| cadaverdog |
I am in the process of designing a quarter wavelength loudspeaker based around the Fostex FE-166E.
The peculiarity of this design is that the low frequency cutoff will be at about 100Hz, so I'm not really concerned with getting the lows "right."
My questions are:
Can truncating the low frequency response of this driver through filtering- yield benefits in the mid/high frequency response??
If a transmission-line theory is based upon exploiting a drivers resonant peak, then what happens if you filter-out the resonant frequency??
Please correct me if I'm wrong:
Because music does not have every frequency occurring at the same time (unless you like listening to white noise) I assume that this design will perform exactly the same as if the missing frequencies were present. |
|
|
| rjbond3rd |
| Hi cadaverdog, your project sounds interesting. May I ask what made you choose QW as the design? Just asking out of curiosity! |
|
|
| cadaverdog |
I am doing an engineering senior design project- and one of the specifications is that it be an original idea.
What I think makes this original is that this will be the first (that I know of) double transmission-line loudspeaker; Meaning, there will be a t-line mid/high section and a t-line low freq. section within the same enclosure.
I cannot change the specification from quarter-wave to something else as that was the entire point of last semester: Defining the goal of the design.
Anyway, there will be built into the cabinet an integrated amplifier using a pair of Hypex UCD-180 amplifiers, a phase control, level controls (for the individual sections)- and a master volume control. All you have to do is plug in a CD player and go.
The second reason I chose quarter-wave is because I prefer the sound of this enclosure to any other type.
I was requesting suggestions because I figured that it might save me a little research.
I have a lot of expensive equipment that I can access from the universities' inventory if I need it, so I will build and test if I have to. |
|
|
| Nihilist |
You plan on tuning the TL to a frequency below 100hz, and then run a highpass filter so the driver only sees from 100hz and up ?
I don't see what the point of this is, as the purpose of the TL is for bass response/tuning.
Am I missing something ?
.................................Blake |
|
|
| Scottmoose |
| I was wondering that myself. |
|
|
| GM |
Well, not sure how original it is per se since what you propose (if I understand you correctly) for the mid/high will be an aperiodic alignment if done properly, so the 1/4 WL part is moot and its analogue can be found in some '50s era two way systems. I imagine it's been mated to a TL (or more likely a TQWT) somewhere along the line also due to their popularity, though no consumer product past or present to cite comes to mind at the moment.
Anyway, just to make sure we're 'on the same page', I'm interpreting your goal is to use a TL as an acoustic series notch filter to ~completely damp the mid/HF driver's resonant peak so that no electrical high pass is required, or that if one is used, that a textbook solution will work as predicted.
WRT to 1/4 WL theory, what you want is a max flat impedance alignment which the casual DIYer can now easily design using Rick Shultz's (aka 'Exolinear') Alpha TL math. Unfortunately, I know of no program that will accurately sim such a short pipe due to its typically high stuffing ratio, so getting it right and/or 'forcing' a 166E to fit the app may prove a formidable undertaking, but have no doubt you'll be pleased with the results once dialed in.
Regardless, anytime you relieve a wide BW driver of high excursion it's sonically beneficial.
GM |
|
|
| cadaverdog |
Please correct my thinking if I'm wrong:
I thought that the purpose of a t-line is to eliminate the backwave of the driver for all frequencies except the tuning frequency.
Isn't a t-line is essentially an infinite baffle at all frequencies except the tuning frequency (which is why resonance is chosen being that there's an excees of energy at this point?)
Even if used in a mid/high frequency application, isn't eliminating the driver's backwave desirable; even if the resonant frequency is never seen??
If not a t-line, what would be a suitable enclosure for the mid/high section??
Thanks for your time |
|
|
| cadaverdog |
| I've noticed that there is somewhat of a midrange "dip" in the frequency response of the FE-166E- so could I simply tune the driver to a frequency that will augment the midrange?? |
|
|
| Scottmoose |
| quote: | Originally posted by cadaverdog
Please correct my thinking if I'm wrong:
I thought that the purpose of a t-line is to eliminate the backwave of the driver for all frequencies except the tuning frequency.
