Hi there,
I was playing with winISD tryin to figure what ports to use, then came across the fact that you could have say 15, 4mm diameter 18mm long ports, which would produce a green vent mach (not sure what this means, but i assume you want it green)
This would allow for the ports to simply be holes straight through 18mm mdf, on one panel.
I'm sure there are reasons not to do this, i have a feeling it might be to do with port noise. But would it work? and would if any port noise be noticeable?
cheers. dave.
I was playing with winISD tryin to figure what ports to use, then came across the fact that you could have say 15, 4mm diameter 18mm long ports, which would produce a green vent mach (not sure what this means, but i assume you want it green)
This would allow for the ports to simply be holes straight through 18mm mdf, on one panel.
I'm sure there are reasons not to do this, i have a feeling it might be to do with port noise. But would it work? and would if any port noise be noticeable?
cheers. dave.
It would be an aperiodic box...
And there is nothing wrong with those either
Infact they tend to be better damped at 'port' resonance than a normal BR.
I say build it, and give it a listen, the worst you'd have to do is fill the holes, and drill a true port.
To that end, try to use a standard dowel size
Owen
And there is nothing wrong with those either
Infact they tend to be better damped at 'port' resonance than a normal BR.
I say build it, and give it a listen, the worst you'd have to do is fill the holes, and drill a true port.
To that end, try to use a standard dowel size
Owen
This is more an experiment, using this driver in a small enclosure, form over function is the name of the game here.
I kind of need a satisfactory result (project deadline is soon) first time.
My plan is:
5 liter box, tuned to 50hz using 101, 1mm diameter 18mm holes.
As as far as I can tell using this driver in this size cab, would otherwise need ports far to long to easily integrate(I have a fair bit of electronics to cram in, active and wireless...)
Or should i just use a sealed enclosure, and forfeit the little extra low end extension?
cheers.
I kind of need a satisfactory result (project deadline is soon) first time.
My plan is:
5 liter box, tuned to 50hz using 101, 1mm diameter 18mm holes.
As as far as I can tell using this driver in this size cab, would otherwise need ports far to long to easily integrate(I have a fair bit of electronics to cram in, active and wireless...)
Or should i just use a sealed enclosure, and forfeit the little extra low end extension?
cheers.
Hi daverave.
If you want to try aperiodic try tuning the box to about 100 - 200 hz (where the Z curve starts to rise) using 10-15 holes about 3/8 - 5/8" then stuff the holes with tontine etc and some extra damping in the box.
Manipulating the damping can flatten out the z curve, improving phase response and good bass response from an undersized cab, at the expense of a bit of efficiency.
It does take some time, and many sweeps to get it right. Don't aim for too flat a z curve, it kills the dynamics.
Geoff
If you want to try aperiodic try tuning the box to about 100 - 200 hz (where the Z curve starts to rise) using 10-15 holes about 3/8 - 5/8" then stuff the holes with tontine etc and some extra damping in the box.
Manipulating the damping can flatten out the z curve, improving phase response and good bass response from an undersized cab, at the expense of a bit of efficiency.
It does take some time, and many sweeps to get it right. Don't aim for too flat a z curve, it kills the dynamics.
Geoff
This is a bug within WinISD in the calculation of port mach. ARe you using "WinISD beta" or "WinISD "Pro" alpha"? Most software doesn't tell you what's inside, and testing these fringe conditions on complicated systems, using simple models, won't get you anywhere.
Simple port equations assume that mass dominates. Anytime you get really small in diameter you will have a mixed mass-resistance element. It no longer acts as a port and starts to act mainly as a leak. You won't gain any extension this way - in fact you may lose some. The resistance in the mass-resistance element will be nonlinear and in order to do a nonlinear sim of a speaker you would need a lot more than Fs, Qts, Vas, etc....
Simple port equations assume that mass dominates. Anytime you get really small in diameter you will have a mixed mass-resistance element. It no longer acts as a port and starts to act mainly as a leak. You won't gain any extension this way - in fact you may lose some. The resistance in the mass-resistance element will be nonlinear and in order to do a nonlinear sim of a speaker you would need a lot more than Fs, Qts, Vas, etc....
Hi Geoff, think I will give that a go cheers.
I guess the best way forward will be to make the side panel removable, and then I can swap and change panels with different numbers of holes etc, to try and find an optimum sound.
How would this type of design affect the crossover? are then any particular types of network suited to this? again I have little experience in this area, and simply looking to build something which I like the sound of.
I was originally going to use Zaphs complete design, but the specification of this project needed something different, and now find myself rather lost without a proven design to follow and set of seas drivers! if all fails, after the project is finished I can re-use the drivers.
I guess the best way forward will be to make the side panel removable, and then I can swap and change panels with different numbers of holes etc, to try and find an optimum sound.
How would this type of design affect the crossover? are then any particular types of network suited to this? again I have little experience in this area, and simply looking to build something which I like the sound of.
