You can find the formulae for port design in Loudspeaker Design Cookbook. It's a basic book that anyone building a speaker should have.
If I planned on building more speakers and designing the cabinets myself then I would indeed buy the book that you suggest SY. Since I am merely building the design that is in the PDF and don't plan on building another pair of speakers maybe you could be so kind as to answer my simple question. Thanks either way.
G
G
G said:
No, I believe that you would need to make the port slightly shorter. Mind you, I'm not a tuned-port speaker man.
Steve
For a tubular duct, the following equation applies:
L = (2118*D^2/(Fb^2*Vb)) - 0.85*D,
where Fb is the box tuning frequency in Hz, L and D are length and diameter of the port, respectively, and Vb is the box volume in cubic feet.
L = (2118*D^2/(Fb^2*Vb)) - 0.85*D,
where Fb is the box tuning frequency in Hz, L and D are length and diameter of the port, respectively, and Vb is the box volume in cubic feet.
Which, according to my calculations means that (assuming the box volume and tuning frequency are constant):
If the length required is 3" for a 3.25" diameter port, then
the length would need to be about 2.4" for a 3" diameter port.
But, to misquote Jim Carey in The Mask:
"Somebody check me."
Steve
If the length required is 3" for a 3.25" diameter port, then
the length would need to be about 2.4" for a 3" diameter port.
But, to misquote Jim Carey in The Mask:
"Somebody check me."
Steve
Ports, dampers volume etc
This is my first ported project (now I'm scared)🙂 but I've done a couple of sealed. I do intend to measure response, and my box is oversize for exactly the reason you cite.
A question on measurement. What I have done is put in a refrence line in my spectrum analyzer of pink noise - 10 second sample, with peak hold on. This yields a frown shape dropping off in both frequency ends in a nice curve. Every speaker I've got more or less conforms to this curve (I normallize all my samples) and I can see the praks and valleys that correspond to what I can hear. So.. I can theorize about why are certain peak happens, and make a correction.
I use a ruler flat condenser in omni mode a foot or two from the speaker. As I pull back I can see the room effects leak in. (Most mikes are flatter in omni mode than cardiod, even though one would like to use a cardiod pattern for such a measurement to reduce room effects. I put the pink noise at about 85 db so the speakers are working inside their sweet spot and priximity effect is sort of in range (ie about how I would listen to them).
My question is: Am I doing this right?
This is my first ported project (now I'm scared)🙂 but I've done a couple of sealed. I do intend to measure response, and my box is oversize for exactly the reason you cite.
A question on measurement. What I have done is put in a refrence line in my spectrum analyzer of pink noise - 10 second sample, with peak hold on. This yields a frown shape dropping off in both frequency ends in a nice curve. Every speaker I've got more or less conforms to this curve (I normallize all my samples) and I can see the praks and valleys that correspond to what I can hear. So.. I can theorize about why are certain peak happens, and make a correction.
I use a ruler flat condenser in omni mode a foot or two from the speaker. As I pull back I can see the room effects leak in. (Most mikes are flatter in omni mode than cardiod, even though one would like to use a cardiod pattern for such a measurement to reduce room effects. I put the pink noise at about 85 db so the speakers are working inside their sweet spot and priximity effect is sort of in range (ie about how I would listen to them).
My question is: Am I doing this right?
I'd do two things:
1. Near field. Get that mike within 1/4 inch of the cone, if you can. Likewise, when you measure the port, get that mike right at the outer surface. If you want to get a full freq response, do a vector sum of the two measurements, scaled to their relative areas. d'Appolito gives gory details in "Measuring Loudspeakers."
2. Measure impedance. Having a crosscheck to validate measurements is a really good idea.
1. Near field. Get that mike within 1/4 inch of the cone, if you can. Likewise, when you measure the port, get that mike right at the outer surface. If you want to get a full freq response, do a vector sum of the two measurements, scaled to their relative areas. d'Appolito gives gory details in "Measuring Loudspeakers."
2. Measure impedance. Having a crosscheck to validate measurements is a really good idea.
To get back to the thread (Damping Materials for Speaker Cabinets) I've got an interesting one.
