Seeing as this is my first foray into any kind of speaker box design, I was horrified to learn that my own home sound system,
a 15" top and a very small 18" sub had never been tuned properly.
I built the top myself a few years ago, from some plans I found in an old book, and I thought it sounded great, it wasn't bad that's for sure.
But I had never heard, nor did anyone ever tell me, about the proper concept of bass reflex and tuning an enclosure.
The sub itself is a cheapo "off - the - shelf" product, given to me by a friend.
Neither of these has any where near the correct port length.
So, I have a query.... About WinISD
Box Volume
From what I have been reading, the box volume in the "box" tab, this is a net figure, right.....?
As in, no driver, no braces, no ports, nothing except the air in the enclosure.....?
Everything I read would seem to indicate this.... Except for the "DIY Audio Guy" youTube channel, who tells me you must "fold" the vent INTO the box volume.
This is not correct, is it.....?
In WinISD, should the box volume bee added to the ports, or are the ports already in the box.
Will somebody please confirm this before I cut the pipe, I want to do this right the first time.
Because if the ports are to bee added to a box already built, this is going to skew the port length a little in the simulation.
Picture of the two stripped down enclosures attached
a 15" top and a very small 18" sub had never been tuned properly.
I built the top myself a few years ago, from some plans I found in an old book, and I thought it sounded great, it wasn't bad that's for sure.
But I had never heard, nor did anyone ever tell me, about the proper concept of bass reflex and tuning an enclosure.
The sub itself is a cheapo "off - the - shelf" product, given to me by a friend.
Neither of these has any where near the correct port length.
So, I have a query.... About WinISD
Box Volume
From what I have been reading, the box volume in the "box" tab, this is a net figure, right.....?
As in, no driver, no braces, no ports, nothing except the air in the enclosure.....?
Everything I read would seem to indicate this.... Except for the "DIY Audio Guy" youTube channel, who tells me you must "fold" the vent INTO the box volume.
This is not correct, is it.....?
In WinISD, should the box volume bee added to the ports, or are the ports already in the box.
Will somebody please confirm this before I cut the pipe, I want to do this right the first time.
Because if the ports are to bee added to a box already built, this is going to skew the port length a little in the simulation.
Picture of the two stripped down enclosures attached
Done !!!
👍👍👍
I just made the pipes as long as they could sensibly bee from the back wall.
Start low and measure.... we'll go from there.
Ballpark simulations were calculated to get some gauge of port volume first.
More questions..... What is the deal with tuning for SPL vs. tuning for a flat response?
Does a hefty sub in the 300 litre range really reproduce bass well..... I did hear someone say the L'Acoustics wouldn't sound too great but EQ can sort that out.
Because you have to..... Not really a problem here, we have decent EQ, but it perked my interest nonetheless.
More soon
👍👍👍
I just made the pipes as long as they could sensibly bee from the back wall.
Start low and measure.... we'll go from there.
Ballpark simulations were calculated to get some gauge of port volume first.
More questions..... What is the deal with tuning for SPL vs. tuning for a flat response?
Does a hefty sub in the 300 litre range really reproduce bass well..... I did hear someone say the L'Acoustics wouldn't sound too great but EQ can sort that out.
Because you have to..... Not really a problem here, we have decent EQ, but it perked my interest nonetheless.
More soon
Maybe later..... It has been 6 months since I posted in this thread.
A messy divorce followed by a summer of wild parties and hanging out on the DJ circuit..... Time to align myself..... Anyway
So, I managed to build the two enclosures above...... The monitor top with it's new porting and added crossover circuit was a huge improvement.
Crossover Circuit Thread - "One Driver - Two Tweeters"
https://www.diyaudio.com/community/threads/one-driver-two-tweeters.407515/
I got my hands on a DATS v3 and tested the monitor and what has now been christened the "MiniSub". They hit 40 and 30 hertz respectively, although l do believe the ports on the sub, that were intentionally left too long (better than too short), were most definitely too long..... I was shooting for 35Hz on the Minisub, so I just low cut the bottom 5 hertz out and the bass sounded much tighter.
