You might be able to do some surgery on your box to convert to a tapered MLTL layout like @Brian Steele 's Boom Unit
https://www.diysubwoofers.org/projects/home/boomunit/
His Excel worksheets take the physical dimensions and generates the inputs to model it in hornresp, and this tends to be more accurate than winisd.
https://www.diysubwoofers.org/projects/home/boomunit/
His Excel worksheets take the physical dimensions and generates the inputs to model it in hornresp, and this tends to be more accurate than winisd.
Note the impedance either side of Fb is much higher than 4 ohms.
95 watts at 4 ohms requires ~19.5volts, 19.5 volts into 20ohms is only 19 watts.
The restriction you placed in the port's right angle turn reducing it's cross sectional area probably would lower Fb.
Better to leave it unrestricted, or equal in cross sectional area as in post #20.
Neither the port SPL maxima or the driver SPL minima are necessarily the box Fb, and as the port of the OP's cabinet is within a few centimeters of the driver, the distinction between the two won't be as clear as an excursion or impedance measurement.You just close mic(couple centimeters away) the driver and port separately. This assumes you have a test mic and software like REW.. which is free. That will produce graphs that may look something like this where the contributions of each are clearly visible.
Art
Thanks for clarifying.
With Eg = 18.97 volts the port outlet velocity is as shown below. Notice how the Strouhal number is less than 1 at maximum velocity.
/////////////////////////////////////////////////////////////////////
The Strouhal number St is a measure relating to the ratio of the acceleration caused by the unsteadiness of the air flow in a bass reflex port tube, and the convective acceleration caused by the non-uniformity of the flow at the port termination.
St is defined as the ratio of the cross-sectional port radius a and the particle displacement u / w, where u is the peak particle velocity and w = 2 * Pi * f.
St = a / (u / (2 * Pi * f))
If the particle displacement value is greater than the port radius value, in other words if the Strouhal number is less than one, then there will be an unsteady separation of the acoustic flow at the port end, and vortices associated with the separation of the flow will form.
/////////////////////////////////////////////////////////////////////
The formula provided refers to 'a' as being the 'cross-sectional port radius'. As the present bass reflex enclosure is using a slotted port, I would expect that this means that the 'a' value might refer to 1/2 the minimum of the width and height dimensions. Would it be correct to assume then that the "effective" Strouhal number is even lower than that which is computed by Hornresp? This would make the turbulence problem a bit more severe, and thus intruding at lower power levels. Added to the other intrinsic problems introduced by sharp corners, etc., things do look a trifle severe.
Last edited:
When asked the question "For a ported (bass reflex) loudspeaker enclosure, what is meant by 'convective acceleration caused by the non-uniformity of the flow at the port termination'?", Microsoft Copilot had this to say:
After asking "How would 90° corners in a slotted port affect performance of the port?", Microsoft Copilot provided the following response:
In a ported (bass reflex) loudspeaker, the term "convective acceleration caused by the non-uniformity of the flow at the port termination" refers to the aerodynamic effects that occur as air moves in and out of the port. Here's a breakdown:
- Convective Acceleration: This happens when the velocity of the airflow changes along its path. In the context of a loudspeaker port, as air is pushed or pulled through the port, its speed and direction can vary due to the shape and design of the port.
- Non-Uniformity of Flow: At the port termination (the ends of the port), the airflow is often not smooth or uniform. This can be due to sharp edges, abrupt changes in the port's geometry, or turbulence caused by the interaction of the airflow with the surrounding air.
- Impact on Sound: These non-uniformities can lead to turbulence and vortex shedding, which may generate unwanted noise (often referred to as "chuffing") and reduce the efficiency of the port. This is particularly noticeable at higher sound pressure levels.
Designers often address these issues by optimizing the port's geometry, such as rounding the edges or using flared ports, to minimize turbulence and ensure smoother airflow. This helps improve the overall performance of the loudspeaker.
After asking "How would 90° corners in a slotted port affect performance of the port?", Microsoft Copilot provided the following response:
90° corners in a slotted port can negatively impact the performance of the port due to the following reasons:
- Airflow Disruption: Sharp corners cause abrupt changes in the direction of airflow, leading to turbulence. This turbulence can result in unwanted noise, often referred to as "chuffing," and reduce the efficiency of the port.
- Energy Loss: The turbulence at the corners dissipates energy, which can lower the overall output of the loudspeaker and reduce its efficiency.
- Nonlinearities: Sharp corners can introduce nonlinearities in the airflow, which may cause distortion and affect the accuracy of sound reproduction.
- Port Tuning: The presence of 90-degree bends can slightly alter the tuning of the port, as the effective length of the port changes due to the geometry of the bend.
Mitigation Strategies
To address these issues, designers often:
- Use rounded corners or curved bends to smooth the airflow and reduce turbulence.
- Optimize the port's dimensions and geometry to balance airflow velocity and minimize energy loss.
- Employ computational fluid dynamics (CFD) simulations to analyze and refine the port design.
Not as far as the strict definition of St is concerned.
In the case of a rectangular cross-section the value of 'a' is calculated as (Width * Height / Pi) ^ 0.5.
Seems like a reasonable assumption to make, but I would not want to try to quantify an "effective" number.
In essence, the port tube is simply too long for the given cross-sectional area.