Isn't a t-line is essentially an infinite baffle at all frequencies except the tuning frequency (which is why resonance is chosen being that there's an excees of energy at this point?)
Even if used in a mid/high frequency application, isn't eliminating the driver's backwave desirable; even if the resonant frequency is never seen??
If not a t-line, what would be a suitable enclosure for the mid/high section??
Thanks for your time |
Looking at your queries in turn:
1\ not really. The problem is that TL is used by a host of people to mean completely different things -sometimes even exact opposites. By strict definition, a pure TL is a max-flat impedance design of the type outlined by GM above. There are other types of line which seem to have come under the catch-all definition -hybrid semi-resonant lines (like Bailey's, Augspurger's etc), and very lightly damped, highly resonant QW lines. Depends what's wanted. In the latter two catagories, the cabinet is used to support the LF regions to varying extents, only slightly in the former catagory, and to a much larger extent in the latter.
2\ Yes. Although heavily damping other types of cab. will probably get you ~'close enough' to render it of questionable utility.
3\ Heavily damped aperiodic or sealed should do the job fine.
It's worth observing though that B&W did a sealed midrange 1/2 wavelength line damping the full BW of said mid. driver in their Nautilus cabinet, although IIRC it was closer to Olney's 1937 Acoustic Labyrinth (or PMC's current range of commercial cabinets, for that matter) with damped walls rather than the fibreous stuffing Bailey introduced with the TL moniker in 1965.
BTW -the 167 might be a better choice for you than the 166.
| quote: | Originally posted by cadaverdog
I've noticed that there is somewhat of a midrange "dip" in the frequency response of the FE-166E- so could I simply tune the driver to a frequency that will augment the midrange?? |
Not really -it'd be the hell of a small enclosure, and you really don't want to hear midrange informartion coming from two different places. Especially if they are physically widely separated. You'll get destructive interference (lobing / phase issues), plus some other problems.
|
|
|
| cadaverdog |
When you say "aperiodic", do you mean a line that's not necessarily 1/4-wave??
If I were to do a sealed midrange, is there any specification for the volume of the enclosure??
Thanks |
|
|
| cadaverdog |
It was a good point that you made about phasing issues in the midrange. I don't want to have to deal with that as well. It'll probably be better to either use a heavily damped line for the midrange- or simply seal the back.
If you would be so gracious as to indulge me one last time, I am adding a phase control to the LF section in order to avoid having to "align" the drivers on the front baffle; In your opinion, is my thinking flawed here?? |
|
|
| Scottmoose |
| quote: | Originally posted by cadaverdog
When you say "aperiodic", do you mean a line that's not necessarily 1/4-wave??
If I were to do a sealed midrange, is there any specification for the volume of the enclosure?? |
Aperiodic is another of those terms the definition of which tends to vary depending on who you ask. Generally, it refers to a leaky sealed box, rather than a 'line.' BTW, be careful applying the QW term to TLs. Strictly, a TL should be either a sealed or open line so heavily damped as to completely suppress the entire BW of the driver, so the question of QW performance is in reality somewhat moot. Only when you remove some of the damping from the open-ended box, to form a semi-resonant or highly resonant line, is QW resonant behaviour particularly relevant.
WRT sealed boxes, yes, of course. It depends on exactly what you want to do, as always. Therefore, may I suggest that before you start anything, a little research would reap dividends? The Loudspeaker Design Cookbook will give you a good start.
| quote: | Originally posted by cadaverdog
I am adding a phase control to the LF section in order to avoid having to "align" the drivers on the front baffle; In your opinion, is my thinking flawed here?? |
Not necessarily flawed, although it's usually better to solve something mechanically IMO by careful design than try to fudge it with a shortcut. Depending on the drivers, the layout & the XO frequencies you may not need to get a perfect alignment however. Remember, good enough is good enough. If you really want to use the 166 though, the XO point will need to be relatively high as Fhm on that driver is about 475Hz, based on the factory specs. (which on current evidence will be the greatest work of fiction since Great Expectations). |
|
|
| GM |
| quote: | Originally posted by Scottmoose
......the 166.......(which on current evidence...........). |
What's its actual specs?
