Folks,
I apologize if this question has been asked ad nauseam, but searching got me nowhere.
Much inspired by Bob Brines' work, I am toying with a straight MLTL for an Alpair 10P-A - the version with the paper cone rather than the metal cone. I'm playing with Martin J. King's MathCad scripts. I'm puzzled by a few things:
- How is the distance from the bottom of the TL to the port center determined?
- How does one decide on a resonance frequency for the TL and the port?
The f0 of the Alpair 10P-A is 42.4 Hz. Qts = 0.330. So I aim for a TL resonance frequency in the high 30ies.
By empirical design (i.e. trial & error) I arrived at an enclosure which provides a nice flat response and f3 of 33 Hz. I noticed that the resonance of the enclosure is very sensitive to the port diameter and length but not very sensitive to port placement (within reason, of course). The resonance frequency depends on the length of the TL, but the port dimensions seem to have the most impact.
It seems there are two knobs for tweaking the TL resonance: The length of the TL and the port. I think my question boils down to this: What's a systematic way for turning those knobs? Which tradeoffs are associated with each knob? I'd like to understand how this works rather than rely on trial and error.
Thanks,
~Tom
I apologize if this question has been asked ad nauseam, but searching got me nowhere.
Much inspired by Bob Brines' work, I am toying with a straight MLTL for an Alpair 10P-A - the version with the paper cone rather than the metal cone. I'm playing with Martin J. King's MathCad scripts. I'm puzzled by a few things:
- How is the distance from the bottom of the TL to the port center determined?
- How does one decide on a resonance frequency for the TL and the port?
The f0 of the Alpair 10P-A is 42.4 Hz. Qts = 0.330. So I aim for a TL resonance frequency in the high 30ies.
By empirical design (i.e. trial & error) I arrived at an enclosure which provides a nice flat response and f3 of 33 Hz. I noticed that the resonance of the enclosure is very sensitive to the port diameter and length but not very sensitive to port placement (within reason, of course). The resonance frequency depends on the length of the TL, but the port dimensions seem to have the most impact.
It seems there are two knobs for tweaking the TL resonance: The length of the TL and the port. I think my question boils down to this: What's a systematic way for turning those knobs? Which tradeoffs are associated with each knob? I'd like to understand how this works rather than rely on trial and error.
Thanks,
~Tom
I have also found that port placement is not that sensitive either when I tweak parameters in akabak for a mltl. Look for the resonance peaks and dips and see if moving the port position can smooth those out. In general, I find that placement of port at 2 to 4 diameters from the bottom makes little difference. If you will be using a slotted shelf type port, placement at the bottom is natural and the preferred location based on ease of construction. The placement of the driver, on the other hand, can be used to reduce the resonance peaks and dips with great effect and is very sensitive to small movements of even a half inch. However, driver placement between 0.2 and 0.33x from the top usually provides a reasonable place to start before tweaking to optimize. In fact, you may find that a quite reasonable and well-performing mltl can be designed based on rules of thumb alone without any tweaking using a simulation as I did - and describe in the Accidental MLTL technique thread: http://www.diyaudio.com/forums/full-range/231951-accidental-mltl-technique.html
Certainly use MJK since you have it, but try the AMLTL as a starting point and see how close to a solution it is.
Can you please post your sim results for the A10p or at least the dimensions of the design?
Good luck.
X
Certainly use MJK since you have it, but try the AMLTL as a starting point and see how close to a solution it is.
Can you please post your sim results for the A10p or at least the dimensions of the design?
Good luck.
X
Hi,
The AMLTL is a long thread but makes a lot of sense. I'd add
depending on the driver you can play around with the vented
box alignment a little before you convert it to the MLTL.
e.g. whilst WINISD suggest 16L tuned to 52Hz,
20L tuned to 42Hz looks quite good to me.
Port placement is not critical at the end of
the box, driver placement is more critical.
rgds, sreten.
The AMLTL is a long thread but makes a lot of sense. I'd add
depending on the driver you can play around with the vented
box alignment a little before you convert it to the MLTL.
e.g. whilst WINISD suggest 16L tuned to 52Hz,
20L tuned to 42Hz looks quite good to me.
Port placement is not critical at the end of
the box, driver placement is more critical.
rgds, sreten.
