i want to build a bipole transmission line FE167e with 2 fe167e per channel.
straight pipe similar to this:
http://www.t-linespeakers.org/projects/tlB/index.html
problem is i don't know what size of enclosure...
first thing is to determine 1/4 wave length at drivers Fs correct? in this case it is 50Hz so the line should be around 5.5 ft high correct? stuffed it should bring it down to the 1/4 wave length to around 2.5 ft high? is this correct?
so which do i use?
the Sd of both drivers together is 40 sg inches.... so the pipe area should be around 48 to 80 (1.2 to 2 times Sd)...
I tried D/L the mathcad worksheets and couldn't figure it out.
can someone quickly run the numbers and tell me if I am on the right track?
Thanks!
straight pipe similar to this:
http://www.t-linespeakers.org/projects/tlB/index.html
problem is i don't know what size of enclosure...
first thing is to determine 1/4 wave length at drivers Fs correct? in this case it is 50Hz so the line should be around 5.5 ft high correct? stuffed it should bring it down to the 1/4 wave length to around 2.5 ft high? is this correct?
so which do i use?
the Sd of both drivers together is 40 sg inches.... so the pipe area should be around 48 to 80 (1.2 to 2 times Sd)...
I tried D/L the mathcad worksheets and couldn't figure it out.
can someone quickly run the numbers and tell me if I am on the right track?
Thanks!
Greets!
>first thing is to determine 1/4 wave length at drivers Fs correct? in this case it is 50Hz so the line should be around 5.5 ft high correct? stuffed it should bring it down to the 1/4 wave length to around 2.5 ft high? is this correct?
====
~13560/4/51.5 = ~65.8" less the terminus end correction. No, it won't shorten it that much, but this much stuffing will 'suck the life' out of a FR driver if it's not a very hi-Q one so I don't recommend so much stuffing it shortens the line.
====
>the Sd of both drivers together is 40 sg inches.... so the pipe area should be around 48 to 80 (1.2 to 2 times Sd)...
====
It's a function of Vb divided by line length. Minimum Vb appears to be ~2.26ft^3, but some series R is required to flatten the response so assuming 4 ohms, we now ideally need ~12.23ft^3:
L = 59.125"
CSA = 323.329"^2 (~7.86 x Sd)
driver down 26.21"
stuff to suit
No BSC should be required in a typical size room and spaced up like the example will get the drivers up to ear height.
====
>I tried D/L the mathcad worksheets and couldn't figure it out.
====
It's pretty simple really, what couldn't you figure out after reading his various docs?
GM
>first thing is to determine 1/4 wave length at drivers Fs correct? in this case it is 50Hz so the line should be around 5.5 ft high correct? stuffed it should bring it down to the 1/4 wave length to around 2.5 ft high? is this correct?
====
~13560/4/51.5 = ~65.8" less the terminus end correction. No, it won't shorten it that much, but this much stuffing will 'suck the life' out of a FR driver if it's not a very hi-Q one so I don't recommend so much stuffing it shortens the line.
====
>the Sd of both drivers together is 40 sg inches.... so the pipe area should be around 48 to 80 (1.2 to 2 times Sd)...
====
It's a function of Vb divided by line length. Minimum Vb appears to be ~2.26ft^3, but some series R is required to flatten the response so assuming 4 ohms, we now ideally need ~12.23ft^3:
L = 59.125"
CSA = 323.329"^2 (~7.86 x Sd)
driver down 26.21"
stuff to suit
No BSC should be required in a typical size room and spaced up like the example will get the drivers up to ear height.
====
>I tried D/L the mathcad worksheets and couldn't figure it out.
====
It's pretty simple really, what couldn't you figure out after reading his various docs?
GM
well...
i d/l the mathcad explorer 8 and the various files for TL and TQWT designs..
but when I tried to load those files the explorer 8 just went to its opening screen and no TL equations or data...
