Speaker boxes and port resonance

Discussion in 'Archived Threads 2001-2004' started by Ted_Polzin, Jan 2, 2003.

  1. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Good morning. I have a few questions about some speakers I want to build.
    I have been using Unibox and have modeled the size of the speaker I want to build. I seems to like 45.7l the best. My first question is how much internal bracing should I use? Would 3 horizontal braces be enough? Should I also use vertical bracing? How about stuffing? What kind of material should I use in towers? Also how much internal volume is used up by 2 6.5" drivers, the tweeter and the bracing, 10%?
    My next series of questions [​IMG] is about port resonance. When modeling the speaker to keep the port speed below 26m/s at f3 I need to have the port big enough that there is a modeled port resonance at about 1500 Hz which is 1000hz below where I want to have my crossover. Which is worse a higher port speed (as high as 65m/s at f3) or a port resonance below the crossover range?
    Thanks for your help
    Ted
     
  2. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    >I have been using Unibox and have modeled the size of the speaker I want to build.
    ====
    Based on sims listed on other threads, I don't have any faith in the accuracy of Unibox or WINISD's port predictions.
    ====
    > I seems to like 45.7l the best.
    ====
    Don't know, but this may be suspect also.
    ====
    > My first question is how much internal bracing should I use? Would 3 horizontal braces be enough? Should I also use vertical bracing?
    ====
    The amount depends on the cab's materials/size. If any bracing is required then all six sides should be tied together so it can't 'breathe'. Ideally, the driver(s) should be mass loaded/braced to the cab. Adire's ventilated 'x' frame works well and makes it easier to add the driver bracing. I prefer to use up scrap no-void plywood strips to effectively do the same thing.
    ====
    > How about stuffing? What kind of material should I use in towers?
    ====
    I use R-19 fiberglass when I stuff, but how tall are the towers? Once you move away from a golden ratio cab, the length begins to impart pipe resonances to the response, increasing with increasing length.
    Stuffing a vented cab to increase its apparent size isn't a good plan since it normally overdamps the speaker/lowering vent output too much so you want to use the absolute minimum that attenuates any mids reflections and/or standing waves. Normally lining the end opposite the vent, one wall, and back is all you should need if you built a stiff cab. Here's a little overview: http://www.geocities.com/kreskovs/Box-1.html
    ====
    > Also how much internal volume is used up by 2 6.5" drivers, the tweeter and the bracing, 10%?
    ====
    Depends on the driver/pipe size. The vent volume can be easily calculated, the driver's close enough by placing it in a waterproof bag and measuring how much water it displaces.
    ====
    >My next series of questions is about port resonance. When modeling the speaker to keep the port speed below 26m/s at f3 I need to have the port big enough that there is a modeled port resonance at about 1500 Hz which is 1000hz below where I want to have my crossover. Which is worse a higher port speed (as high as 65m/s at f3) or a port resonance below the crossover range?
    ====
    Hmm, since ~17m/s is considered the upper limit unflared and ~34m/s flared, my concerns WRT accuracy, and with no other info, I'd say you may need to re-think your design. That said, there are practical limits to pipe size to keep its resonances from audibly comb filtering the response. Here's a popular rule-of-thumb from DW:
    MaxLength = 13560/(20*Fb)
    Where
    MaxLength is in inches
    Fb is the tuning frequency in Hz
    For area, the following equation would work:
    MaxArea = (H*W)/10
    Where
    MaxArea is in square inches
    H is the height of the panel containing the vent, in inches
    W is the width of the panel containing the vent, in inches
    GM
     
  3. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    First thank you for your response. I have a few more questions if you don't mind [​IMG]
    >I have been using Unibox and have modeled the size of the speaker I want to build.
    ====
    >>Based on sims listed on other threads, I don't have any faith in the accuracy of Unibox or WINISD's port predictions.
    I agree. I'm not sure what else to use though. What would be the best way to determine the port size then?
    > I seems to like 45.7l the best.
    ====
    >>Don't know, but this may be suspect also.
    I have verified this from several sources. Madisound who sold me (actually my dad who bought me the drivers for Christmas) the drivers has also said this [​IMG]
    >> Ideally, the driver(s) should be mass loaded/braced to the cab. Adire's ventilated 'x' frame works well and makes it easier to add the driver bracing. I prefer to use up scrap no-void plywood strips to effectively do the same thing.
    How would I load/brace the drivers themselves to the cab? So you are saying I should have horizontal braces then vertical brace them to the top and bottom?
    >> I use R-19 fiberglass when I stuff, but how tall are the towers? Once you move away from a golden ratio cab, the length begins to impart pipe resonances to the response, increasing with increasing length.
    >>Stuffing a vented cab to increase its apparent size isn't a good plan since it normally overdamps the speaker/lowering vent output too much so you want to use the absolute minimum that attenuates any mids reflections and/or standing waves. Normally lining the end opposite the vent, one wall, and back is all you should need if you built a stiff cab.
    yeah I do not plan on using stuffing to increase the apparent size. I want to use it to dampen any standing waves etc. Would fiberglass work good here? How about stuff like Dacron? The size I want to make it is 45.5 inches tall x 9 inches wide x 11.5 inches deep, out of MDF. .75 inches for all but front baffle, front baffle 1.5 inches. The drivers I have are from this page http://www.madisound.com/audaxsale.html
    AW025S1 1" SHIELDED PURE TITANIUM DOME TWEETER (for the tweeter)
    HT170F0 6-1/2" FIBERGLASS CONE WOOFER (for the mid/woofer)
    Not the world best speakers but should do OK for a first set of speakers. I have built a pair of stereo amps from ESP (http://sound.westhost.com/project3a.htm) and a active crossover from him. I plan on actively biamping the speakers.
    I do have a stock crossover design from Madisound to work from for crossover points etc. I am just concerned about the boxes and where I should go. I am buying the MDF tonight and will probably start cutting it tomorrow so any help would be awesome!!!!
    Thanks for all your time.
    Ted
     