I followed Bob Brines' design procedure. If I recall correctly, I ended up with about 135~140 cm for the internal height. The cross sectional area was 4*Sd. I started with the driver at 0.2*L and tweaked the position slightly to remove the slight SPL wobble caused by the 5th harmonic.
I forget the exact port dimensions, but I seem to recall using either 2.5" or 3" diameter; 3" length.
~Tom
I forget the exact port dimensions, but I seem to recall using either 2.5" or 3" diameter; 3" length.
~Tom
Mr. Brines explains it well: Nature of the Quarter Wave Resonator.
On that page the tradeoffs between TL area, port area, and driver placement are explained. No word on port placement, though. I'm getting the feeling it isn't overly critical.
Thanks Bob for making that information available.
~Tom
On that page the tradeoffs between TL area, port area, and driver placement are explained. No word on port placement, though. I'm getting the feeling it isn't overly critical.
Thanks Bob for making that information available.
~Tom
Port size and length are both significant. You want to use the absolutely smallest port diameter that will support the SPL required. Determine this by looking at the port velocity. Port length is important because , as mentioned above, the are two degrees of freedom here. You can find an infinite number of pipe lengths and port lengths that will give the same Fc.
Driver and port position are both important and rather touchy. After setting the stuffing to zero, move the driver to get the best suppression of the first overtone.
Now the port. There are three rather large harmonics between 700-100Hz. Move the port up from the bottom until it suppresses the middle harmonic. This produces the best FR plot. Now stuff the pipe until you get the degree of smoothness you desire.
Bob
Driver and port position are both important and rather touchy. After setting the stuffing to zero, move the driver to get the best suppression of the first overtone.
Now the port. There are three rather large harmonics between 700-100Hz. Move the port up from the bottom until it suppresses the middle harmonic. This produces the best FR plot. Now stuff the pipe until you get the degree of smoothness you desire.
Bob
Hi,
Though that page mentions Mass Loading by tapered lines,
it is not the same as a MLTL, and they are not discussed.
Hence nothing on port placement and nothing on if the
the port actually suppresses the higher modes output.
rgds, sreten.
Last edited:
Smaller port diameter --> higher velocity. From what I understand, higher velocity creates turbulence in the port, which adds distortion. I don't know if this is science or internet hearsay, though. What's a good tradeoff between "as small diameter as possible" and port velocity? In other words, what's the speed limit or the limit on the minimum port diameter?
Awesome. Thank you!
Another question: How do you model the port placement in Martin's "Sections" sheet? Do you just assume that the port is placed at the end of the TL? After all, the "Sections" sheet is intended for use with a TL that's open in the end. A TL with a port located 5" up from the bottom, is closed in the end, but open 5" up. I don't see a logical way to include this in the "Sections" sheet, yet, on your website you mention that you use it.
~Tom
Last edited:
Isn't MLTL = Mass Loaded Transmission Line?
I believe the port mass loads the TL, so a straight pipe with a hole in it would qualify as a MLTL. No?
~Tom
Note the comments about port sizes and bandpass boxes near the bottom of this page:
tech note 14
~Tom
tech note 14
~Tom
MLTL Experts. ?? Is there a BIGGER design picture??
0) BUILD AN ADJUSTABLE PORT. TL stuffing is an art. Your absorption materials may be different from the model. Your stuffing technique may be different from the model. Your room will have more effect on bass SPL than the MLTL port placement. Think about how you can built an adjustable port. From my experience, a 0.5" change in port length is audible.
1) A large radius on all port edges is very inportant. External and Internal.
2) SIMULATE A PLAN to adjust the port shape/volume for room tuning. Floor gain? Against wall port? Corner? Away from all walls?
3) Most smart stuffing solutions absorb >200Hz resonances.
DOW thick 125hz 250hz 500hz 1000hz 2000hz 4000hz
703, 3" 0.53 1.19 1.21 1.08 1.01 1.04
Few MLTL solutions include room effect port compensation plans.
0) BUILD AN ADJUSTABLE PORT. TL stuffing is an art. Your absorption materials may be different from the model. Your stuffing technique may be different from the model. Your room will have more effect on bass SPL than the MLTL port placement. Think about how you can built an adjustable port. From my experience, a 0.5" change in port length is audible.
1) A large radius on all port edges is very inportant. External and Internal.
2) SIMULATE A PLAN to adjust the port shape/volume for room tuning. Floor gain? Against wall port? Corner? Away from all walls?