I've just put together a prototype for my new speaker. It's a tall, narrow column and I designed the cabinet to require minimum damping as I believe that too much can make the speaker sound lifeless. I'm talking here about the 100Hz to 20kHz range, which I cover using four Bandor 2" units (see my link below for my original speaker using this concept).
Anyway, I think I may have overdone it. I've listened to the speaker with no damping at all and it sounds fine. Of course, I'll have to spend some time tuning by ear and also by measurement and I look forward to doing this next week (I want better than fine). But I'm curious to get other views ...
My walls are pretty dead to vibration so let's leave out panel damping and concern ourselves with:
A) Putting some stuffing in the middle of the column (wool or fibreglass wadding) and
B) Sticking something on the inner walls (carpet, wool, foam, etc.)
So guys, in your considerable collective experience, how much of A and B is too much and how much is not enough? I'm interested in anyone who has played around with this either by ear or measurement.
Steve
Oh, and Theo - I've just seen your message - if you can see discrepancies in the measurements and correlate them with what you hear, I'd say: Yes, you're doing just fine.
I've just put together a prototype for my new speaker. It's a tall, narrow column and I designed the cabinet to require minimum damping as I believe that too much can make the speaker sound lifeless. I'm talking here about the 100Hz to 20kHz range, which I cover using four Bandor 2" units (see my link below for my original speaker using this concept).
Anyway, I think I may have overdone it. I've listened to the speaker with no damping at all and it sounds fine. Of course, I'll have to spend some time tuning by ear and also by measurement and I look forward to doing this next week (I want better than fine). But I'm curious to get other views ...
My walls are pretty dead to vibration so let's leave out panel damping and concern ourselves with:
A) Putting some stuffing in the middle of the column (wool or fibreglass wadding) and
B) Sticking something on the inner walls (carpet, wool, foam, etc.)
So guys, in your considerable collective experience, how much of A and B is too much and how much is not enough? I'm interested in anyone who has played around with this either by ear or measurement.
Steve
Oh, and Theo - I've just seen your message - if you can see discrepancies in the measurements and correlate them with what you hear, I'd say: Yes, you're doing just fine.
SY said:
Thanks for that SY. I've read a couple of reviews and put the book on order. It's a "must read".
Steve
All - thx for your help on the measurement stuff.
Steve on the choice between 1 and 2 I'm confused. This is a transmission line right? Stuffing the line makes the line "longer" - retuning it, so I assume you tune that by measurement to the intended frequency. Stuffing = tuning in a t-line.
The box section counts as part of the transmission line, so I would not stuff it either except to retune it. I don't know why damping would mess up liveliness except to the degree where it makes for more "stored energy" simply because it impedes the flow to the transmission line. The "special egg shape" supposedly eliminates internal vibrations (not sure how) - so what would you need damping for?
In any case I'd put any stuffing in the transmission line way down inside, and then measure and look for outages that relate to multiples of the box dimensions at certain wavelengths. Like if the box is 20 inches wide and you have a dip at a the frequency that corresponds to a 40 inch long wavelength. If you don't, no harm, no foul, right?
Steve on the choice between 1 and 2 I'm confused. This is a transmission line right? Stuffing the line makes the line "longer" - retuning it, so I assume you tune that by measurement to the intended frequency. Stuffing = tuning in a t-line.
The box section counts as part of the transmission line, so I would not stuff it either except to retune it. I don't know why damping would mess up liveliness except to the degree where it makes for more "stored energy" simply because it impedes the flow to the transmission line. The "special egg shape" supposedly eliminates internal vibrations (not sure how) - so what would you need damping for?
In any case I'd put any stuffing in the transmission line way down inside, and then measure and look for outages that relate to multiples of the box dimensions at certain wavelengths. Like if the box is 20 inches wide and you have a dip at a the frequency that corresponds to a 40 inch long wavelength. If you don't, no harm, no foul, right?
TheoM said:
Actually no, my original design was a TL but my new speaker is a sealed box. As I'm only taking it down to 100Hz (and then handing over to a pair of bass cabinets) I thought that I wouldn't do a TL this time.