Both enclosures were simulated in WinISD and the two graphs over laid to find a crossover point using the SPL plot graph.
I;m not sure if that is the correct way to do this this, but when set to 69 hertz, the sound system sounded even better again and I am very happy with those two completed speakers.
More on the "Monster Front Loaded BR sub" shortly..... It is almost completely assembled and is at least ready for some kind of testing now.
I will update here later tonight.
A messy divorce followed by a summer of wild parties and hanging out on the DJ circuit..... Time to align myself..... Anyway
So, I managed to build the two enclosures above...... The monitor top with it's new porting and added crossover circuit was a huge improvement.
Crossover Circuit Thread - "One Driver - Two Tweeters"
https://www.diyaudio.com/community/threads/one-driver-two-tweeters.407515/
I got my hands on a DATS v3 and tested the monitor and what has now been christened the "MiniSub". They hit 40 and 30 hertz respectively, although l do believe the ports on the sub, that were intentionally left too long (better than too short), were most definitely too long..... I was shooting for 35Hz on the Minisub, so I just low cut the bottom 5 hertz out and the bass sounded much tighter.
Both enclosures were simulated in WinISD and the two graphs over laid to find a crossover point using the SPL plot graph.
I;m not sure if that is the correct way to do this this, but when set to 69 hertz, the sound system sounded even better again and I am very happy with those two completed speakers.
More on the "Monster Front Loaded BR sub" shortly..... It is almost completely assembled and is at least ready for some kind of testing now.
I will update here later tonight.
Attachments
Finally..... Something to show
So I made an enclosure for testing...... Based on Mr. Flexys KS28 Single 18" Sub Design
Linked again for the convienience of othes.
From here on in the design has recieved the designation of KS-18x..... After all, why call a single 18 KS28 anything but..... and the small x stands for "experimental"
I have made a few small changes to the original youtube plan..... Firstly, it was scaled up for 25mm MDF...... I'm still not sure if this is a good idea but what the hell..... It will have wheels and I like a solid box so bit of a moot point really.
Internal are identical..... Except it is 50mm wider..... And 30mm deeper inside the flat "floor".
And I also included two 12mm quadrant rails for the curve...... Which is now effectively, a "kitchen drawer" and can bee replaced by as many "kitchen drawers" as I can make of whatever curvature I desire for testing.
I am still waiting on some board from the suppliers of different thickness for the flat "floor".
The idea being that these can bee juggled around to get a flat floor of varying heights.
One thing that is missing is the "wobbly" board in the throat.
I have left it out for now but I know it is essential..... If anyone would like to take a guess at how "wide" the wobbles should bee I would must appreciate it.
So I made an enclosure for testing...... Based on Mr. Flexys KS28 Single 18" Sub Design
Linked again for the convienience of othes.
From here on in the design has recieved the designation of KS-18x..... After all, why call a single 18 KS28 anything but..... and the small x stands for "experimental"
I have made a few small changes to the original youtube plan..... Firstly, it was scaled up for 25mm MDF...... I'm still not sure if this is a good idea but what the hell..... It will have wheels and I like a solid box so bit of a moot point really.
Internal are identical..... Except it is 50mm wider..... And 30mm deeper inside the flat "floor".
And I also included two 12mm quadrant rails for the curve...... Which is now effectively, a "kitchen drawer" and can bee replaced by as many "kitchen drawers" as I can make of whatever curvature I desire for testing.
I am still waiting on some board from the suppliers of different thickness for the flat "floor".
The idea being that these can bee juggled around to get a flat floor of varying heights.
One thing that is missing is the "wobbly" board in the throat.
I have left it out for now but I know it is essential..... If anyone would like to take a guess at how "wide" the wobbles should bee I would must appreciate it.
Attachments
Last edited:
So, even without the "wobbly" board and the "flat floor" boards I went ahead anyway and set up for testing.
Because of my earlier work on the two flared pipe BR builds further up this page it was as simple as replace one sub for the other.
Annd..... It sounds absolutely awful..... Very quiet.... Nowhere near as loud as the much smaller BR sub beside it.
