Ok ...just read my own post and it reads like a loaded (no pun intended) question.
During the frequency response test what I found was that the woofer closer to the port had a much higher excursion that the other near the box resonance frequency.
Also the impedance curve was not as well defined as I would have expected.
This made me quite unhappy. So I redesigned the front panel so that the port to woofer (centre) distance was the same for both woofers. My perceived problem went away, i.e. both woofers now behave exactly the same.
That seemed to make sense to me. But now I'm thinking all the dual woofer boxes that I have seen to date are not symetrical, i.e. the distance between each woofer and the port are different. So now I'm wondering are the designers of these trying to achieve something like maybe balancing or extending response?
Cheers
During the frequency response test what I found was that the woofer closer to the port had a much higher excursion that the other near the box resonance frequency.
Also the impedance curve was not as well defined as I would have expected.
This made me quite unhappy. So I redesigned the front panel so that the port to woofer (centre) distance was the same for both woofers. My perceived problem went away, i.e. both woofers now behave exactly the same.
That seemed to make sense to me. But now I'm thinking all the dual woofer boxes that I have seen to date are not symetrical, i.e. the distance between each woofer and the port are different. So now I'm wondering are the designers of these trying to achieve something like maybe balancing or extending response?
Cheers
Port Location;
If the air inside a box is tactile i.e. compresses and expands with the woofer's movement and if sound has a defined speed inside the box then the distance from the woofer to the port must matter.
But I have been unable to find a formula or a guide on what this effect is.
Or is this a non issue because most boxes are not big enough for this to be a problem? Some articles that I have read have stated that the actual location of the port is uninportant as long as the dimensions are correct. What about dual woofer boxes with only one port at one end? The port to woofer distance is different for each woofer.
Any ideas ?
So many questions........that won't go away ...my head hurts...
Sigh
If the air inside a box is tactile i.e. compresses and expands with the woofer's movement and if sound has a defined speed inside the box then the distance from the woofer to the port must matter.
But I have been unable to find a formula or a guide on what this effect is.
Or is this a non issue because most boxes are not big enough for this to be a problem? Some articles that I have read have stated that the actual location of the port is uninportant as long as the dimensions are correct. What about dual woofer boxes with only one port at one end? The port to woofer distance is different for each woofer.
Any ideas ?
So many questions........that won't go away ...my head hurts...
Sigh
I would have thought that since the woofers act in unison to compress the air inside the box as a whole, the distance to the vent would not matter. The box is operating in pressure mode not propagation mode, so the 'time to travel' theory should not hold. However, your finding is interesting. Had you not moved the vent to a central location and found a change, I would have said that the driver specs are simply quite different.
It is also surprising considering that so much of this kind of thing is measured not in inches, but in wavelengths.
The port carries frequencies from the box resonsance frequency to one octave above it. The port output is greatest at the box resonance frequency, and gradually rolls off until there is little left at 100 Hz. If your port is tuned to 50 Hz, (a fairly high box resonance frequency), then the port is active from 50 Hz to 100 Hz, essentially inactive above that. The cone output supersedes the port output from 70 Hz on up, though it is aided by the port output until 100 Hz or so.
The wavelength of 100 Hz is over 11 feet. So wherever the port is located in a practical sized box is likely to be a small percentage of that. If there are two woofers in the box, the difference in length between them is likely to be negligible compared to the wavelength of the frequencies passing through the ports.
I suppose it is possible to build a seven foot tall box with one woofer on the top, one on the bottom, and the port on the bottom. There, you would have an appreciable difference in wavelength. Unless it is something like that, though, I must admit I am surprised-but I do not doubt your observation. Just surprised the phenomenon exists.
The port carries frequencies from the box resonsance frequency to one octave above it. The port output is greatest at the box resonance frequency, and gradually rolls off until there is little left at 100 Hz. If your port is tuned to 50 Hz, (a fairly high box resonance frequency), then the port is active from 50 Hz to 100 Hz, essentially inactive above that. The cone output supersedes the port output from 70 Hz on up, though it is aided by the port output until 100 Hz or so.
The wavelength of 100 Hz is over 11 feet. So wherever the port is located in a practical sized box is likely to be a small percentage of that. If there are two woofers in the box, the difference in length between them is likely to be negligible compared to the wavelength of the frequencies passing through the ports.
I suppose it is possible to build a seven foot tall box with one woofer on the top, one on the bottom, and the port on the bottom. There, you would have an appreciable difference in wavelength. Unless it is something like that, though, I must admit I am surprised-but I do not doubt your observation. Just surprised the phenomenon exists.