  4. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    >First thank you for your response. I have a few more questions if you don't mind
    ====
    You're welcome, why am I not surprised? [​IMG]
    ====
    >I have been using Unibox and have modeled the size of the speaker I want to build.
    >>Based on sims listed on other threads, I don't have any faith in the accuracy of Unibox or WINISD's port predictions.
    >I agree. I'm not sure what else to use though. What would be the best way to determine the port size then?
    ====
    Unless there's a freebie program out there that's at least as good that I'm not aware of, then $25 for BoxPlot 3.0 is a good value if speaker building is a sometime hobby, or if you're serious, then I recommend you step up to Lspcad.
    If you don't mind not being able to save your designs, then download the Mathcad demo and the various spreadsheets from http://www.quarter-wave.com/
    ====
    > I seems to like 45.7l the best.
    >>Don't know, but this may be suspect also.
    >I have verified this from several sources. Madisound who sold me (actually my dad who bought me the drivers for Christmas) the drivers has also said this
    ====
    Ok, finally found some more specs than what's at Madisound, and while they're different from what I have that's preloaded in BP, they both model ~the same at 45.71L/46Hz Fb. A 4" dia vent 6.81" long looks good assuming a golden ratio cab.
    ====
    >> Ideally, the driver(s) should be mass loaded/braced to the cab. Adire's ventilated 'x' frame works well and makes it easier to add the driver bracing. I prefer to use up scrap no-void plywood strips to effectively do the same thing.
    How would I load/brace the drivers themselves to the cab? So you are saying I should have horizontal braces then vertical brace them to the top and bottom?
    ====
    Correct, if all sides are tied together they can't expand or contract. All they can do is develop multiple individual resonances. If these areas are then subdivided with diagonally placed braces, then you're done since triangular shapes have a hard time flexing with any well defined Fs (low Q). [​IMG]
    You mass load/brace the drivers by making a cradle to hold the motor in place while applying a very slight, but constant, force against it when installed on the baffle.
    ====
    >> I use R-19 fiberglass when I stuff, but how tall are the towers? Once you move away from a golden ratio cab, the length begins to impart pipe resonances to the response, increasing with increasing length.
    >>Stuffing a vented cab to increase its apparent size isn't a good plan since it normally overdamps the speaker/lowering vent output too much so you want to use the absolute minimum that attenuates any mids reflections and/or standing waves. Normally lining the end opposite the vent, one wall, and back is all you should need if you built a stiff cab.
    >yeah I do not plan on using stuffing to increase the apparent size. I want to use it to dampen any standing waves etc. Would fiberglass work good here?
    ====
    If the cab is stiff enough, otherwise you'll need to use Deflex pads and fiberglass (or equal).
    ====
    > How about stuff like Dacron?
    ====
    Don't have any experience with it, nor do I recall reading about anyone using it for just lining walls, just for stuffing.
    ====
    > The size I want to make it is 45.5 inches tall x 9 inches wide x 11.5 inches deep, out of MDF. .75 inches for all but front baffle, front baffle 1.5 inches.
    ====
    I don't have time right now, but this needs to be modeled in MK's MLTQWT spreadsheet to see how it compares to the golden ratio Vb/Fb/vent size, and to find out how much stuffing at the top is required. It may be fine, but if it were me I wouldn't start cutting until the results are in.
    ====
    > The drivers I have are from this page http://www.madisound.com/audaxsale.html
    AW025S1 1" SHIELDED PURE TITANIUM DOME TWEETER (for the tweeter)
    HT170F0 6-1/2" FIBERGLASS CONE WOOFER (for the mid/woofer)
    >Not the world best speakers but should do OK for a first set of speakers. I have built a pair of stereo amps from ESP (http://sound.westhost.com/project3a.htm) and a active crossover from him. I plan on actively biamping the speakers.
    >I do have a stock crossover design from Madisound to work from for crossover points etc.
    ====
    OK. Looks like good challenge, have fun! [​IMG]
    ====
    I am just concerned about the boxes and where I should go. I am buying the MDF tonight and will probably start cutting it tomorrow so any help would be awesome!!!!
    ====
    Aarg! I guess I'm too late. Bummer. [​IMG] 3/4" BB plywood is the Hot Ticket for tower designs.
    ====
    >Thanks for all your time.
    Ted
    ====
    You bet! [​IMG]
    GM
     