I didn't expect it to bee perfect but I am very underwhelmed.
Another quick and dirty measuremnet with the DATS v3 and it tells me the built combination is 30Hz.
It should bee about that with the "floorboards" and "wobble"..... Not that it matters, I just wanted to flex my new DATS muscles and test something.
If anybody would like to input opinions and/or suggestions I would much appreciate that too.
More photos..... Just for fun
Because of my earlier work on the two flared pipe BR builds further up this page it was as simple as replace one sub for the other.
Annd..... It sounds absolutely awful..... Very quiet.... Nowhere near as loud as the much smaller BR sub beside it.
I didn't expect it to bee perfect but I am very underwhelmed.
Another quick and dirty measuremnet with the DATS v3 and it tells me the built combination is 30Hz.
It should bee about that with the "floorboards" and "wobble"..... Not that it matters, I just wanted to flex my new DATS muscles and test something.
If anybody would like to input opinions and/or suggestions I would much appreciate that too.
More photos..... Just for fun
Attachments
Last edited:
If you are not cutting the top end off with a steep LP ~100Hz you'll be hearing the awful upper pipe resonances of the long port:
The smaller sub is probably tuned higher, so may sound louder.
If you are calling this the "wobbly" board:
it might affect tuning slightly, but won't make much difference in sound, your Fb is what it is, slightly over 30Hz.
Who knows what "Mr. Flexy's actual box (if there is one) Fb is, but the simulated response in the video would suggest it's higher than yours.
Make the port wider in the center to raise Fb to ~36Hz, the impedance peaks either side of Fb will be close to even, frequency response will flatten out, and the box will put out more low end.
Try your comparisons with the subs in a room corner or against the wall, the back wall reflection could be killing the low end.
Cheers,
Art
Get outside and do a ground plane measurment. No matched crossover and a small room are not the way you judge a sub. You are at the point where your ear is not the right tool to find out what is right or wrong.
The impedance sweep doesn't look too bad, the resonace you see above 100 Hz don't do much harm if you cross it like 60Hz 18dB/oct or steeper.
You went for 25mm MDF because you wanted to save on bracing? Sorry, but that doesn't work. Only makes it heavier. Even 30mm will not helb with the large surfaces of a 18" cabinet.
It may be a good idea to measure it closed, so you compare how effectice the port is, as you mentioned "not loud enough". Measure the impedance of any cabinet you compare, a 4 Ohm box will be much louder than one with 8 Ohm. You got to match voltages to get identical energy/ Watt input.
The impedance sweep doesn't look too bad, the resonace you see above 100 Hz don't do much harm if you cross it like 60Hz 18dB/oct or steeper.
You went for 25mm MDF because you wanted to save on bracing? Sorry, but that doesn't work. Only makes it heavier. Even 30mm will not helb with the large surfaces of a 18" cabinet.
It may be a good idea to measure it closed, so you compare how effectice the port is, as you mentioned "not loud enough". Measure the impedance of any cabinet you compare, a 4 Ohm box will be much louder than one with 8 Ohm. You got to match voltages to get identical energy/ Watt input.
Last edited:
Maybe I should have been clearer..... That's what I get for posting 3 times in 2 days without a reply.
1. The KS-18x is crossed over at about 70Hz..... Exactly how the smaller sub was set up so I just switched one for the other just for testing.
I also have 31-band EQ for removing any bandwidths if I need to.
2. Both subs measured about 30Hz
Note - The measuring rig was different for each subs..... I used a Speakon adaptor on the MiniSub but the L'Acoustics clone was hard wired through a sealed grommet .
I am awaiting screw - on speakon plugs in the post because I can't solder.
This will eliminate a variable and should bee sorted after the weekend.
I will attach pictures of my measuring gear below.
3. Yes, that is the "wobble"
4. I am also beginning to doubt that "Mr. Flexy's Single KS28 Sub" ever existed
5. I have already made the port 30mm higher than the one in the youTube video to allow for some tuning.
Mr. Flexys port is 70mm at the narrowest point and 170mm at the mouth.