12 cu ft is a lot of enclosure... do you have the response curves??
i kinda wanted to run the 2 drivers with NO series resistance...
also i was thinking of a straight pipe with the drivers at top and opening at bottom... similar to link i provided.
can you provide me with the formulas you used? or did you use the mathcad worksheets?
Thanks!!!
i d/l the mathcad explorer 8 and the various files for TL and TQWT designs..
but when I tried to load those files the explorer 8 just went to its opening screen and no TL equations or data...
12 cu ft is a lot of enclosure... do you have the response curves??
i kinda wanted to run the 2 drivers with NO series resistance...
also i was thinking of a straight pipe with the drivers at top and opening at bottom... similar to link i provided.
can you provide me with the formulas you used? or did you use the mathcad worksheets?
Thanks!!!
Greets!
> i d/l the mathcad explorer 8 and the various files for TL and TQWT designs..
but when I tried to load those files the explorer 8 just went to its opening screen and no TL equations or data...
====
I had this problem also back when I had Norton 2003, turning off the scripting protection solved the problem. Once I upgraded to 2004, this particular problem went away.
====
>12 cu ft is a lot of enclosure... do you have the response curves??
====
I have the MathCad sim and can send it to you as a MC8 .mcd, but you'll have to solve your problem before viewing it. WRT size, TLs require big Vbs to get any LF gain so you can make it smaller, just understand that its F3 rises with decreasing Vb regardless of what the pipe's effective 1/4WL tuning is.
To get a relatively short pipe and LF gain requires mass loading it either by reverse tapering it or adding a vent, making it an ML-TL.
====
>i kinda wanted to run the 2 drivers with NO series resistance...
====
Well, you didn't say, plus you won't get a ~tonally balanced response unless stuck in a corner due to the driver's medium low Qts, so I chose a design that would perform well out in the room where bipoles ideally need to be.
Regardless, you can run this with less/no series R as required/desired, though you will need to use some other form of EQ to flatten the response, so 'pick your poison', so to speak.........
====
>also i was thinking of a straight pipe with the drivers at top and opening at bottom... similar to link i provided.
====
This is a straight pipe with bottom exit, just the driver is placed along the line where it will yield the smoothest response without having to stuff the life out of it.
If you look at the response of the linked example, it vividly shows what sort of response to expect with an end loaded design. Acceptable for a multi-way with woofer where you want max gain, but not an FR that can't decouple all the 'ripple' effects this much loading causes without stuffing the bejeesus out of it, defeating the reason for end loading it in the first place.
====
>can you provide me with the formulas you used? or did you use the mathcad worksheets?
Thanks!!!
====
You're welcome!
Now-a-days I get the basics (Vb, desired Qes, Fb, minimum vent diameter) using a simple box program (BoxPlot 3.07 in my case), then load these into a proprietary spreadsheet I developed to determine optimum length, driver position. I only use MathCad to get an idea of how much stuffing is required in some designs and/or vent length since no other program is as accurate once the shape morphs into a high aspect ratio. Being rather math challenged, I never could figure out how to calc the pipe's effect on it so almost always had to test the speaker to determine how much to shorten it. After seeing MJK's math, it's no wonder as to why!
For doing your own from scratch, I highly recommend you start with MJK's 'Classic' design tables to get a first approximation, or possibly final, design if you don't want to gamble on mine.
Anyway, if you insist on doing an end loaded design using a relatively short/low Vb pipe with no series R, then the F3 could wind up as high as 80-100 Hz if you don't mass load it. That's a lot of BW/efficiency to give up to keep the pipe small. Might as well just stick them in a sealed cab.
Indeed, for best transient response and near IB performance, make a sealed pipe of whatever length works best and the largest CSA you can tolerate (at least 2-3 x Sd depending on length), then EQ flat in-room.
GM
> i d/l the mathcad explorer 8 and the various files for TL and TQWT designs..
but when I tried to load those files the explorer 8 just went to its opening screen and no TL equations or data...
====
I had this problem also back when I had Norton 2003, turning off the scripting protection solved the problem. Once I upgraded to 2004, this particular problem went away.
