/' notation to define this multiplication factor indicates to me that I should be dividing rather than multiplying with it. Better IMO to label it Sdx or somesuch even though his WS shows that it's a multiplier.
FYI, I sent MJK a 'heads up' to this post, so maybe he'll take time to comment.
GM
The equation for the CSA at the bottom of page 3 is shown below.
S0 = (p c Sd^2 DZ DR Re) / (BL)^2
I have set the scale factors DZ and DR to produce an area that will lead to a particular SPL response curve. This curve is the goal that I use to design TL enclosures. Different people might have different design goals which would require different values of CSA. In general the larger the CSA the higher the bass response. I have selected what I consider to be a good compromise between bass output and enclosure size. Other people might make different trade-offs and recommend different CSA values. This is only my recommendation. To each his own.
The equation on page 13 is the same equation as shown above but with Sd pulled to the left side to form the ratio I like to use when describing a TL. If you divid both sides of the equations above by Sd you will get the following.
S0/Sd = (p c Sd DZ DR Re) / (BL)^2
I tend to express the CSA of a TL in terms of a ratio with the driver cone area Sd. So the equation above yields S0/Sd = 2.872 on page 13. Since SL/S0 = 1 for the example, SL = 2.872 Sd. I don't see any error, what am I missing?
The "/" sign indicates a ratio between or a division of one variable by another. While you may not agree with my alignment, and you may have a better method of your own, I do not consider anything above an error. So far the feedback I have recieved about TLs designed from the tables has been very positive.
Got the e-mail and I have responded.
Hope that helps,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.