Test box is 100mm and 200mm, respectively.
I thought I had left enough extra space for some port tuning, but maybe not.
The only way I can widen the port any more is by making a shallower curve.
Picture below.
6. I am still awaiting more boards from the supplier..... Once I get a delivery I will have the ability to block the port altogether or even slide in a flat sheet where my "kitchen drawer" should bee..... They should bee here after the weekend.
7. All enclosures are 4 ohm.
The "MiniSub" has an 18Sound 18LW2400 (1200w rms) and the L'Acoustics clone is wearing a FaitalPro 18XL1600 (1600w)
1. The KS-18x is crossed over at about 70Hz..... Exactly how the smaller sub was set up so I just switched one for the other just for testing.
I also have 31-band EQ for removing any bandwidths if I need to.
2. Both subs measured about 30Hz
Note - The measuring rig was different for each subs..... I used a Speakon adaptor on the MiniSub but the L'Acoustics clone was hard wired through a sealed grommet .
I am awaiting screw - on speakon plugs in the post because I can't solder.
This will eliminate a variable and should bee sorted after the weekend.
I will attach pictures of my measuring gear below.
3. Yes, that is the "wobble"
4. I am also beginning to doubt that "Mr. Flexy's Single KS28 Sub" ever existed
5. I have already made the port 30mm higher than the one in the youTube video to allow for some tuning.
Mr. Flexys port is 70mm at the narrowest point and 170mm at the mouth.
Test box is 100mm and 200mm, respectively.
I thought I had left enough extra space for some port tuning, but maybe not.
The only way I can widen the port any more is by making a shallower curve.
Picture below.
6. I am still awaiting more boards from the supplier..... Once I get a delivery I will have the ability to block the port altogether or even slide in a flat sheet where my "kitchen drawer" should bee..... They should bee here after the weekend.
7. All enclosures are 4 ohm.
The "MiniSub" has an 18Sound 18LW2400 (1200w rms) and the L'Acoustics clone is wearing a FaitalPro 18XL1600 (1600w)
Attachments
So you did not like your ks18x ?
The plan was all wrong and the resulting sound to your ears is horrible?
The plan was all wrong and the resulting sound to your ears is horrible?
It has been a while since I have posted here but in between other projects, this one was not forgotten.
After a night or two of playing with EQ and amp gain settings the Faitals seemed to "break - in"
Huge improvement.
It was used in this configuration for a while and tested with plenty of actual music.
Not bad then.
This is "Test 1"
Soon after this I realised I had made a mistake in my measurements, the box (not the port) was 25mm shorter in height.
I had already widened the box by 50mm and scaled up from 18mm board to 25.... Any way....
Adjustments were made and I ended up with 25mm extra net box volume and while I was at it I decided to widen the port a little as well.
This raised the tuning to 36Hz but the impedance spikes were quite uneven.
"Test 2"
Substituting the curve for a flat board raised the tuning again only slightly but very much evened out the impedence graph.
This is the bigger box with more port volume...... Looks much better.
"Test 3"
One final idea was tried with the curve sitting on a fitted flat floor (effectively keeping things the same as
"Test 3" but making the box smaller)
And this is "Test 4"
Comments and feedback anyone.....?
After a night or two of playing with EQ and amp gain settings the Faitals seemed to "break - in"
Huge improvement.
It was used in this configuration for a while and tested with plenty of actual music.
Not bad then.
This is "Test 1"
Soon after this I realised I had made a mistake in my measurements, the box (not the port) was 25mm shorter in height.
I had already widened the box by 50mm and scaled up from 18mm board to 25.... Any way....
Adjustments were made and I ended up with 25mm extra net box volume and while I was at it I decided to widen the port a little as well.
This raised the tuning to 36Hz but the impedance spikes were quite uneven.
"Test 2"
Substituting the curve for a flat board raised the tuning again only slightly but very much evened out the impedence graph.
This is the bigger box with more port volume...... Looks much better.
"Test 3"
One final idea was tried with the curve sitting on a fitted flat floor (effectively keeping things the same as
"Test 3" but making the box smaller)
And this is "Test 4"
Comments and feedback anyone.....?