====
>12 cu ft is a lot of enclosure... do you have the response curves??
====
I have the MathCad sim and can send it to you as a MC8 .mcd, but you'll have to solve your problem before viewing it. WRT size, TLs require big Vbs to get any LF gain so you can make it smaller, just understand that its F3 rises with decreasing Vb regardless of what the pipe's effective 1/4WL tuning is.
To get a relatively short pipe and LF gain requires mass loading it either by reverse tapering it or adding a vent, making it an ML-TL.
====
>i kinda wanted to run the 2 drivers with NO series resistance...
====
Well, you didn't say, plus you won't get a ~tonally balanced response unless stuck in a corner due to the driver's medium low Qts, so I chose a design that would perform well out in the room where bipoles ideally need to be.
Regardless, you can run this with less/no series R as required/desired, though you will need to use some other form of EQ to flatten the response, so 'pick your poison', so to speak.........
====
>also i was thinking of a straight pipe with the drivers at top and opening at bottom... similar to link i provided.
====
This is a straight pipe with bottom exit, just the driver is placed along the line where it will yield the smoothest response without having to stuff the life out of it.
If you look at the response of the linked example, it vividly shows what sort of response to expect with an end loaded design. Acceptable for a multi-way with woofer where you want max gain, but not an FR that can't decouple all the 'ripple' effects this much loading causes without stuffing the bejeesus out of it, defeating the reason for end loading it in the first place.
====
>can you provide me with the formulas you used? or did you use the mathcad worksheets?
Thanks!!!
====
You're welcome!
Now-a-days I get the basics (Vb, desired Qes, Fb, minimum vent diameter) using a simple box program (BoxPlot 3.07 in my case), then load these into a proprietary spreadsheet I developed to determine optimum length, driver position. I only use MathCad to get an idea of how much stuffing is required in some designs and/or vent length since no other program is as accurate once the shape morphs into a high aspect ratio. Being rather math challenged, I never could figure out how to calc the pipe's effect on it so almost always had to test the speaker to determine how much to shorten it. After seeing MJK's math, it's no wonder as to why!
For doing your own from scratch, I highly recommend you start with MJK's 'Classic' design tables to get a first approximation, or possibly final, design if you don't want to gamble on mine.
Anyway, if you insist on doing an end loaded design using a relatively short/low Vb pipe with no series R, then the F3 could wind up as high as 80-100 Hz if you don't mass load it. That's a lot of BW/efficiency to give up to keep the pipe small. Might as well just stick them in a sealed cab.
Indeed, for best transient response and near IB performance, make a sealed pipe of whatever length works best and the largest CSA you can tolerate (at least 2-3 x Sd depending on length), then EQ flat in-room.
GM
well...
hey thanks for the detailed reply.
anyhow... i guess what i really care about is being able to run them as bipoles and fullrange.
how would a ML-TL look for a bipole design?
or perhaps just a simple BR design?
maybe 2 "ports" one on front and one on back?
Thanks!
hey thanks for the detailed reply.
anyhow... i guess what i really care about is being able to run them as bipoles and fullrange.
how would a ML-TL look for a bipole design?
or perhaps just a simple BR design?
maybe 2 "ports" one on front and one on back?
Thanks!
I´d guess the simplest you could do is take Bob Brines FT1600 and just double the cross-section-area.
http://geocities.com/rbrines1/
You´ll find some ML-TL bipole examples here.
Examples with a FE-126/7 : http://www.diyaudio.com/forums/showthread.php?threadid=46934
I just build a BR for the FE-167E (15l recommended) and have to say it was not really worth. It´s nice but bass is a bit light (no problem in small rooms though and/or with a sub)and well, you need a stand anyway so could rather do a floorstander from the beginning on.
The BR would certainly be enough for a rear speaker in a HT-setup.(IMO)
greets
Oh wow, I'm just about to finish up a pair of these and now everyone is abandoninig these drivers in BR boxes!