Yes kelticwizard;
That's what I thought too, until I saw what I saw. That is the speaker next to the port had a much higher excursion than the one further away. Both woofers by the way were next to each other with their frames no more than 50mm apart.
I examined a friends PA setup with 3 PA drivers (dunno why he had 3) and saw the same thing but now with 3 speakers. Guess what, the driver excursion decreased as the distance from the port increased. I.e. the middle woofer moved more than the furthest woofer but less than the closest woofer.
This cannot be explained by wavelength as you suggested so what is it?
Is it the actual localised complaince of the air volume.
If driver distortion increases with excursion then it may be worth pursuing the reasoning and subsequent fix.
Cheers
That's what I thought too, until I saw what I saw. That is the speaker next to the port had a much higher excursion than the one further away. Both woofers by the way were next to each other with their frames no more than 50mm apart.
I examined a friends PA setup with 3 PA drivers (dunno why he had 3) and saw the same thing but now with 3 speakers. Guess what, the driver excursion decreased as the distance from the port increased. I.e. the middle woofer moved more than the furthest woofer but less than the closest woofer.
This cannot be explained by wavelength as you suggested so what is it?
Is it the actual localised complaince of the air volume.
If driver distortion increases with excursion then it may be worth pursuing the reasoning and subsequent fix.
Cheers
Re: Yes kelticwizard;
Raises a lot of interesting questions. To keep from burdening the discussion with too many questions at once, I will just ask this one for now: were the woofers hooked up in series or in parallel? Does this phenomenon change when you switch between series and parallel hookup?
quasi said:
Raises a lot of interesting questions. To keep from burdening the discussion with too many questions at once, I will just ask this one for now: were the woofers hooked up in series or in parallel? Does this phenomenon change when you switch between series and parallel hookup?
Higher excursion in the woofer tends to happen if the port is too close because the woofer begins to function the inside of the port like an open baffle. Take an extreme example: if you have a port the same diameter as the woofer and it terminates close to the back of the woofer, the woofer will basically see itself as inside an open tubular baffle and pretty much ignore your nice big box you so carefully calculated. The open baffle will provide negligible cone reinforcement thus allowing it to flap in the breeze.
Re: Re: Yes kelticwizard;
My woofers were 2 x 16 ohm 10" foam suspension types wired in parallel to give 8 ohms. I did not try a series configuration.
Some of the responses indicate that a port too close to the woofer will not provide enough loading or suspension on the cone. So while it will depend on the speakers and the boxes characteristics, how far from the woofer should the port be? Can there be a formula?
Cheers
kelticwizard said:
My woofers were 2 x 16 ohm 10" foam suspension types wired in parallel to give 8 ohms. I did not try a series configuration.
Some of the responses indicate that a port too close to the woofer will not provide enough loading or suspension on the cone. So while it will depend on the speakers and the boxes characteristics, how far from the woofer should the port be? Can there be a formula?
Cheers
Ahhh....I'm sitting here typing and and the same thing is happening to me. I'm just trying to determine the freq of the MTM box I have just constructed. It has a couple of 8 inch drivers and a four inch port. The boxes are about 900 mm (3ft) high and the two drivers are above each other with the tweeter in the middle, one in the upper part of the baffle the other in about the centre of the baffle. The port is in the bottom part of the baffle. My calculations put the box at about 68 litres and the port tuning at about 36 Hz. When I feed a signal of 35 Hz into the box the top driver is virtually stationary and there is lots of output from the port. However the lower driver ( not that low as it is still about 11 inches centre to centre from the port) still has plenty of excursion. When I up the sig generator up to about 48 Hz the lower driver has it's minimum excursion!
Confusing! I expect that the upper speaker is behaving correctly, and the lower one is being unduly influenced by the port, but the port isn't that close. There are plenty of single driver boxes with the driver closer to the port. My guess is that the upper speaker also interacts with the lower one in conjunction with the port and for some reason (unknown to me) together they force the resonance of the lower (centre) speaker higher. It is almost like it is at the fulcrum between the two. The three elements (top, bottom and port) act together to create a much more complex system than is described by the thiele/small equations which really concerns itself with two elements.
I can only hope that the end result is a port that contributes to extending the response even more. There are plenty of expensive examples of the MTM design, so perhaps it isn't a problem. The alternative is that as the MTM design relies on both drivers behaving identically, this descrepancy will really screw up the sound. However as this is occuring a very long way below the crossover point, perhaps it is not as important as it first appears.