Attachments
Last edited:
Sorry to go off-topic slightly, but I wondered if you had any links to white papers or studies relating noise spectra to hearing damage.
In my 40-odd years in acoustics, we so often hear the myth that bass doesn't damage hearing as quickly as higher frequencies (usually from car stereo kids), but despite searching I have never seen hard data to support it.
@ maxolini - Thank you Sir,
It has been 6 months since I posted, I was busy in the meantime trying to learn as much as I could while also
engaged with other projects.... But there is no substitute for hands on practical experience.
Knowing what I know now, I think I should have enlarged the port (raised the tuning) instead of making the box
itself bigger first.
I didn't really know what I was looking at then, and I am still no expert now.
I would like to find some kind of pattern in the impedance plots as I make changes but al ot of it was confusing
and the curve was steep (no pun intended).
The magnitude of the peaks changed a lot while it was much harder to move the Fc around more than a couple of Hertz.
Am I correct in saying that the reflex impedance graphs, with it's two peaks are a function of excursion (cone movement).
This excursion is controlled by the amount of air the driver is pushing, so as air is added (by increasing box size or port volume),
these spikes are getting smaller and smaller respectively, indicating less cone movement caused by increased acoustic resistance.
For example, if the right spike is higher, this means the tuning is high (higher than the FS of the driver), and visa versa.
Lowering the tuning by narrowing or lengthening the port (more port air resistance I do believe) seems to bee almost identical to doing the opposite -
making the box bigger, all other things being equal.
And raising the tuning (by widening or shortening the port) seems to correlate to again, making the net box volume smaller.
It took me a long time of reading, and no one ever explicitly explained to me, to see that making the port longer and narrower lowers the tuning....
And making the port wider and shorter does the inverse.
I think I have it now..... Or, at least a firm grasp of it.... Am I standing on a good foundation here.....?
Feel free to correct me at any time or call me out.
More coming soon....
It has been 6 months since I posted, I was busy in the meantime trying to learn as much as I could while also
engaged with other projects.... But there is no substitute for hands on practical experience.
Knowing what I know now, I think I should have enlarged the port (raised the tuning) instead of making the box
itself bigger first.
I didn't really know what I was looking at then, and I am still no expert now.
I would like to find some kind of pattern in the impedance plots as I make changes but al ot of it was confusing
and the curve was steep (no pun intended).
The magnitude of the peaks changed a lot while it was much harder to move the Fc around more than a couple of Hertz.
Am I correct in saying that the reflex impedance graphs, with it's two peaks are a function of excursion (cone movement).
This excursion is controlled by the amount of air the driver is pushing, so as air is added (by increasing box size or port volume),
these spikes are getting smaller and smaller respectively, indicating less cone movement caused by increased acoustic resistance.
For example, if the right spike is higher, this means the tuning is high (higher than the FS of the driver), and visa versa.
Lowering the tuning by narrowing or lengthening the port (more port air resistance I do believe) seems to bee almost identical to doing the opposite -
making the box bigger, all other things being equal.
And raising the tuning (by widening or shortening the port) seems to correlate to again, making the net box volume smaller.
It took me a long time of reading, and no one ever explicitly explained to me, to see that making the port longer and narrower lowers the tuning....
And making the port wider and shorter does the inverse.
I think I have it now..... Or, at least a firm grasp of it.... Am I standing on a good foundation here.....?
Feel free to correct me at any time or call me out.
More coming soon....
Port tuning is very roughly independent of its dimensions for a given volume - it is the mass of air in the port which matters.
In a ported speaker the air in the enclosure is a spring and the air in the port is a mass. Imagine a mass hanging on long soft spring, gently bouncing up and down - the frequency of oscillation is not affected by the shape of the mass! (In reality we have friction, air resistance, and other losses so these generalisations are not accurate to ten decimal places...).
A few general rules about ports:
In a ported speaker the air in the enclosure is a spring and the air in the port is a mass. Imagine a mass hanging on long soft spring, gently bouncing up and down - the frequency of oscillation is not affected by the shape of the mass! (In reality we have friction, air resistance, and other losses so these generalisations are not accurate to ten decimal places...).
