Port close to the wall--effect on tuning?

[diesel_88]

I understand the rule of thumb is to locate a port at least one port diameter from all cabinet walls (side and back). What are the correction factors when you have to get closer than that?

Came across this useful diagram, but couldn't quite follow how the coefficients apply to enclosure tuning calculations. If someone could explain it a little more clearly, I'd be much obliged:

http://www.diyaudio.com/forums/showt...24#post1148924

Also, how do I correct the calculations when I have to run a port closer than one diameter to the back wall of the cabinet?

[richie00boy]

It's not so much you have to apply a correction factor, it's that the port becomes strangled thus not acting fully as it should.

[diesel_88]

I'm assuming (perhaps incorrectly) that I don't ever want to get any closer to the back wall than one-quarter the port diameter--that would definitely be strangled. Even then, I'd have to leave a zone well-clear of damping material, round all edges, etc.

But in between one diameter and one-quarter diameter, what are the effects on box tuning and performance (theoretical, empirical, or anecdotal)? Likewise for a port that runs really close to the side wall (or corner) of a box.

[Svante]

Originally posted by diesel_88

But in between one diameter and one-quarter diameter, what are the effects on box tuning and performance (theoretical, empirical, or anecdotal)? Likewise for a port that runs really close to the side wall (or corner) of a box.

The effect is that the tuning gets lower the more the opening of the tube is obstructed. The obstruction acts as a lengthening of the tube. The amount of lowering of the resonance frequency depends on the degree of obstruction in relation to the length of the tube. A long tube needs a tighter obstruction for the same effect on the tuning frequency.

Apart from this effect, there will be increased turbulence at the inner end if the opening is too close to the back wall. 1/4 of the diameter seems too close, intuitively, but I have no hard data to support this.