TIA,
GM |
|
|
| GM |
| quote: | Originally posted by cadaverdog
When you say "aperiodic", do you mean a line that's not necessarily 1/4-wave?? |
Right. Aperiodic means 'without period', ergo a max flat impedance TL alignment by definition is Aperiodic.
GM |
|
|
| cadaverdog |
The LF section will use a pair of 8" subwoofers in push-pull configuration in order to eliminate cabinet vibrations. I am hoping that the combination of a high-power amplifier and a highly-stuffed line will get me into the lower registers.
There are probably much more elegant solutions as you have so graciously let me know of, but I only have about 5 months to bring this project from concept to reality. I just wanted to make sure that what I was planning is even remotely practical.
Therefore, as a result of our discussion, I am going to modify my design such that I don't go out of my way to build something that one of you referred to as "of questionable utility."
Chances are, if you noticed a design flaw this early-on, then the board of professors who'll be evaluating this beast will probably see the same thing.
Thank you all for your input. |
|
|
| GM |
You're welcome!
If by 'push-pull' you mean a bipolar layout, then yes, it will minimize cab vibrations.
If you use a low Fs driver and tune it to damp it, then you will get a near IB frequency response to below Fs with a flatter acoustic phase response (a good thing in theory), so the driver's Qts will define the TL's roll off slope (Qtp), i.e. for a tonally balanced response to near Fs in half space requires a 0.707 Qts and 0.5 Qts for a critically damped one (1st order roll off).
Anyway, good luck with your project! Please save the link to this thread and post what you built, preferably with pictures, and how it was graded.
GM |
|
|
| cadaverdog |
| Will do, thanks. |
|
|
| Nihilist |
"If by 'push-pull' you mean a bipolar layout, then yes, it will minimize cab vibrations.
If you use a low Fs driver and tune it to damp it, then you will get a near IB frequency response to below Fs with a flatter acoustic phase response (a good thing in theory), so the driver's Qts will define the TL's roll off slope (Qtp), i.e. for a tonally balanced response to near Fs in half space requires a 0.707 Qts and 0.5 Qts for a critically damped one (1st order roll off)."
Gm,
Is the tuning you describe applicable to the Bipolar design only, or was the Bipolar part only meant for the minimisation of cabinet vibration ?
If the tuning you describe is for the TL , regardless of mono or bipole action as I suspect, then I have another question for you. Let's use a simple example of a driver with a Qts.7 and a Fs of 25hz (example only, no particular driver in mind here) .
We would need to tune the TL to 25hz to achieve the response shape you mention. Would this be a function of the length of the line, the amount of stuffing, or both ?
Could one use an undamped line of full length, or a shorter stuffed line to achieve similar bass response ? I know that the sound may be different , but I would imagine that would have to do with cabinet artifacts and/or differences in reflections between a damped/undamped line.
Thanks.................................Blake |
|
|
| cadaverdog |
...............that I titled the project "Quarter Wavelength Loudspeaker"- being that this loose definition can give me some creative license to build something a bit more practical.
Whatever form this project takes, I have to achieve critical damping (which is what I told the review board to expect.)
I am currently using Mr. Augspurgers' "TLWRX" software to model the various components described here. If you have any suggestions for better software, let me know.
The Gantt chart for the project looks something like this:
Design and build the integrated amplifier circuit (already done)
Model the two lines in software in order to get "ballpark" parameters.
Build a test setup using cardboard tubes of various lengths and various stuffing.
Settle on final volumes/stuffing for both HF and LF enclosures.
Test the synergistic whole using the amplifier circuit described above.
Build the cabinet and do final assembly.
Present the finished project.