A few general rules about ports:
- Begin the design with a port area equal to the speaker's effective area (Sd) and make it smaller until it fits in the box. (I have only ever once achieved port area = cone area!).
- Use a single circular port.
- Ensure that the port length is no longer than three times its diameter.
- Radius the port at both ends to ensure equality of air movement at each end to avoid port rectification.
- Make the port pretty rigid, and support large ports if necessary.
- Don't fret about decimal places when port tuning - in the real world it will change several whole Hertz simply with temperature and output level.
Based on some empirical research I've done, the "classic" "port tuning is independent of volume dimensions" only applies if the volume occupied by the port is less than 10% of the overall volume. Once you go above that, it's time to perhaps consider modeling the enclosure differently if possible (e.g. ODTL).
Some notes about the impedance peaks:
When the impedance peaks in a bass reflex system are not the same height, it indicates an asymmetry in the system’s response that can reveal specific design characteristics or issues. Let’s break down what this means and why it happens.
Impedance Curve Basics in Bass Reflex
In a well-tuned bass reflex enclosure, the impedance curve typically shows:
Two peaks flanking the tuning frequency (Fb), where the port resonates and the driver’s excursion is minimized.
The lower-frequency peak (below Fb) is influenced by the driver’s free-air resonance (Fs) interacting with the enclosure’s compliance (box stiffness).
The higher-frequency peak (above Fb) reflects the port’s resonance interacting with the driver.
A dip at Fb, where the port takes over most of the acoustic output.
Ideally, these peaks are roughly equal in height in a critically damped or optimally tuned system (e.g., a Butterworth alignment), though exact equality isn’t always the goal—specific alignments like Chebyshev or QB3 may intentionally skew them.
What Unequal Peaks Mean
When the peaks differ in height, it’s a sign that the system’s damping, tuning, or volume isn’t perfectly balanced. Here’s how to interpret it:
Lower Peak Taller Than Upper Peak
Cause: This typically happens when the box volume (Vb) is too small relative to the driver’s Vas (compliance equivalent volume).
The driver’s resonance dominates because the enclosure’s stiffness overly restricts its motion, boosting the lower-frequency peak.
Effect on Response: The system leans toward over-damping. The SPL graph might show a weaker port contribution, with bass rolling off earlier than desired and lacking depth.
Implication: The tuning frequency (Fb) may be too high for the box size, or the driver’s Qts is higher than ideal for the design.
Upper Peak Taller Than Lower Peak
Cause: This occurs when the box volume is too large or the port is tuned too low. The port resonance becomes more dominant, amplifying the higher-frequency peak as the system oscillates more freely.
Effect on Response: The system becomes under-damped, with a boomy peak in the SPL graph near or just above Fb, indicating excessive energy around the tuning frequency.
Implication: The driver may not be well-controlled by the enclosure, leading to poor transient response and a “loose” sound.
Other Factors
Driver Parameters: A driver with a high Qts (>0.4) naturally skews the lower peak higher in smaller boxes, while a low Qts (<0.3) might emphasize the upper peak in larger boxes
Port Design: If the port is lossy (e.g., too narrow, too long, or poorly flared), it can dampen the upper peak by reducing the port’s efficiency, making the lower peak appear taller.
Leaks or Damping Material: Box leaks or excessive internal stuffing can suppress one peak (usually the upper one) by absorbing port energy, throwing off the balance.
Should You Worry?
Not Necessarily: Unequal peaks don’t automatically mean the design is “wrong.” Some alignments (e.g., extended bass shelf or sub-Chebyshev) intentionally produce uneven peaks to achieve a specific frequency response trade-off, like deeper extension at the cost of flatness.
Listen and Measure: Check the SPL graph and listen to the system. If the sound is boomy (upper peak dominant) or weak (lower peak dominant), adjust the design. The impedance curve is a diagnostic tool, not the final verdict.
As I intimated, it's the ideal, but rarely achievable; it's to be taken as a starting point.