  5. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Once again thank you [​IMG]
    >>Correct, if all sides are tied together they can't expand or contract. All they can do is develop multiple individual resonances. If these areas are then subdivided with diagonally placed braces, then you're done since triangular shapes have a hard time flexing with any well defined Fs (low Q).
    How about diagonal bracing through out? From top to back to front to back to bottom. Would that work?
    >>You mass load/brace the drivers by making a cradle to hold the motor in place while applying a very slight, but constant, force against it when installed on the baffle.
    ?
    Still a little confused here. Would the basket be pressing into the speaker or pulling it away from it. Would it attach to the magnet? If so would that make the driver permanently installed so it could not be removed?
    >> the cab is stiff enough, otherwise you'll need to use Deflex pads and fiberglass (or equal).
    Ok this goes back to the bracing. How do I determine if it is stiff enough? Also is standard fiberglass thin enough or would I have to split it?
    >>>The size I want to make it is 45.5 inches tall x 9 inches wide x 11.5 inches deep, out of MDF. .75 inches for all but front baffle, front baffle 1.5 inches.
    ====
    >>don't have time right now, but this needs to be modeled in MK's MLTQWT spreadsheet to see how it compares to the golden ratio Vb/Fb/vent size, and to find out how much stuffing at the top is required. It may be fine, but if it were me I wouldn't start cutting until the results are in.
    I was talking to a friend of mine tonight who is a cabinet maker and we decided to change the dimensions a little. How does 41" high x 10 " wide x 11.5" deep with the same MDF dimensions sound (.75 for all but front which is 1.5")
    >>Aarg! I guess I'm too late. Bummer. 3/4" BB plywood is the Hot Ticket for tower designs
    Yeah I thought of BB plywood. Cost kept me away from it for this project. The cost for a 5x5 3/4" sheet is $80.00 CAD here where as MDF is $35.00 CAD and I need 2 sheets of MDF for everything I want to do. So I probably would have needed 3 or 4 sheets of BB. Next time I will use it. There is always a next time [​IMG]
    >> OK. Looks like good challenge, have fun!
    Thanks. I hope all goes well and thanks again for your help.
    Ted
     
  6. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    >Once again thank you
    ====
    Sure! Beats rehashing subs ad nauseum (bbbooorrriiinnnggg).
    ====
    >How about diagonal bracing through out? From top to back to front to back to bottom. Would that work?
    ====
    As long as it accomplishs the goals I listed, your imagination is the limit. I've seen the wood equivalent of a Birdcage Maserati chassis for bracing, so whatever floats yer boat. [​IMG] There is a point of diminishing returns though.
    Considering the dims of your towers, a second layer of MDF all around should suffice since stiffness goes up at the cube of thickness and energy storage decreases at the square of stiffness, and it saves you lots of work. This would raise the MDF's base MOE from ~0.33x to ~2.7x what 3/4" BB's average is. Then all you need is a horizontal shelf brace between the two drivers, or one behind each if that's the way you want to mass load/brace them. The tradeoff is weight, though in a tower design that's not necessarily a bad thing.
    Regardless, with a 1.5" thick baffle you should chamfer out the driver holes on the backside at a 45deg so they don't 'feel' like they're in a short tunnel.
    ====
    Still a little confused here. Would the basket be pressing into the speaker or pulling it away from it. Would it attach to the magnet? If so would that make the driver permanently installed so it could not be removed?
    ====
    Hmm, you want to do this, or the functional equivalent: http://melhuish.org/audio/images/press-screw.gif
    If you still don't follow, I'm out of explanations. [​IMG]
    ====
    >Ok this goes back to the bracing. How do I determine if it is stiff enough?
    ====
    I'm sure a mechanical engr. can calculate it, but it's an experience thing for me. I don't use MDF for several reasons, so all my advice is based on using high MOE (stiff) materials. With MDF you have to mass it up to get stiffness/damping, so bracing doesn't do much unless it's pretty comprehensive, like Adire's 'x' shelves. If you choose not to just 'double up' per the above, then considering the aspect ratio, the shelf braces behind the drivers and another one dividing the lower expanse in a ~0.62:1.62 ratio with a hardwood closet pole running top to bottom through them and all glued together will suffice.
    ====
    > Also is standard fiberglass thin enough or would I have to split it?
    ====
    I use the ~1" thick stuff you find on the back of acoustic ceiling tile, or what you buy at PE, etc., though for stuffing like will be required in the top, I use the ~4-8" thick stuff (depending) you stick between wall studs, just don't leave the paper backing on.
    ====
    >I was talking to a friend of mine tonight who is a cabinet maker and we decided to change the dimensions a little. How does 41" high x 10 " wide x 11.5" deep with the same MDF dimensions sound (.75 for all but front which is 1.5")
    ====
    MK's spreadsheets rule IMO, so without modeling, I can't say what's best. I do recommend you make the width/depth inside dims in a golden ratio though, so if these were the final i.d. dims that represented the optimum cross sectional area, then I would make them 8.42" and 13.62" to spread out the standing waves amplitudes (low Q) that can reflect back through the cones. God is in the details. No single one may have much, if any, audible effect, but a bunch of them makes for superior performance both audible and perceived.
    ====
    >Yeah I thought of BB plywood. Cost kept me away from it for this project. The cost for a 5x5 3/4" sheet is $80.00 CAD here where as MDF is $35.00 CAD and I need 2 sheets of MDF for everything I want to do. So I probably would have needed 3 or 4 sheets of BB. Next time I will use it. There is always a next time
    ====
    Hmm, the ratio is about the same here, but ya'll's higher inflation takes a bigger bite. [​IMG]
    ====
    >> OK. Looks like good challenge, have fun!
    Thanks. I hope all goes well and thanks again for your help.
    Ted
    ====
    You bet! Later dude, it's pushing 3:00am here and I'm getting a little long in the tooth for 22hr days. [​IMG]
    GM
     
  7. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Have I said how much I appreciate your help? Really it's been great!!! I am learning so much [​IMG]
    >>Sure! Beats rehashing subs ad nauseum (bbbooorrriiinnnggg).
    Whats wrong with subs [​IMG]
    >>As long as it accomplishes the goals I listed, your imagination is the limit. I've seen the wood equivalent of a Birdcage Maserati chassis for bracing, so whatever floats yer boat. There is a point of diminishing returns though.
    [​IMG] Makes sense. I'll talk with Mac (my cabinet maker friend) and see what he thinks. I would imaging the main thing is to solidly have all 6 sides connected together.
    >>Considering the dims of your towers, a second layer of MDF all around should suffice since stiffness goes up at the cube of thickness and energy storage decreases at the square of stiffness, and it saves you lots of work. This would raise the MDF's base MOE from ~0.33x to ~2.7x what 3/4" BB's average is. Then all you need is a horizontal shelf brace between the two drivers, or one behind each if that's the way you want to mass load/brace them. The tradeoff is weight, though in a tower design that's not necessarily a bad thing.
    Hmm I had thought about and discarded going to .75 inch all around for weight and money reasons. Would simply adding .25 of a inch to make a total of 1" for all but front and 1.5 in front do? I would like to keep these to under 100lbs each [​IMG]
     