My logic for building the enclosure is to go from cardboard (round) tubes to a wooden (square) tube enclosure. In my thinking, this is fine as long as the (cross-sectional) area and the length is the same. Maybe I can play with the stuffing in order to fine tune the system.
I will post the entire project on this board- so that you can all have a part in it if you'd like.
I originally thought that I've quagmired myself in an impossible project, but thanks to you guys- I'm OK with the prospect now. I don't pretend to know everything about these mysterious enclosures (as I am just learning about them), but I am an aspiring engineer and a music lover- and I'm sure that I know what sounds good.
Thanks guys, I'll keep you posted |
|
|
| GM |
| quote: | Originally posted by Nihilist
Gm,
Is the tuning you describe applicable to the Bipolar design only..........?
We would need to tune the TL to 25hz to achieve the response shape you mention. Would this be a function of the length of the line, the amount of stuffing, or both ?
Could one use an undamped line of full length, or a shorter stuffed line to achieve similar bass response ? |
No, you can bipole any alignment.
It will be a function of the line length and net Vb plus the amount/type of stuffing used.
Well, you won't get a similar response AFAIK no matter how you juggle the numbers due to the damping effect of the stuffing having a much different friction coefficient than air through a duct. The other problem is that its third harmonic dip will be at different points with it being a higher Q one higher in frequency in an un-stuffed line, so for sub duty where the un-stuffed TL's greater gain is normally desirable, its Fp is normally a function of where its third harmonic dip is since normally you want to use it to increase the XO's effective slope order.
That said, if size and Qp is more important than gain BW in a sub, then the TL's CSA can be dramatically reduced and while it will have to be lengthened for a given third harmonic dip frequency, increasing net Vb, the net reduction can still be considerable. For a theoretical example, a FE166E's Fp = Fs stuffed TL is ~2.883 ft^3 Vs the un-stuffed's minuscule ~0.8657 ft^3, a ~3.33x reduction. While still performance wise considerably different to the stuffed TL, it will be much closer frequency response wise than the full size high gain one.
GM |
|
|
| cadaverdog |
I was brainstorming yesterday at work- and I found some large paper cylinders (similar to concrete forms) that I thought would be perfect for this project. They are 12" ID and 13.25" OD. They are 68" long.
I could have the woofers at the base of the tubes (the length of which will be tuned for the woofers themselves)- and I could have the FE-166E's about half-way up (or ear level); The FE-166E's would use the tubes as a "leaky-sealed or aperiodic" enclosure. The benefits would be a simple enclosure with a round front baffle (minimal diffraction.)
The drivers on the front baffle would be spaced so that they are not further than one wavelength of the crossover frequency from each other- and the enclosure itself would be tilted (possible adjustable) for optimal listening at ear level.
Other than looking like a pair of large cigarretes in my dorm room, what do you think?? |
|
|
| Nihilist |
Sounds good. The paper tubes may honk , and the 68" tube will be tuned to about 50hz sans stuffing.
Would be very simple , seems like it should work well.
I've built speaker using 12" id PVC , using 8" Morel woofers in a sealed/stuffed configuration. Sounded pretty darn good. One of the best sounding speakers I've built.
.........................Blake |
|
|
| cadaverdog |
I think I'm excited.
I am designing the PCB for the power supply right now- and I should be getting back the boards in two weeks.
There will be a "box" mounted on the back of this enclosure in order to house the amplifiers, crossover and power supply. Being class-D, the amplifiers wil be minimally sized.
I've decided not to complicate the circuit anymore than it already is- so I won't be using a phase control on this project.
I'm thinking that if I make the enclosure "tilt" adjustable- then I could avoid using electronic phase correction.
I have to have both the enclosure and amplifier design completed by late July- so I'll keep you posted.
Thanks |
|
|
| cadaverdog |
I have finalized my design- and I now have a few more questions:
The final design will still use a t-line for the subs in push-pull configuration. This, once again, is for <100Hz.