Is it possible to model a constant radius port in Hornresp? I'm building a box now with a slot port but am converting area of the round port to the slot (I used the Excel file provided in the thread about constant radius ports). Sorry if the information I'm asking for already exists in that thread.
Ports in hornresp are considered round (constant radius?) tubes.
If you refer to "constant radius" as a flare radius used to define "normalized flare rate" (by salvatti/devantier/button) this is not possible in hornresp.
I made an excel spreadsheet to calculate tuning such a port here.
For generating an optimized "normalized flare rate" port use my optimizer tool.
You can combine hornresp and the optimizer excel tool. Model the enclosure in hornresp, using a sensible port surface and then use tuning frequency and enclosure volume (and max desired spl) to create an optimized port.
A few 'optimum port' rules-of-thumb which have never let me down:
If the you cannot get anywhere near any of these, then either the design is flawed, it's an inappropriate drive unit, or, most commonly, you're trying to get too much bass extension - more is usually not more, and the main reason that most small ported speakers sound so dreadful...
- Port area should be as close to Sd as possible.
- Length should not exceed three times its diameter.
- Use one straight circular port.
- Airflow at both ends of the port should be equalised to avoid rectification.
- Generous flaring radii should be employed. (Around 1/3 to 1/2 the diameter).
- maximum velocity should not exceed 15m/s.
I would not second this. Mostly for modern smaller (~5") drivers the enclosure volume is usually small and a reasonably low tuning needs a long and/or narrow port. The alternative is either passive radiator(s) or an optimized port, see my tools above.
After my investigations I would also not confirm a straight port is optimum.
The optimum I found and that is reflected in the roozen and the salvatti et al papers has a mid port surface about half of the opening surfaces, not accounting for the required roundover.
Useful port area and maximum velocity are very much dependent on the maximum desired output level AND the tuning frequency. low tuning frequencies need much bigger port surfaces to avoid turbulence.
I suppose the max 15 m/s velocity (or 5% velocity of sound) rule may come from a very common tuning frequency of around 40 Hz.
@hepi34
Anyways, best of luck. Either way, you know what, why, how. The decision is yours. It is not so uncommon to scrap a design under development, or needing to iterate it once more. In all pain, it is better to walk through the mistakes instead of luckily nailing it for the first time. Only then you really know.
Most of the time, it works, until it doesn´t. I have witnessed the issue dew times with both WinISD and Hornresp. It just didn´t do what was projected. So it is best to verify in the development.
Well, you don´t have to cut the whole port ideally. Arguably, if it is just a tuning issue, and you are undertuned, just cut the port in a way that allows. Just make a wide "wedge" on top of the vertical part to make it effectively shorter.
This one would be much more tedious and difficult, but in light of "scrappable design" anyways, you might give it a go. Just draw the line inside, making the corner dimension you would like to round, and drill it out. One hole after another, till you done. Then you can implement new, better, more rounded corner. This part I believe is "crucial" for getiing better performance. The inner corner really needs to be round.
Anyways, best of luck. Either way, you know what, why, how. The decision is yours. It is not so uncommon to scrap a design under development, or needing to iterate it once more. In all pain, it is better to walk through the mistakes instead of luckily nailing it for the first time. Only then you really know.
I have a question about #1. What happens if the port area is lower (half of sd)? Or port area is bigger (double sd)? [note that I pulled half and double out of thin air]
Bigger ports generally have lower losses, so the port output peak is usually higher (can be simulated in hornresp - try out very small ports!).
Big ports also transmit more higher frequency content, mostly port and enclosure resonance peaks.
Bigger ports are less critical regarding roundovers due to their lower air speed.
Also, bigger pors need to be (much) longer, as the relation of port surface and length is roughly constant. That means: double port diameter -> quadruple length.
The rule of thumb ("port surface similar to Sd") neglects the influence of Xmax and tuning frequency.
Both increased Xmax and decreased tuning frequency increase the quantity of displaced air in the port and therefore need a bigger port ( that is if the driver will be driven to its Xmax).
When port area is too small the port output lags in amplitude compared to the driver and port noises occur at lower drive levels.
The larger a port is the longer it must be for the same tuning frequency, an exceedingly large port would add unnecessary volume to the enclosure.
Usually, >1/3 of Sd is a good starting point.
You could build a folded transmission line where the whole enclosure is the port.
Positive flare = efficient + large enclosure.
Straight flare = ease of construction. No angles.
Negative flare = low tune + small enclosure.
Hofmann's Iron Law = efficient, low, small...pick 2.
You could build a folded transmission line where the whole enclosure is the port.
Positive flare = efficient + large enclosure.
Straight flare = ease of construction. No angles.
Negative flare = low tune + small enclosure.
Hofmann's Iron Law = efficient, low, small...pick 2.
Okay, 1/3 sd makes some more sense to me. I don't think I'd seen port area = sd as a recommendation before
If you're designing with port area=Sd in mind, you may as well design it as a TL, because that port is going to be so long that it's likely going to introduce bandwidth-limiting resonances into the overall response.
In fact, if you're going to use a large, long port, you may as well design the box as a tapered TL or MLTL because it will allow you to determine where those resonances appear, and how to control them with careful driver placement along the path.
In fact, if you're going to use a large, long port, you may as well design the box as a tapered TL or MLTL because it will allow you to determine where those resonances appear, and how to control them with careful driver placement along the path.