Paul Joppa from Bottlehead gave up and now here are two more who are ready to throw in the towel. Oh well, it didn't cost me anything but time since I already had the drivers and some extra wood lying around. I'll finish them up anyway and give them quick listen and hopefully wlll come up with a "different" conclusion. 😀
Don´t get me wrong I really like ´em, it´s just that I don´t have a proper speaker stand and prefer listening to music without a sub.
(which it needs in my room)
Give yourself some time with the speaker. Depending on your room you might very much like it as it is. Try it with a sub if you have one. Quite a marvellous combo for it´s simplicity.
I guess also that the Fostex recommended is not the last word on BR.
If it´s supposed to be a BR, maybe it should be more like the Omega Grande 6 having more than 20l (>0.7 cu.ft).
FT1600 is on the list as well...
(which it needs in my room)
Give yourself some time with the speaker. Depending on your room you might very much like it as it is. Try it with a sub if you have one. Quite a marvellous combo for it´s simplicity.
I guess also that the Fostex recommended is not the last word on BR.
If it´s supposed to be a BR, maybe it should be more like the Omega Grande 6 having more than 20l (>0.7 cu.ft).
FT1600 is on the list as well...
They're not awful, it's just that they will definitely need low end support. If that's your goal then it will work well.
AudioGeek:
I think GMs numbers are good to go with (sure is a big box). You mentioned that you didn't like the idea of a series resistor but my experience has been that they can be quite beneficial. Another option would be to wire the drivers in series but in this particular situation I would go with parallel wiring and a series resistor first.
Personally, I would shy away from a box that size but there's always a compromise. You could build a bipole with the WR125ST but you'd have to give up efficiency. Not a lot, but some. Add a super tweeter and you're in business.
AudioGeek:
I think GMs numbers are good to go with (sure is a big box). You mentioned that you didn't like the idea of a series resistor but my experience has been that they can be quite beneficial. Another option would be to wire the drivers in series but in this particular situation I would go with parallel wiring and a series resistor first.
Personally, I would shy away from a box that size but there's always a compromise. You could build a bipole with the WR125ST but you'd have to give up efficiency. Not a lot, but some. Add a super tweeter and you're in business.
Greets!
You're welcome!
Well, normally I just run the numbers and post them because I know from experience they will perform as good as the accuracy of the specs used. BB, MJK, and numerous others are pleased with their theoretically somewhat less than optimum alignments so decided to try a different tact that lies between the two design 'extremes' since my 'optimum' design with no series R dictates a ~high Fb, consequently a fairly short pipe with the drivers so low that a considerable amount of space/materials is wasted in an extended false bottom to get them up to ear height.
Anyway, all things considered it looks somewhat better than I thought it would and is good down to a ~35 Hz Fb using MJK's? measured specs, so with some BSC and the rear driver rolled off to suit it should be a strong performer:
L = 55.09"
CSA = 100.388"^2 (suggest 12.75"w x 7.875"d)
driver down 19.93"
vent = 3" dia. x 0.75" near the bottom
suggest 0.25 lbs/ft^3 polyfil fluffed out from the top down ~44"
GM
GM:
is the posted design for a bipole then?
Also, can i run without a BSC? and run the rear full?
series or parallel drivers?
finally, if I wanted to add a series resistor... is there a design that can be built going to say 40 - 50Hz and both drivers run full range in a not so large enclosure?
Thanks!!!
p.s. i want to stick with the fe167e cause i have 2 allready.
pps i will run from tube amp with an output impedance of probably 3 ohms... (does this account for anything?)
is the posted design for a bipole then?
Also, can i run without a BSC? and run the rear full?
series or parallel drivers?
finally, if I wanted to add a series resistor... is there a design that can be built going to say 40 - 50Hz and both drivers run full range in a not so large enclosure?
Thanks!!!
p.s. i want to stick with the fe167e cause i have 2 allready.
pps i will run from tube amp with an output impedance of probably 3 ohms... (does this account for anything?)