  8. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    >Have I said how much I appreciate your help? Really it's been great!!! I am
    learning so much [​IMG]
    ====
    I hope so, I'd hate to think I've done all this typing just for the
    practice! [​IMG]
    ====
    >Whats wrong with subs [​IMG]
    ====
    Nothing. Started building them long before they became popular, but there's
    only a minimal amount of know-how required to designing/building a good one,
    then it's time to move on.
    ====
    >[​IMG] Makes sense. I'll talk with Mac (my cabinet maker friend) and see what
    he thinks. I would imaging the main thing is to solidly have all 6 sides
    connected together.
    ====
    Uh, this is the point I've been making all along.
    ====
    >Hmm I had thought about and discarded going to .75 inch all around for
    weight and money reasons. Would simply adding .25 of a inch to make a total
    of 1" for all but front and 1.5 in front do? I would like to keep these to
    under 100lbs each [​IMG]
    ====
    Well, every little bit helps, but is way shy of my standards, though that's
    neither here nor there. This increases stiffness from ~0.33x BB ply to
    ~0.69x. If you're only going to increase it a 1/4", then it would be better
    to use a much stiffer material, such as BB, Oak ply/whatever. This at least
    would pump it up to ~1.4x 3/4" BB ply's stiffness.
    ====
    >Once again how would I brace the drivers with a shelf? I think I understand
    the rod way you told me about just not with a shelf.
    ====
    Oh ye of little imagination...... if the screw/block puts a constant
    pressure on the motor, isn't that the same as a notch in the shelf that's
    the right distance to apply a little constant pressure when the driver's
    screwed into place?
    ====
    >I think I catch you now. Would I simply use small wood blocks, tee nuts and
    a metal rod? Then I can allow myself to adjust the pressure from the back?
    ====
    Hallelujah! [​IMG] Yes, it's one popular way.
    ====
    > is so how much pressure and how would I know when I have enough?
    ====
    When it sounds 'tight'. Since it varies, make a little sealed cab out of
    scrap and experiment. If these drivers are typical of other small point
    source drivers I'm familiar with the change won't be subtle.
    ====
    >I took a good look at the spreadsheets. They look good. Not sure what it
    all means with the graphs though but I will learn.
    ====
    Yeah, he made those for himself to learn with so there's no real structure
    to them. Someday he may take the time to formalize them in a GUI.
    ====
    >> I do recommend you make the width/depth inside dims in a golden ratio
    though, so if these were the final i.d. dims that represented the optimum
    cross sectional area, then I would make them 8.42" and 13.62" to spread out
    the standing waves amplitudes (low Q) that can reflect back through the
    cones. God is in the details. No single one may have much, if any, audible
    effect, but a bunch of them makes for superior performance both audible and
    perceived.
    Would using those dimensions really help? I chose the ones I did for
    stability as well as optimization reasons. As well I can use 12" veneers
    then [​IMG]
    ====
    Naw, 'course not, I just thought I'd baffle you with some low tech sounding
    BS. Would you notice a worthwhile difference if you built one per your dims
    and one per mine? Maybe, maybe not, depending on all the variables. If one
    speaker was built basic and the other with attention to all the little
    details, then the difference would be audibly smoother.
    If you choose to use your dims, especially not knowing if there's enough
    cross sectional area for a smooth midbass (and I'm betting it's not since
    most drivers need 3x-7x Sd, while yours is only ~2.6x), then more Deflex (or
    equal) padding will be required as well as more stuffing, which will also
    affect tuning output. Getting good performance in a tower design isn't as
    simple as with a sub box.
    ====
    >Yeah and even worse for where I live. MDF is $35 in town here (small town)
    where as it is $20.00 in Edmonton.
    ====
    That's cheap for Canadian rates isn't it? I mean it's ~$22/sheet down here
    in Hotlanta, which is only 2hrs away from a Ga-Pacific MDF mill.
    ====
    > No way to transport it up here though. My main source for good BB is in
    Edmonton.
    ====
    Some POS in a bio-degradeable bag stole my truck last May, and the insurance
    wasn't enough to replace it with anything I'd feel safe driving, and since a
    two seat roadster isn't much of a hauler I'm SOL too. Not that it matters
    much since I can't build anything anytime soon.
    ====
    > I use them for all my finishing stuff. They will even get me all my cherry
    veneer cut from the same log [​IMG]
    ====
    Cool! I don't think I have that option, but I was never big on fancy
    finishes, so never checked into it. I let the end users do anything better
    than wallpaper or basic painting. I do have some 152+yr old weathered
    barnwood that I planed down a couple of years ago to make some tower
    speakers for a friend, and they turned out georgeous, even before sealing:
    http://www.myhifisite.50megs.com/gm/tower1.jpg They now sport a hand rubbed
    clear satin gloss finish, planed 3/4" x 2.5" Redwood tripod base and solid
    polyfil damping pad sprayed with a tan vinyl paint.
    ====
    >Thanks a lot!!!!!!!!
    Ted
    ====
    Sure thing! [​IMG]
    GM
     