As far as the Fostex FE-166E is concerned (used for >100Hz), I have decided to go with a sealed box tuned for critical damping. The issue I found with this was the fact that even for a QB=.5, the enclosure was really small. I think I unwittingly "solved" this by the fact that the t-line tube will run right through the center of the Fostex enclosure. All I have to do is subtract the area taken by the tube and adjust the height accordingly. In my mind, this has the added benefit of better prevention of standing waves by having the backwave strike something cylindrical rather than a flat-backed enclosure. I can add some light damping to "taste".
Anyway, here are the questions:
1.) Do I really need to mount the subs front to back, or can I simply have both cones facing outward. Is that "in is not equal to out" stuff just an audiophile myth??
2.) I have placed the drivers on the front baffle approx 32" apart which is just 1" below the quarter wavelength value of 100 Hz. Is this too far apart- and will I get lobing?
3.) I wanted to tilt the front baffle to align the acoustic centers of the drivers, but the more I think about it- I don't think the extra work will pay-off. I don't always stay planted to one spot when I listen, so I'm not sure the extra "gold plating" will make a difference. Am I wrong here??
Thanks for all of your help...................... |
|
|
| Nihilist |
"The final design will still use a t-line for the subs in push-pull configuration. " . . . . . . . . .
. . . . . . . . . " 1.) Do I really need to mount the subs front to back, or can I simply have both cones facing outward. Is that "in is not equal to out" stuff just an audiophile myth??"
I guess the question is, are you running the subs in "Push-Pull" , or in a compound loaded (Isobaric) "Push-Pull" . If you are meaning to say compound loaded (Isobaric) , then you do not HAVE to run the subs front to back, although there is a claimed benefit, which I will not dispute.
If you are only running the subs "Push-Pull" , then yes , you do have to run one sub facing "backwards" to accomplish this.
A speakers motor is NOT symetrical. So it's fore and aft stroke will be influenced differently by the motor. It seems quite reasonable that it will make a difference. I can't say I've seen any measurements to prove it, but some of the most musical subs use a "Push-Pull" format.
"2.) I have placed the drivers on the front baffle approx 32" apart which is just 1" below the quarter wavelength value of 100 Hz. Is this too far apart- and will I get lobing?"
You are pushing it , at least mathematically. In real world it might be ok , but I'd rather have some safety margin , if it were my project.
"3.) I wanted to tilt the front baffle to align the acoustic centers of the drivers, but the more I think about it- I don't think the extra work will pay-off. I don't always stay planted to one spot when I listen, so I'm not sure the extra "gold plating" will make a difference. Am I wrong here??"
Well, if it isn't too much work, do it. The drivers will still be time aligned , regardless of your position. It will just be more noticeable when you are in the "sweet spot" .
http://img507.imageshack.us/img507/1451/pushpullui9.png
...............................Blake |
|
|
| cadaverdog |
Thanks for your reply; I was beginning to wonder if anybody read this post anymore.
I'll put the drivers a bit closer, turn one of the subs around, and angle the front baffle.
My instincts told me that these things were right to do in the first place, but it's in my nature to ask questions when I'm in the company of so many knowledgable people.
I'm going to start on the cabinets on Monday night.
I'll post pictures- and the technical paper once this ball gets rolling a bit.
Thanks for your help. |
|
|
| Nihilist |
You're welcome. Good luck with the build. Keep us posted how it turns out !
............................Blake |
|
|
| digitalstore |
| quote: | Originally posted by cadaverdog
Thanks for your reply; I was beginning to wonder if anybody read this post anymore.
I'll put the drivers a bit closer, turn one of the subs around, and angle the front baffle.
My instincts told me that these things were right to do in the first place, but it's in my nature to ask questions when I'm in the company of so many knowledgable people.
I'm going to start on the cabinets on Monday night.
I'll post pictures- and the technical paper once this ball gets rolling a bit.
Thanks for your help. |
Don't forget to place it here;)
I am eagerly waiting for it. :bigeyes: |
|
|
|