GM said:
No other reason than having a big box. A CSA of 300+ sq. in. is bigger than I would want to have. The sound quality would probably be pretty good though.
Greets!
>is the posted design for a bipole then?
====
You're kidding, right?
====
>Also, can i run without a BSC? and run the rear full?
====
Sure! It wouldn't be something I'd want to listen to, but then that doesn't matter.
====
>series or parallel drivers?
====
Parallel, though with no BSC I'd try series just to see how it sounds. Usually though, drivers aren't matched close enough for it to work well without a parallel resistor to average them out, especially FR drivers.
====
> finally, if I wanted to add a series resistor... is there a design that can be built going to say 40 - 50Hz and both drivers run full range in a not so large enclosure?
====
Sure, just so long as you understand that there won't be much output below ~100 Hz. Read the other posts about the pitiful response of the Fostex recommended BR.
====
>Thanks!!!
p.s. i want to stick with the fe167e cause i have 2 allready.
====
Whatever.
====
>pps i will run from tube amp with an output impedance of probably 3 ohms... (does this account for anything?)
====
Ya know, you'd think that as many times as folks have tossed this out after a few rounds of posts I'd learn to ask this on the front end........ The amp's output impedance has the same effect on the driver's Qts as a series resistor, so for a maximally flat ML-TL, it needs to be somewhat bigger than my current design otherwise it will have boomy midbass (Altec style 'West Coast' sound) with a rapid roll off below ~60Hz. Sticking it in an even smaller cab will require it to be an aperiodic design.
Anyway, it's apparent now that performance is a distant second to size, so using the numbers in your first post: 30" high/80"^2 CSA, position the drivers down 12.25" from the top with a 2" dia. x 3" long vent down near/at the bottom, or out the bottom if you prefer, and stuff it till it sounds as smooth as it's going to get, just don't expect them to maintain any sense of tonal balance below ~300Hz without any BSC, and even with it the LF will fall like a stone below ~70 Hz if there's not much boundary/room gain to boost the lower octaves.
GM
>is the posted design for a bipole then?
====
You're kidding, right?
====
>Also, can i run without a BSC? and run the rear full?
====
Sure! It wouldn't be something I'd want to listen to, but then that doesn't matter.
====
>series or parallel drivers?
====
Parallel, though with no BSC I'd try series just to see how it sounds. Usually though, drivers aren't matched close enough for it to work well without a parallel resistor to average them out, especially FR drivers.
====
> finally, if I wanted to add a series resistor... is there a design that can be built going to say 40 - 50Hz and both drivers run full range in a not so large enclosure?
====
Sure, just so long as you understand that there won't be much output below ~100 Hz. Read the other posts about the pitiful response of the Fostex recommended BR.
====
>Thanks!!!
p.s. i want to stick with the fe167e cause i have 2 allready.
====
Whatever.
====
>pps i will run from tube amp with an output impedance of probably 3 ohms... (does this account for anything?)
====
Ya know, you'd think that as many times as folks have tossed this out after a few rounds of posts I'd learn to ask this on the front end........ The amp's output impedance has the same effect on the driver's Qts as a series resistor, so for a maximally flat ML-TL, it needs to be somewhat bigger than my current design otherwise it will have boomy midbass (Altec style 'West Coast' sound) with a rapid roll off below ~60Hz. Sticking it in an even smaller cab will require it to be an aperiodic design.
Anyway, it's apparent now that performance is a distant second to size, so using the numbers in your first post: 30" high/80"^2 CSA, position the drivers down 12.25" from the top with a 2" dia. x 3" long vent down near/at the bottom, or out the bottom if you prefer, and stuff it till it sounds as smooth as it's going to get, just don't expect them to maintain any sense of tonal balance below ~300Hz without any BSC, and even with it the LF will fall like a stone below ~70 Hz if there's not much boundary/room gain to boost the lower octaves.
GM
- Status
- Not open for further replies.