  9. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Here we go again. I think I almost have it all [​IMG] You've been great help. I don't know how I could ever repay you [​IMG]
    >>>Hmm I had thought about and discarded going to .75 inch all around for
    weight and money reasons. Would simply adding .25 of a inch to make a total
    of 1" for all but front and 1.5 in front do? I would like to keep these to
    under 100lbs each
    ====
    >>Well, every little bit helps, but is way shy of my standards, though that's
    neither here nor there. This increases stiffness from ~0.33x BB ply to
    ~0.69x. If you're only going to increase it a 1/4", then it would be better
    to use a much stiffer material, such as BB, Oak ply/whatever. This at least
    would pump it up to ~1.4x 3/4" BB ply's stiffness.
    Ok I'm not sure what I will do yet then... I'll have to think about it. I'm going to cut the inner box tomorrow anyways. I can always add an extra sheet to the outside later.
    Update ... I Think I will do the double layer of MDF as per your recomendations. [​IMG]
    ====
    >>> is so how much pressure and how would I know when I have enough?
    ====
    >>When it sounds 'tight'. Since it varies, make a little sealed cab out of
    scrap and experiment. If these drivers are typical of other small point
    source drivers I'm familiar with the change won't be subtle.
    So I would play a test tone through the speaker and increase the force?
    >> Oh ye of little imagination...... if the screw/block puts a constant
    pressure on the motor, isn't that the same as a notch in the shelf that's
    the right distance to apply a little constant pressure when the driver's
    screwed into place?
    Ok would this be similar to above? Figure out how far then do it to all via test tone? I would have to believe it is not a lot of pressure. What would I listen for. Does it go from bad to good to really bad?
    >>Naw, 'course not, I just thought I'd baffle you with some low tech sounding
    BS. Would you notice a worthwhile difference if you built one per your dims
    and one per mine? Maybe, maybe not, depending on all the variables. If one
    speaker was built basic and the other with attention to all the little
    details, then the difference would be audibly smoother.
    >>If you choose to use your dims, especially not knowing if there's enough
    cross sectional area for a smooth midbass (and I'm betting it's not since
    most drivers need 3x-7x Sd, while yours is only ~2.6x), then more Deflex (or
    equal) padding will be required as well as more stuffing, which will also
    affect tuning output. Getting good performance in a tower design isn't as
    simple as with a sub box.
    Sigh I get your point :b I will do as you say [​IMG] So how much cross sectional area do I get from your design. Is it 8.42 x 13.6? Is that my inner dimension? I believe it is. So then I can just figure out my height easily enough. I'll do that tomorrow [​IMG] Hmm that would make my height only 26". That is almost to short. I would like to keep it at least 36" high. How would you recommend I do that?
    Update. If I read you right I should just use the golden ratio of 1.618 for my front and back dimention. If I had a I.D front of 7.5 and a depth of 12.25 then my height would be 33.5. Would these work?
    >>Yeah and even worse for where I live. MDF is $35 in town here (small town)
    where as it is $20.00 in Edmonton.
    ====
    That's cheap for Canadian rates isn't it? I mean it's ~$22/sheet down here
    in Hotlanta, which is only 2hrs away from a Ga-Pacific MDF mill.
    Duno at least I know it is not all that bad of a price then [​IMG]
    Thanks again!!
    Ted
     
  10. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    >Here we go again. I think I almost have it all [​IMG] You've been great help. I don't know how I could ever repay you [​IMG]
    ====
    A bag full of $1000US bills works for me! [​IMG]
    ====
    >Ok would this be similar to above? Figure out how far then do it to all via test tone? I would have to believe it is not a lot of pressure.
    ====
    Depends on the driver size/construction. Probably not much on yours though.
    ====
    >What would I listen for. Does it go from bad to good to really bad?
    ====
    No, it tightens up. You're mass loading it, damping resonances. Once there's no more audible difference, then any more may damage it.
    ====
    >Sigh I get your point :b I will do as you say [​IMG] So how much cross sectional area do I get from your design.
    ====
    Same as yours, though the cross sectional area probably isn't optimum. I was just using your value as an example.
    ====
    > Is it 8.42 x 13.6? Is that my inner dimension?
    ====
    Yes. When discussing volume it should always be i.d./net.
    ====
    > I believe it is. So then I can just figure out my height easily enough. I'll do that tomorrow [​IMG]
    ====
    So you're going to use this cross sectional area anyway, without simming in mathcad, even after my explaining why it's probably not close to optimum?!
    GM
     
  11. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Great. I am getting close [​IMG] Thank you!!
    > I believe it is. So then I can just figure out my height easily enough. I'll do that tomorrow
    ====
    >>So you're going to use this cross sectional area anyway, without simming in mathcad, even after my explaining why it's probably not close to optimum?!
    Not really sure how much bigger I could go on the cross sectional area. I also have some space limits on the area I am going to put it in. I really can't have a short fat speaker. I know it may not the the most optimal but I am hoping I can make it work. I would like to use the i.d. of 7.5 wide, a depth of 12.25 then my height would be 33.5. The cross sectional area should be close to a golden ratio.
    I am going to use a double thickness of MDF all around as per your recommendations. It will give me a area of a little over 50L from which I can subtract the port area, bracing and speaker volume. It should be close to 45L and with a little stuffing to tame the standing waves I should be pretty close to optimal volume.
    How much Deflex (or acoustic ceiling tiles) should I use to optimize things.
    Also would a port of 4" with a length of 6.81" still work? If not how big should I go. As well does the position of the drivers and port make a big difference? I would like to use a MTM arrangement. Should I off set the tweeter? I see a few other speakers using it now.
    Thanks a lot for your help. I don't know how far I would have gotten with out it.
    Ted
     