So, stuffing or padding (let's call lining the inside walls "padding" to differentiate it from bracing or damping the cabinet) is to reduce internal reflections and resonances which, presumably, have an audible effect by their actions on and through the drive units.
Steve
I came up with a port length of 2.27" for an enclosure of 45 liters with a port diameter of 3" and an FB of 50Hz. Thanks for the equation Sy. Would sanding the inner and outer inside edges of the tube reduce turbulence?
G
G
stuffing or padding?
Sealed: Stuff it. Kill the backwave. Actually, I'd add a low end damper to the walls like that roofing felt (stuffing does not do well with low end), hanging loosely along the walls, AND stuffing. I've seen high end sealed cabs stuffed to the gills - packed tight. Some people advocate for moderate stuffing which I read as 1/2 full - but I'd fill it up without packing it in, so to speak. I'd also spray the inside walls with that spray damper (on sale at parts express) just for good luck and extra sealing. I've heard that the sum of the micropores in a wooden box adds up to significant leakage. ted.
Sealed: Stuff it. Kill the backwave. Actually, I'd add a low end damper to the walls like that roofing felt (stuffing does not do well with low end), hanging loosely along the walls, AND stuffing. I've seen high end sealed cabs stuffed to the gills - packed tight. Some people advocate for moderate stuffing which I read as 1/2 full - but I'd fill it up without packing it in, so to speak. I'd also spray the inside walls with that spray damper (on sale at parts express) just for good luck and extra sealing. I've heard that the sum of the micropores in a wooden box adds up to significant leakage. ted.
At 3" diameter with that high of a tuning frequency, I doubt that you'll have a turbulence problem. If you're worried about port noise, put the port on the back of the cabinet.
Steve, maybe it's worth trying some lessons from Stealth technology; have the surface behind the driver be absorptive and dissipative (a density gradient with densest material on the cabinet side of the absorber might be interesting to try, using a nice, lossy foam) and structure the surface as a group of angles, like an F117 surface. If it doesn't work any better than normal stuffing, it can still be hyped as a Major Breakthrough to help you awe magazine reviewers.
Stealth damper diffuser
The stealth idea is good. I recently did something exactly like it, based on the same idea. Three things going on here. Check the attachment. 1) All dampers also diffuse/reflect. 2) What gets through the damper bounces off the back wall and takes another trip through the damper. 3) When the wave enters a denser medium it loses energy AND diffracts, changing its direction. So to scatter the energy, reducing big nodes and standing waves behind the driver I attached a slice of an auralex bass trap behind the driver. Because of the random and steep angles it creates a sort of chaos in which random events are more likely to occur - and by definition more of them will be nodes or cancellations, but because the energy is so disorganized after leaving the trap all the nodes would be relatively small. I'm sure there is chaos theory to back this up - but I can't do the math. Also, since little would reflect directly back to the driver there would be little left to impinge on the driver directly. Also, because its thicker behind the driver, the highest energy interference is least likely to occur. Bottom line our backwave either reflects obliquely, or gets absorbed, or gets attenuated AND refracted TWICE on its first trip across the box. Take that, Backwave.
The stealth idea is good. I recently did something exactly like it, based on the same idea. Three things going on here. Check the attachment. 1) All dampers also diffuse/reflect. 2) What gets through the damper bounces off the back wall and takes another trip through the damper. 3) When the wave enters a denser medium it loses energy AND diffracts, changing its direction. So to scatter the energy, reducing big nodes and standing waves behind the driver I attached a slice of an auralex bass trap behind the driver. Because of the random and steep angles it creates a sort of chaos in which random events are more likely to occur - and by definition more of them will be nodes or cancellations, but because the energy is so disorganized after leaving the trap all the nodes would be relatively small. I'm sure there is chaos theory to back this up - but I can't do the math. Also, since little would reflect directly back to the driver there would be little left to impinge on the driver directly. Also, because its thicker behind the driver, the highest energy interference is least likely to occur. Bottom line our backwave either reflects obliquely, or gets absorbed, or gets attenuated AND refracted TWICE on its first trip across the box. Take that, Backwave.
- Status
- Not open for further replies.