3) Most smart stuffing solutions absorb >200Hz resonances.
DOW thick 125hz 250hz 500hz 1000hz 2000hz 4000hz
703, 3" 0.53 1.19 1.21 1.08 1.01 1.04
Few MLTL solutions include room effect port compensation plans.
Attachments
Just ask how the whole process is determined and done. The actual air resistance or acoustic impedance determines that the speaker behaves in
a certain way and that the acoustic load determined by some restrictions
in the bacwave path makes behave the cone itself in another way ( change of some parameters , like in a mass load the column of air in motion is coupled to the movement of the cone till the fluidity of air and velocity of vibrations -frequencies- don't allow for that), and what you see at the end of a TL is just ( hopefully)
the end of processing thru the 1/4 wave rule. Also a BR is a product of internal and external pressure.
Last edited:
Hi,
Yes IMO a MLTL is a pipe loaded by a distinct port at the end rather than a Tapered TL.
That is a port with a much smaller CSA (with a length) into a much bigger pipe CSA.
However at this point hard facts get distinctly murky, which is unusual for MJK.
e.g. if you build this : https://sites.google.com/site/undefinition/diy/amiga
Stuff the top half and move the port near the bottom you've got a MLTL.
rgds, sreten.
A port shouldn't have any peaks 700Hz to 1KHz other than its own length resonance.
Last edited:
Right. Although I think it was originally GM who first coined or at least popularised the term MLTL for an untapered QW pipe mass-loaded by a vent. A reverse taper pipe, i.e. one narrowing toward the terminus (which oddly enough used to be called a TQWT, though it is the opposite of the conical horns usually described thus these days -I think GM & I are the only people who hold to the original definition) is inherently mass loaded given its geometry.
Vent placement in an MLTL can & does have an effect on the response & particularly the upper BW of the vent output, though as might be expected somewhat less than driver location has unless large changes are made. Generally they aren't, and it's rare to see an MLTL (or at least a box designed as such) with the vent offset from the nominal end of the main pipe more than, say, 5in. Unless the box is heavily damped then the differences are usually related to the acoustic location of the main pipe's harmonic modes. Me, I usually have the vent in an MLTL at the end of the main pipe or thereabouts since it's ultimately the most efficient load over the widest BW for a given tuning &c. That tends to mean you might need a little more damping; so be it. Depends on your design priorities. Equally, I tend to prefer slightly oversized (in terms of diameter) vents, providing (providing) length doesn't exceed, say, 6in. The latter is somewhat arbitrary, but it should keep the vent's own harmonic modes fairly high (> 1KHz) where they're less likely to cause problems. Works for me anyway. YMMV as always of course.
Vent placement in an MLTL can & does have an effect on the response & particularly the upper BW of the vent output, though as might be expected somewhat less than driver location has unless large changes are made. Generally they aren't, and it's rare to see an MLTL (or at least a box designed as such) with the vent offset from the nominal end of the main pipe more than, say, 5in. Unless the box is heavily damped then the differences are usually related to the acoustic location of the main pipe's harmonic modes. Me, I usually have the vent in an MLTL at the end of the main pipe or thereabouts since it's ultimately the most efficient load over the widest BW for a given tuning &c. That tends to mean you might need a little more damping; so be it. Depends on your design priorities. Equally, I tend to prefer slightly oversized (in terms of diameter) vents, providing (providing) length doesn't exceed, say, 6in. The latter is somewhat arbitrary, but it should keep the vent's own harmonic modes fairly high (> 1KHz) where they're less likely to cause problems. Works for me anyway. YMMV as always of course.
The magic number for port velocity is 10m/sec give or take depending on your reference. Figure the SPL you will actually use and then figure your port diameter from that.
Don't use "sections" for MLTL's. Use "ported box". "Ported box" will model sealed boxes, BR's and MLTL's and is much simpler than "sections". Use "sections" only if you are doing a tapered TL, "Voigt" pipe,or back horn.
Bob
10 m/s or about 0.03*(speed of sound). I like your philosophy of designing for typical listening levels rather than rock concert levels.
Ah. That makes more sense. Thank you very much for your help.
~Tom
Having discussed this with Scott before, i dug into the lit, and the term TQWT (as commonly used today) was esssentially coined in Voigt's patent on the Voigt pipe.
dave