  12. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    Been trying to email this for the past 2 hrs, but MSN has decided not to let me send any emails for whatever reason so maybe you'll notice it on the HTF:
    Hmm, I wrote a response, but the forum was down on several occasions when I
    tried to post it. Looks like I ultimately forgot about it, pretty common for
    me I'm afraid, so sorry 'bout that:
    >Great. I am getting close [​IMG] Thank you!!
    ====
    Hmm, not too much thanks, I don't see bags of $$ on my doorstep. [​IMG]
    ====
    >Not really sure how much bigger I could go on the cross sectional area. I
    also have some space limits on the area I am going to put it in. I really
    can't have a short fat speaker.
    ====
    Hmm, since they're going to be towers, why would it be short? Once you move
    into pipe design, Vb no longer is calculated like a standard T/S cab. It's a
    whole new ballgame.
    ====
    > I know it may not the the most optimal but I am hoping I can make it work.
    I would like to use the i.d. of 7.5 wide, a depth of 12.25 then my height
    would be 33.5. The cross sectional area should be close to a golden ratio.
    ====
    7.5*1.618 = 12.13", or ~91"^2, a ~21% reduction from the previous area. In a
    pipe design, it's this cross sectional area that is the most relevant WRT
    midbass/mids FR.
    Anyway, for an MTM you want the HF at ~ear height, so unless you're really
    short or sit on the floor, this is way too short since normally the HF would
    be up around 36-38" off the floor, so you'll need at least a 48"i.d. height.
    ====
    >I am going to use a double thickness of MDF all around as per your
    recommendations.
    ====
    Good plan.
    ====
    > It will give me a area of a little over 50L from which I can subtract the
    port area, bracing and speaker volume. It should be close to 45L and with a
    little stuffing to tame the standing waves I should be pretty close to
    optimal volume.
    ====
    OK, if you stick to a ~golden ratio cab.
    ====
    >How much Deflex (or acoustic ceiling tiles) should I use to optimize
    things.
    ====
    Depends on what you build.
    ====
    >Also would a port of 4" with a length of 6.81" still work?
    ====
    If you haven't changed the original Vb and it's ~ a golden ratio cab.
    ====
    >If not how big should I go.
    ====
    Depends on what you finally build. With tower designs, its length affects
    the vent dims.
    ====
    > As well does the position of the drivers and port make a big difference? I
    would like to use a MTM arrangement. Should I off set the tweeter? I see a
    few other speakers using it now.
    ====
    Yes, The vent needs to be at least 3" from any boundary, and if a pipe
    design then it needs to be near an end. Offsetting drivers in a golden ratio
    helps to spread baffle edge diffraction resonances, so I recommend it.
    ====
    >Thanks a lot for your help. I don't know how far I would have gotten with
    out it.
    ====
    You would have built what you wanted and delt with any sonic issues as best
    you could and pronounced it the best speakers you ever heard, or at least
    close to it. It's the bane of DIYers, we rationalize away anything. [​IMG]
    GM
    ----- Original Message -----
    From: "Ted_Polzin"
    To:
    Sent: Tuesday, January 07, 2003 11:30 AM
    Subject: Speaker
    > Hello:
    >
    > Any chance you could respond to my last questions. I am hoping you might
    be able to give me a little more help regarding how to optimize my speakers
    within what I have to live with. I know it might not be the best but it is
    already quite a bit better than when I first started. There must be ways to
    compensate for not having the most optimal cross sectional area and if you
    could help me find them I would be most grateful.
    >
    > Thanks a lot
    >
    > Ted
     
  13. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    >Great. I am getting close Thank you!!
    ====
    >>>Hmm, not too much thanks, I don't see bags of $$ on my doorstep.
    No worries. It is in the mail [​IMG] of course I mailed Canadian dollars so a bag .... after exchange .... and bank fees ... your probably net $.50
    >Not really sure how much bigger I could go on the cross sectional area. I
    also have some space limits on the area I am going to put it in. I really
    can't have a short fat speaker.
    ====
    >>>??Hmm, since they're going to be towers, why would it be short? Once you move
    into pipe design, Vb no longer is calculated like a standard T/S cab. It's a
    whole new ballgame.
    I am limited to around 50l ... If I have a large cross sectional area I need a shorter speaker ... or taller and a smaller cross section. What do you mean by a pipe design?
    > I know it may not the the most optimal but I am hoping I can make it work.
    I would like to use the i.d. of 7.5 wide, a depth of 12.25 then my height
    would be 33.5. The cross sectional area should be close to a golden ratio.
    ====
    >>>7.5*1.618 = 12.13", or ~91"^2, a ~21% reduction from the previous area. In a
    pipe design, it's this cross sectional area that is the most relevant WRT
    midbass/mids FR.
    Actually it is a little larger. I originally gave you the O.D. Given the speaker has a 140CM^2 Sd (21.5 ") and the sd is 91"^2 I should have 4.23 x Sd. If I am reading Mathcad right that is not too bad.
    A>>nyway, for an MTM you want the HF at ~ear height, so unless you're really
    short or sit on the floor, this is way too short since normally the HF would
    be up around 36-38" off the floor, so you'll need at least a 48"i.d. height.
    hrmmmm ... I could add a bottom peace that is not part of the internal volume. Would that work to make it higher? I still want to keep the overall height under 48". Would a 47" Speaker be fine .. 33.5 ID. I would have a bottom chamber that is "empty". Can it be single mdf or should I double that as well.
    >How much Deflex (or acoustic ceiling tiles) should I use to optimize
    things.
    ====
    >>>Depends on what you build.
    Ok Mathcad seems to like ~.8-.9 lbs / CF. What would that be .. the back and a side? I would then need around 1.3-1.4 lbs total.
    > As well does the position of the drivers and port make a big difference? I
    would like to use a MTM arrangement. Should I off set the tweeter? I see a
    few other speakers using it now.
    ====
    >>>Yes, The vent needs to be at least 3" from any boundary, and if a pipe
    design then it needs to be near an end. Offsetting drivers in a golden ratio
    helps to spread baffle edge diffraction resonances, so I recommend it.
    Ok. If I offset the tweeter would I make it flush with the edges of the mid base (umm yeah flush in a vertical line) also I would imagine you want the speakers opposite each other.
    You would have built what you wanted and delt with any sonic issues as best
    you could and pronounced it the best speakers you ever heard, or at least
    close to it. It's the bane of DIYers, we rationalize away anything.
    Actually these should be my first decent set of speakers so that goes with out question. I am using some Fluance speakers right now ....
    For a question on the mathcad spreadsheet. I'm using ML TQWT. The graph to look at for freq response is the SPL one in Far Field transmission line systems right? Also what does Driver Position ratio mean and how do I affect it?
    Thanks
    Ted
     