What do you think?
Confusing! I expect that the upper speaker is behaving correctly, and the lower one is being unduly influenced by the port, but the port isn't that close. There are plenty of single driver boxes with the driver closer to the port. My guess is that the upper speaker also interacts with the lower one in conjunction with the port and for some reason (unknown to me) together they force the resonance of the lower (centre) speaker higher. It is almost like it is at the fulcrum between the two. The three elements (top, bottom and port) act together to create a much more complex system than is described by the thiele/small equations which really concerns itself with two elements.
I can only hope that the end result is a port that contributes to extending the response even more. There are plenty of expensive examples of the MTM design, so perhaps it isn't a problem. The alternative is that as the MTM design relies on both drivers behaving identically, this descrepancy will really screw up the sound. However as this is occuring a very long way below the crossover point, perhaps it is not as important as it first appears.
What do you think?
I can't speak for Quasi, but I ran tests on my drivers and they were not identical. However Fs is within one hertz of each other and Qts is within 3%, and I think that such a large difference in behaviour between the drivers in the box can not be explained by variation in the driver specs.
I must admit I am suprised at what I see. It is certainly not a theoretical thing. I can reach out from my desk here and touch the drivers and feel the difference. Look.... I'm reaching out now and can feel the cones....There must be many members out there with similar boxes. Perhaps most measure the box res freq by measuring the impedance at the speaker terminals, however my understanding is that the most direct measurement is the speaker movement (or really lack of movement at resonance when most of the output in a reflex cabinet is from the port, and it is this phenomonem is that causes the changes in measured impedence)
I must admit I am suprised at what I see. It is certainly not a theoretical thing. I can reach out from my desk here and touch the drivers and feel the difference. Look.... I'm reaching out now and can feel the cones....There must be many members out there with similar boxes. Perhaps most measure the box res freq by measuring the impedance at the speaker terminals, however my understanding is that the most direct measurement is the speaker movement (or really lack of movement at resonance when most of the output in a reflex cabinet is from the port, and it is this phenomonem is that causes the changes in measured impedence)
Hi richie00boy. There is a bit over six inches between the inside end of the port and the back of the box. There is no stuffing at all, but all surfaces including the bracing is covered with 12mm (1/2") acoustic foam. However as the box is very heavliy braced, some of the braces are reasonably close to the port. However I don't think this would explain more than a hertz or two of variation.
Sounds OK then. Normally you should allow at least a vent diameter between the vent end and any walls. Stuffing should be confined to lining the walls. All of which you seem to have complied with perfectly.
How about if you seal the vent and feed the box a sine wave sweep signal? How do the drivers move?
How about if you seal the vent and feed the box a sine wave sweep signal? How do the drivers move?
I just pressed a heavy book (CRC handbook of chem and physics! Doesn't get much heavier than that) against the port to seal it. As I swept up the freq there was not any one freq with no driver cone movement for either driver. I guess you asked me to do this to check if there were any other leaks etc that may contribute to a sort of secondary port action??
I once had that problem in a sealed enclosure and it wasn't a difference in the drivers. I had run the input leads to one driver and then the other. The first driver was moving more than the second. Hard to believe but true. To cure this I simply ran the + to one driver and the - to the other driver and then patched them over. Problem solved. Still can't figure out why.
Cal
Cal
Cal, I don't quite understand what you mean by "...ran the + to one driver and the - to the other ..." Do you mean that they started as parallel and ended up as in series?
The drivers are nominally 4 ohms and are wired in series and have identical Re, so both are experiencing the same current and voltage drop. And it isn't a case of one moving more than the other, it is rather a case of one ceasing to move at one frequency and the other ceasing to move at a much higher frequency.
This is not quite true. The top speaker appears to be acting close to what would be expected if it was the only driver in the box and almost (not quite) stops moving at about the box res.~35Hz. The second driver between the top one and the port however almost stops at a much higher freq. ~ 48Hz. However, even at this freq it is still moving more than the top one at ~35Hz.
The drivers are nominally 4 ohms and are wired in series and have identical Re, so both are experiencing the same current and voltage drop. And it isn't a case of one moving more than the other, it is rather a case of one ceasing to move at one frequency and the other ceasing to move at a much higher frequency.
This is not quite true. The top speaker appears to be acting close to what would be expected if it was the only driver in the box and almost (not quite) stops moving at about the box res.~35Hz. The second driver between the top one and the port however almost stops at a much higher freq. ~ 48Hz. However, even at this freq it is still moving more than the top one at ~35Hz.