[Trevor]

On 21/01/2015 6:48 AM, Scott Dorsey wrote:
Les Cargill wrote:
In this way, you can completely dial in the ideal port length to compensate
for any errors in box volume calculation as well as variances in the drivers.

I suspect this is outside the envelope of measurement.

In that case, it's either not important, or else you need better measurement
hardware in order to a proper design job.

Between the 1x ( 0.525 cu ft ) box and the 2x ( 1.05 cu ft ) box, there
is one hump in the group delay that goes from 4ish to 8ish msec, around
75-80 Hz in the frequency response.

Doubling the box nearly has to be more of a perturbation than 1.25 mm of
port depth.

But then again, ports are critically sensitive elements. Still, I am
skeptical that "just take it outside and measure it" is going to
allow enough accuracy to be able to tell.

That's entirely possible. The problem is that the simulation has even less
accuracy in most cases. So either you need better measurement or better
simulation, or else you need to not worry so much and live with the error.

I will say that if you don't have good measurement, you are more likely to
get a design that meets parameters on the first try with a sealed box design
than a ported design.

It's pretty easy for anybody with computer sound to make an impedance
sweep these days, free software readily available, and that's all you
need to tune the box. Not to mention you can also measure the driver for
T/S parameters before you design the box.
But yes, a sealed box is always easier to design and construct if that's
what you want.

Trevor.