  14. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    Well, I think I finally got my email system problems rectified (I hope). F@#$%^& Microsoft.
    >No worries. It is in the mail [​IMG] of course I mailed Canadian dollars so a bag .... after exchange .... and bank fees ... your probably net $.50
    ====
    Uh, I stipulated US currency. [​IMG]
    ====
    I am limited to around 50l
    ====
    Why?!
    ====
    >... If I have a large cross sectional area I need a shorter speaker ... or taller and a smaller cross section. What do you mean by a pipe design?
    ====
    When you make one dim. considerably longer, its eigentones (standing waves) begin to affect the FR in dramatic, but predictable ways. Since both closed, open, and stopped pipes are common in audio, we often refer to tower designs as pipe designs.
    ====
    >Actually it is a little larger. I originally gave you the O.D.
    ====
    OK, FWIW, cab dims should always be expressed as internal (i.d.) for what I hope are obvious reasons.
    ====
    >Given the speaker has a 140CM^2 Sd (21.5 ") and the sd is 91"^2 I should have 4.23 x Sd. If I am reading Mathcad right that is not too bad.
    ====
    OK, but you're using two, correct? This means the Sd the box sees is 21.5*2 = 43"^2, or ~2.11x, not enough for good performance. If 4.23x works well, then the cab's cross sectional area has to double.
    ====
    >hrmmmm ... I could add a bottom peace that is not part of the internal volume. Would that work to make it higher? I still want to keep the overall height under 48". Would a 47" Speaker be fine .. 33.5 ID. I would have a bottom chamber that is "empty". Can it be single mdf or should I double that as well.
    ====
    You're not getting with the plan here. [​IMG] The cab needs to be this high to get not only get the tweeter at an acceptable height, but just as important is to keep the top mid driver away from the top of the cab.
    ====
    >Ok Mathcad seems to like ~.8-.9 lbs / CF. What would that be .. the back and a side? I would then need around 1.3-1.4 lbs total.
    ====
    Maybe, maybe not. From this post I get the strong impression that more simming is required, which will probably change this a bunch. In general though, you want a design that looks good with between 0.25-0.5lbs/ft^3.
    It need not be perfectly flat. Indeed, most folks can't tell the difference between flat and +/-3dB in this pipe's BW, so unless you're a perfectionist, +/- 1-2dB is plenty good enough.
    ====
    >Ok. If I offset the tweeter would I make it flush with the edges of the mid base (umm yeah flush in a vertical line) also I would imagine you want the speakers opposite each other.
    ====
    >Actually these should be my first decent set of speakers so that goes with out question. I am using some Fluance speakers right now ....
    ====
    Not familiar with these, but done ~like I've outlined, then once you get the XO dialed in these should be quite good by consumer audio standards.
    ====
    >For a question on the mathcad spreadsheet. I'm using ML TQWT.
    ====
    Right, this is the one to use.
    ====
    > The graph to look at for freq response is the SPL one in Far Field transmission line systems right?
    ====
    Right, with a comparison to an IB's response.
    ====
    > Also what does Driver Position ratio mean and how do I affect it?
    ====
    This where you tell the program where along the length of the pipe the driver(s) is/are located. It makes a major difference! The optimum is L/2, which means it's at the center point. For example, a 48" i.d. pipe, this equates to 24" down or up. But you have two drivers, so if they were butted up against each other, this would be the acoustic center, i.e. one driver above, one below. You need the tweeter at say 37" though, so if 24" = L/2, then 37" = 37/24 = ~1.54, so change the '2' to 1.54. If you can make the baffle so they will fit this way with the tweeter nestled in the 'v' of the two driver, then keep fiddling with cross sectional area/damping. If nothing looks good, try a longer L, which will shift the drivers closer to the middle, and adjust the driver position. I always start with a 74" pipe just to see what I get.
    Notice too that there's a total TL length which includes an end correction a little further down from where you input the stuffing density. This is the i.d. dim. you use to make the pipe, which changes depending on what's inputted, so you have to factor this in each time you play with driver position if it changes.
    If for whatever reason you can't make the baffle wide enough to allow this layout, then spread the mids apart only enough to fit the tweeter in. Don't worry about meeting the 1WL tweeter to mid D'Appolito spacing, it causes too much off axis comb filtering anyway. You just don't want to exceed it.
    If you wind up with much of a gap between them, then ideally you'll need to model both heights and mentally overlay them to see what their combined response will be.
    Just to make sure we're on the same page, MK's design is a tapered one, so you should input your new cross sectional area in both SO and SL. Also, in the driver's specs you need to double Vas, Sd, and change Bl to whatever it is for one times 1.414 since you're using two drivers wired in parallel.
    Last, but not least, check driver excursion down at the 'woofer displacement' graph. The vertical numbers list the one way excursion for 1W only, so for every doubling of power, whatever the graph shows will go up 2x. This will tell you how much power the drivers can safely handle.
    Unfortunately he didn't do a vent mach graph or calculation, so go as big as practical just to be safe, or play with values in another program until you get the same dia./length to check it.
    I know all this is a bit tedious/time consuming when all you want to do is drown yourself in sawdust, but trust me, get it fairly close and you'll be way ahead of the pack and will never go back to bookshelf speakers unless it's strictly due to space problems. I built these types designs for decades, experimenting/measuring, trying to figure out what was going on since there was nothing of use printed on the subject beyond rules-of-thumb that didn't work as well as what I'd figured out, and lacked the math skills to do what MK has done, so know what kind of performance they're capable of. His program is the 'real deal'.
    GM
     
