Flared ports and turbulence

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks for that Dave.

I had heard about ports "pumping up" where one end was unflanged but had not considered that a different flare each end could exhibit the same effect.

Those articles certainly sound like worthwhile reading. I'll grab myself a copy.

With regard to the idea of using a voltage measurement, I've not had a lot of luck. As I said earlier in this thread, the power as calculated from the voltage across the drivers did not equal the power determined graphically in WinISD using the measured SPL.

I ran this past the WinISD people and they suggested that the variations in impedance (vs frequency) made this voltage method difficult. I imagine that the T/S characteristics of the driver would need to be accurately determined also.

With this in mind, I'm satisfied that the SPL method of determining power and thus port velocity is the best approach.


I agree that nearfield measuring gets rid of room effects, but that only gives me the output from the driver.
I needed to measure the contribution from both the drivers and the port in order to get the SPL of the total ported system.

Ideally this would be done outside. I used the in-room approach because the room could add together the outputs from a front mounted port with a pair of side mounted drivers.

Had I used a single driver (front mounted), I would not have been able to move enough air to test the larger flares.
No doubt commercial researchers would use a single, more expensive driver.

I could have front mounted both drivers and the port but the box would have been much larger. This would have lead to a shorter port, which would have been more flare than port, possibly invalidating the results.

Such are the compomises we have to make as DIY'ers.

Anyway, the next stage is to have a look at how 100mm pipe compares. I'll have a look at those articles first though...

Collo
 
I have downloaded the AES article on Maximising Performance from Loudspeaker Ports (preprint 4855) as you have suggested.

It was written by Audio Industry professionals and I found it to be full of interesting tests and recommendations.

Much of the testing was done with ports that were tapered as well as having radiused ends. Such a construction is a bit outside the realms of "DIY", but the info was still useful. I think that the B&W "flowport" is tapered and now understand why!

One example was that they recommended that the radius on the ends of the port be a minimum of 20% of the port diameter. This would rule out using a small router to cut the rollover unless airspeed is kept below 10m/sec.

Thanks again for the list of AES articles - I've added a page on my site, (along with a link back to diyAudio of course) to keep the info alive.

Collo
 
Hi Coolin,

the article makes the distinction between a "flared port" and a "radiused end"

What I have been calling a "flared" port throughout this discussion, is where the port is essentially a straight port with an end that "bells out". This is the same type of port that "Precison Ports" in the US sells, and is within the ability of an ambitious diy'er to make.
The article refers to this as a "radiused end". The only recomendation they make is that of having a minimum radius of 20% of the port diameter is sufficient to prevent noise generated when air enters the port (from either direction)

The majority of the article covers the style of port where the CSA of the port varies thoughout the entire length of the port. It is this style that they refer to as a "flared" port, and I would call "tapered"
They cover compression effects in detail for different flares and conclude that the best all-round design is to have a flare radius that is equal to the length of the port
 

Attachments

  • aes-optimal-port.gif
    aes-optimal-port.gif
    2.2 KB · Views: 313
That wouldnt be extremely difficult for the DIYer to make as a slot port, but as a round port, I wouldnt try it.

About the Haliaetus site.... They claim a low frequency directional speaker yet all I'm seeing is a ported sub with a Giant flare nozzle for a port.... Am I missing something?
 
They only have the port on one side which gives greater output on that side thus causing the imballance to the figure 8 SPL chart.

Oh stop it.... Just because one side has greater output than the other doesnt mean it will be directional. The outputs are still in phase and the wavelengths are still so long that there will be omnidirectional radiation. If what you said were true then any ported box with the port on the backside would be directional. If their speaker is in fact directional(which I wont believe until I hear a reasonable explination) then they've got something else going on.
 
I guess you havent read the info that has been provided. Its not THAT directional. Its still a dipole just more radiation on the front than the back.

The spacing beween front of port and driver is also quite large, i'm guesing 3-4 feet.

The cancelation is because the two are out of phase. If i'm not mistaken regular ports are also "mostly" out of phase compared to the driver output.

There are also others that do it this way with two drivers that are adjustable. Even a commercial product but i dont remember which one.

What i do wonder about is fow effective this will work when placed in a real room. The speaker will be near a wall and this will distort the cancelation abilities.
 