I was originally going to use Zaphs complete design, but the specification of this project needed something different, and now find myself rather lost without a proven design to follow and set of seas drivers! if all fails, after the project is finished I can re-use the drivers.
Removable side sounds like a plan.
Wrt the port - the boundary layer is the key to resistive flow - the more holes, the bigger this effect is in relation to the 'sprung' mass of the air in the ports.
I would assume that your port diameter is around 0.5 mm less than the hole you'll be drilling - that should account for the boundary losses, but the number of holes will increase.
This could sound really, really nice, as the port will in essence be damped by the resistive flow at the edges, and that could take the resonant Q of the port closer to .707 (critically damped), although some stuffing behind the 'portlets' will take the damping beyond .707, and will take the box into aperiodic behaviour.
So, get drilling a temporary removable side.
Start with the original calculated number, measure the FR, increment the number of holes, measure, and so on, until you find an FR that sounds great. This way you'll have some idea of the deviation from the model.
One thing is certain - the 'peak' of the port will be lower, but broader than a standard port, I'm guessing, you could end up with an ebs style response...
Owen
Wrt the port - the boundary layer is the key to resistive flow - the more holes, the bigger this effect is in relation to the 'sprung' mass of the air in the ports.
I would assume that your port diameter is around 0.5 mm less than the hole you'll be drilling - that should account for the boundary losses, but the number of holes will increase.
This could sound really, really nice, as the port will in essence be damped by the resistive flow at the edges, and that could take the resonant Q of the port closer to .707 (critically damped), although some stuffing behind the 'portlets' will take the damping beyond .707, and will take the box into aperiodic behaviour.
So, get drilling a temporary removable side.
Start with the original calculated number, measure the FR, increment the number of holes, measure, and so on, until you find an FR that sounds great. This way you'll have some idea of the deviation from the model.
One thing is certain - the 'peak' of the port will be lower, but broader than a standard port, I'm guessing, you could end up with an ebs style response...
Owen
I use this online calculator with predictable results.
www.ajdesigner.com/phpvent/subwoofer_vent_port_equation_length_l.php
Make sure to use the correct end factor: vent has two flanged ends = 0.850
To play with dia and number of ports, use the "back" button on the browser to save re-entering all the data.
Not sure about an EBS response, but these go down to 30Hz, with the advantage of corner loading. Very firm bass, no boom, and less phase shift than other designs.
Fr (drver) is 28Hz. in 29 Litres. Holes are 5/8" on 18mm baffle. Smaller holes could be stuffed with cigarette filters -7.4mm dia, 15mm long. A bit more predictable than santa's whiskers!
Definitely the way to go if space is a premium.
Geoff.
www.ajdesigner.com/phpvent/subwoofer_vent_port_equation_length_l.php
Make sure to use the correct end factor: vent has two flanged ends = 0.850
To play with dia and number of ports, use the "back" button on the browser to save re-entering all the data.
Not sure about an EBS response, but these go down to 30Hz, with the advantage of corner loading. Very firm bass, no boom, and less phase shift than other designs.
Fr (drver) is 28Hz. in 29 Litres. Holes are 5/8" on 18mm baffle. Smaller holes could be stuffed with cigarette filters -7.4mm dia, 15mm long. A bit more predictable than santa's whiskers!
Definitely the way to go if space is a premium.
Geoff.
Attachments
Daverave
If you try to use a large number of small holes as ports you will need to use a larger cabinet volume than if you use one larger port. However having said that this is a system that is similar to one used previously by Goodmans and in a larger cabinet does produce good sound. However for your proposed cabinet size I believe you will find one vent the best.
If you have not done this before you might find it best to find the answer by picking a vent size - say 25mm - and then experimenting with different lengths of vent to get the best sound.
Don
If you try to use a large number of small holes as ports you will need to use a larger cabinet volume than if you use one larger port. However having said that this is a system that is similar to one used previously by Goodmans and in a larger cabinet does produce good sound. However for your proposed cabinet size I believe you will find one vent the best.
If you have not done this before you might find it best to find the answer by picking a vent size - say 25mm - and then experimenting with different lengths of vent to get the best sound.
Don
The cabinet volume must remain the same!
The port area is determined by the mass of air that is required to resonate against the 'spring' of the air in the cabinet.
So, 1 large port or 2 ports at 1/2 the area, 4 ports at 1/4 the area and so on.
The key differentiator will be that the smaller the 'ports', and the more numerous they are, the greater the mass of air that is affected by the boundary conditions, and not free to move as a piston in the port itself. This will damp the resonance of the port some-what, and lower the Q of the resonant peak. This can however be used to a great advantage as a 70 Hz port (normal at a reference level) could drop 3db by lowering the Q of the port by using many small 'portlets' - this will also extend the range of the port (up and down), so the bass will appear to go deeper, but as it is below reference, less likely to excite room modes (ie become muddy o boomy). This is the technique used in the Extended Bass Shelf (or EBS) alignment, and that is renouned for fast, tight, deep bass.