  15. Ted_Polzin

    Ted_Polzin Stunt Coordinator

    Joined:
    Aug 12, 2002
    Messages:
    81
    Likes Received:
    0
    Trophy Points:
    0
    Sorry for taking so long. I got sent on a trip at work. Damn work getting in the way of speaker building [​IMG]
    >>Uh, I stipulated US currency.
    Oppse [​IMG] ...
    >>>I am limited to around 50l
    ====
    >>Why?!
    Thats what the drivers seemed to like in all tests.
    >>OK, FWIW, cab dims should always be expressed as internal (i.d.) for what I hope are obvious reasons.
    Yup. Won't make that mistake again [​IMG]
    >>hrmmmm ... I could add a bottom peace that is not part of the internal volume. Would that work to make it higher? I still want to keep the overall height under 48". Would a 47" Speaker be fine .. 33.5 ID. I would have a bottom chamber that is "empty". Can it be single mdf or should I double that as well.
    ====
    >>You're not getting with the plan here. The cab needs to be this high to get not only get the tweeter at an acceptable height, but just as important is to keep the top mid driver away from the top of the cab.
    But if I like the internals could I not add height with out adding to the internal size by just adding a bottom box that is glued similar to a speaker stand. This would raise the tweets up.
    ====
    >>> Also what does Driver Position ratio mean and how do I affect it?
    ====
    >>This where you tell the program where along the length of the pipe the driver(s) is/are located. It makes a major difference! The optimum is L/2, which means it's at the center point. For example, a 48" i.d. pipe, this equates to 24" down or up. But you have two drivers, so if they were butted up against each other, this would be the acoustic center, i.e. one driver above, one below. You need the tweeter at say 37" though, so if 24" = L/2, then 37" = 37/24 = ~1.54, so change the '2' to 1.54.
    OK your talking in relation to the top of the box. So /2 would be in the middle ... /1.5 would be in the top 1/3?
    >>If you can make the baffle so they will fit this way with the tweeter nestled in the 'v' of the two driver,
    In the "v" so they are in a complete vertical line or so the tweet is offset to one side. Like this
    xxxxx
    x m x
    xt x
    x m x
    xxxxx
    I sent you a possible response via e-mail. Can you let me know what you think. It seemed to be pretty flat and within a few decibels top to bottom.
    Thanks again
    Ted
     
  16. Greg Monfort

    Greg Monfort Supporting Actor

    Joined:
    May 30, 2000
    Messages:
    884
    Likes Received:
    0
    Trophy Points:
    0
    > Sorry for taking so long.
    ====
    Why? It's your thread. I figured you were simming away, looking for the perfect alignment. [​IMG]
    ====
    > I got sent on a trip at work. Damn work getting in the way of speaker building [​IMG]
    ====
    Yeah, how dare 'life' get in the way of our hobbies! [​IMG]
    ====
    > >>Why?!
    >
    > Thats what the drivers seemed to like in all tests.
    ====
    Even when using pipe dims?
    ====
    > >>You're not getting with the plan here. The cab needs to be this high to get not only get the tweeter at an acceptable height, but just as important is to keep the top mid driver away from the top of the cab.
    >
    > But if I like the internals could I not add height with out adding to the internal size by just adding a bottom box that is glued similar to a speaker stand. This would raise the tweets up.
    ====
    I find it hard to believe that a pair of 6.5" drivers won't need at least a 4ft pipe, but if they don't, then center the driver cluster in the middle of the pipe and either space them up, or extend the cab all the way down and *completely* fill the void with kitty litter to add some mass and damp them.
    ====
    > >>This where you tell the program where along the length of the pipe the driver(s) is/are located. It makes a major difference! The optimum is L/2, which means it's at the center point. For example, a 48" i.d. pipe, this equates to 24" down or up. But you have two drivers, so if they were butted up against each other, this would be the acoustic center, i.e. one driver above, one below. You need the tweeter at say 37" though, so if 24" = L/2, then 37" = 37/24 = ~1.54, so change the '2' to 1.54.
    >
    > OK your talking in relation to the top of the box. So /2 would be in the middle ... /1.5 would be in the top 1/3?
    ====
    Correct.
    ====
    >
    > >>If you can make the baffle so they will fit this way with the tweeter nestled in the 'v' of the two driver,
    >
    > In the "v" so they are in a complete vertical line or so the tweet is offset to one side. Like this
    ====
    Offset.
    ====
    >
    >
    > I sent you a possible response via e-mail. Can you let me know what you think. It seemed to be pretty flat and within a few decibels top to bottom.
    ====
    Didn't get it, or if I did I automatically deleted it. If you sent me an attachment, I don't download them off the server unless I know about it in advance. I get 200-300 emails daily and with only a slow dial-up don't want to waste the time. Send it again and I'll look for it.
    GM
     

Share This Page