Hi!

For the directivity, it is easy to understand. Let's refer to microphones: you can turn any kind of mic into any king of mic!
Take an omni capsule the pattern is O, do some pinch holes on the back of the housing and you have a cardio caps!
To have a figure 8 pattern, do a hole of exactly the same area than the actual caps. you can continuously shift from an O to a 8 including any kind cardio (super, hyper….).
The reciprocal is true, with a cardio you can do an omni, just by sealing the holes.
The pattern is due to the balance between the pressure that reach the front and the rear of the capsule’s diaphragm at the same time (you can add some time delay between the two sides too but it is not our aim here).
The same for the BR boxes, under the Fb the box has a funny directivity depending on the mutual coupling of the driver(s) and the port(s).
So BassAwdyo, yes any BR box as a dipole pattern directivity (or cardio or 8 depending upon the area of the port…) if the driver and the port are on opposite sides. I think that Jean Kergomard issued a paper in the JASA I think about that! The trick: for a midrange driver tuning (stupid habit but…) in a 3ways the point is to tune it high and put the port exactly on the back of the unit to improve room coupling ….
The nozzle use this in a special way, see AES pdf and my former mails.

@+
Maiky
 
Updated Results

Hi again!

I finally finished testing with three different diameter ports to see how the useable velocity might change.

Whilst I was at it, I re-tested the five 86mm ports using voltage measurement to determine input power as suggested earlier in this saga by RonE.

The tests with DVD sequences and SPL meter were abandoned.

This gives more accurate determination of input power and eliminates possible errors from:
-SPL meter adjustments
-Room nodes and meter placement issues
-Conversion of WinISD predicted SPL for "half-space" into a value for the test environment.

The re-testing has given a slightly different set of results to those that were published earlier which were based on SPL readings

The new comparison between ports of three different diameters has shown that larger ports are able to use higher velocity

Testing was done to detect just audible "chuffing" close to the port. At the seating position an extra 2dB is required for it to be just audible with a sine wave test. For normal content, this rises to 3dB, which is equivalent to an increase in port velocity of around 40%. I have factored this in to my recommendations so that ports are no larger than they need to be in a real world situation

An all-encompasing equation has been developed, and for those who don't like maths, I have supplied a downloadable calculator.

All the info has been written up on the flare testing page, along with a link to download the calculator.
http://www.users.bigpond.com/bcolliso/flare-testing.htm


Attached is a new table of maximum design velocities (in metres per second) for common port and flare combinations

as before,
all constructive criticism is welcome

Collo
 

Attachments

  • new-table.gif
    new-table.gif
    14.5 KB · Views: 225
Thanks for those comments guys!

Paul, that velocity is correct.

There is of course the principle that the unflared section of the port also has a speed limit. :irked:


I suspect that larger ports have a higher allowance than smaller ones, but have not been able to find any definitive literature on it.

The closest I have come is a statement in the AES paper "Maximising Performance in Loudspeaker ports" where they
indicate that heavily flared ports suffer from more compression, but they don't expand this into a general principle.

I suspect that a minimally flared port runs into trouble before the unflared section has a chance to get play up. Adding a flare allows you to lift the speed to the point where turbulence and compression in the straight section become an issue.

It would be nice to know what velocity this occurs at for different diameter pipes. Perhaps yet another round of building test ports.

I have looked at most of the commercial flares and all seem to be 30mm radius or less.

Consider the monster 6 inch Aeroports:
http://2loud4you.ca/shop/index.php?...s_id=3&zenid=b915fc78891bb55ba2461d9f56376034

The unflared section is 6 inch diameter and the outside of the flare is 9 inch. Subtracting 1/4 inch all round for the mounting flange and dividing by two gives a flare radius of 1-1/4 inch (ie 30mm)

The precision ports and lightning ports are similar.

I suspect that they are not available in larger radii because of the afforementioned problems.

In the meantime, I'll stick with the notion that anything over 35 m/sec is questionable.


I also would have liked to have tested a 6 inch port to check the accuracy of my predictions, but the test box just won't push enough air!

Maybe I should add a "Warnings and limitations" paragraph to my notes! :att'n:

regards
Collo
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.