So, suck it and see. I think you'd like how it sounds.
Owen
I would like to see it this way. Starting with small signals.
Let's assume an ordinary design with one tube of 10 cm length, cross-sectional area 40 cm². Roughly, but not exactly, this system is equivalent with one that has two tubes of 10 cm length and 20 cm² area each. One could go on dividing the area into smaller tubes like this, but the length of the tube will remain the same (almost). For a reasonable number of tubes, the response of the speaker will remain the same, and the "mach number" (= the velocity of the air inside the port) will remain the same.
Three things will happen, though, when the tube is divided into several smaller tubes.
1. The Reynolds number will decrease, which will result in less risk for turbulence inside the tube, if each tube is more narrow. (remember there are many tubes, and air speed is the same, but there is less space to form vortices)
2. If the tubes gets very narrow, say less than a few centimeters in diameters, there will be an increased flow resistance due to the viscosity of the air. This will lower the Q of the port, which in turn leads to a demand for a larger box volume (and other changes).
3. The airflow outside the vent will be affected, and it is probably harder to make properly flanged port endings if they are many.
Bottom line here is that while it may be useful to divide large (100 cm² +) tubes in several smaller tubes, there is no reason to go for extremely narrow tubes.
And the narrower tubes should each be as long as the single tube, at least almost. The wall thickness will definitely be too short in almost all cases.
Let's assume an ordinary design with one tube of 10 cm length, cross-sectional area 40 cm². Roughly, but not exactly, this system is equivalent with one that has two tubes of 10 cm length and 20 cm² area each. One could go on dividing the area into smaller tubes like this, but the length of the tube will remain the same (almost). For a reasonable number of tubes, the response of the speaker will remain the same, and the "mach number" (= the velocity of the air inside the port) will remain the same.
Three things will happen, though, when the tube is divided into several smaller tubes.
1. The Reynolds number will decrease, which will result in less risk for turbulence inside the tube, if each tube is more narrow. (remember there are many tubes, and air speed is the same, but there is less space to form vortices)
2. If the tubes gets very narrow, say less than a few centimeters in diameters, there will be an increased flow resistance due to the viscosity of the air. This will lower the Q of the port, which in turn leads to a demand for a larger box volume (and other changes).
3. The airflow outside the vent will be affected, and it is probably harder to make properly flanged port endings if they are many.
Bottom line here is that while it may be useful to divide large (100 cm² +) tubes in several smaller tubes, there is no reason to go for extremely narrow tubes.
And the narrower tubes should each be as long as the single tube, at least almost. The wall thickness will definitely be too short in almost all cases.
Lots of little holes is a pretty traditional way of pushing a BR towards being an aperiodic box. I add a backing layer of a specific open cel foam on the inside right against the ports to push it all the way.
There is very little analytical/quantitative information out there to help get an aperiodic box right off the bat.
I will approach an aperiodic design one of 2 ways.
I will model a BR with a very specific target shape to the roll-off knee and then give the port a very high aspect ratio and then fine tune with foam inserted in the port.
Or
I will model a sealed box with Q = 1.1 or a bit higher and then make a WAG (guided by experience) to arrive at a number of small holes or a dynaco-like slot, and then tune with the thickness of the foam or density of the fiberglass (dyna port)
http://www.t-linespeakers.org/classics/dynaco.html
My last effort (4.5 litre for CSS FR125S) was nailed 1st time... plans are here http://planet10-hifi.com/boxes-CSS.html (7th entry down, v1.0 plans link in the text)
dave
There is very little analytical/quantitative information out there to help get an aperiodic box right off the bat.
I will approach an aperiodic design one of 2 ways.
I will model a BR with a very specific target shape to the roll-off knee and then give the port a very high aspect ratio and then fine tune with foam inserted in the port.
Or
I will model a sealed box with Q = 1.1 or a bit higher and then make a WAG (guided by experience) to arrive at a number of small holes or a dynaco-like slot, and then tune with the thickness of the foam or density of the fiberglass (dyna port)
http://www.t-linespeakers.org/classics/dynaco.html
My last effort (4.5 litre for CSS FR125S) was nailed 1st time... plans are here http://planet10-hifi.com/boxes-CSS.html (7th entry down, v1.0 plans link in the text)
dave
"There is very little analytical/quantitative information out there to help get an aperiodic box right off the bat."
Dave's spot on there. He knows I tried looking for it. So, it is an area worth investigating, and experimenting.
2 known models are the Goodmans ARA and the Dynacos. 2 very different approaches, so there's plenty of middle ground to explore.
Geoff.
Dave's spot on there. He knows I tried looking for it. So, it is an area worth investigating, and experimenting.
2 known models are the Goodmans ARA and the Dynacos. 2 very different approaches, so there's plenty of middle ground to explore.
Geoff.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.