silly ? on port's

[rolfy72]

We all know that when we "port" a enclosure for extended bass response we tune to a particular frequency...
Why cant we tune for several frequencys(ie, 2-3 ports of varying lengths)
Will this cause cancellation problems?
Has it been tried yet?(stupid ? but U never know...)
Am I just being a silly bugga?! =)

[pinkmouse]

Yes, it is possible, it's called staggered tuning, but I don't know of any software to simulate it!

In the past, I have used one port as per standard, and another lower tuned port to extend bass downward, but it was not a simple project. It involved about two weeks of altering both ports, with much measuring and listening in between. It did extend bass downwards to a degree in the end, but for the amount of work was probably not worth it. But it was good fun trying!

[SY]

What you'll find is that if the two ports port the same enclosed volume, you still end up with a single resonance. Since the air masses in the ports are coupled through the same "spring" (i.e., internal volume), you have something that looks like a single port with an air mass equal to the sum of the air masses in the two ports (plus the end correction factors). They're virtually in parallel.

Doing what you suggest may in some circumstances be helpful, like if you need to replace a single port whose geometry leads to a noticeable pipe resonance. But that's a pretty special case, and normally more easily dealt with by other means.

[pinkmouse]

What you'll find is that if the two ports port the same enclosed volume, you still end up with a single resonance. Since the air masses in the ports are coupled through the same "spring" (i.e., internal volume), you have something that looks like a single port with an air mass equal to the sum of the air masses in the two ports (plus the end correction factors). They're virtually in parallel.

Theoretically I have to agree with you totally, SY. But for some reason it didn't seem to work quite that way in practice.

But then again, this was about 8 years ago, measurements were done with a cheap signal generator and RS SPL meter, and my memory is vague at the best of times

Perhaps I will have to try this again sometime, and see if I can duplicate my results!

[SY]

Well, I'm using an IMP system, which isn't exactly the last word in resolution. But in the few dual port systems I've tested, I still have only seen the normal two impedance peaks. So, to at least the twelve bits and .1 Hz or so that I can trust my measurement tool...

There was an article in Speaker Builder maybe 10 years ago on using equivalent circuits to predict acoustic response using pSpice. If I can find that issue, I'll post the cite and we can at least do a theoretical look at this matter- unfortunately, I have no dual port systems in my living room at the moment to repeat the experiment- and try to derive some design equations.

[roddyama]

Sy,

I’m not as sharp on circuits as maybe I should be for this conversation, but in the equivalent circuit of a ported box speaker, the ports air mass is in // with the box compliance. If we were to add another port with a different air mass, it would have to be in // as well. Electrically, we would place a resistor between the two // (port) arms of the circuit to isolate them. This would allow them to exhibit different characteristics with the resistor becoming an integral part of the result, whatever that maybe. I don’t see how we would implement the resistor in the conversion from the electrical model back to the real speaker. Are on the Aperiodic design discussion again?

Rodd Yamashita

[SY]

Rodd, your understanding seems to be just fine. However, there's really no (or negligable) resistance if the two ports are firing into the same enclosure. Sure, you can divide the enclosure and couple the two volumes with a port or some resistance, but now we're talking about a VERY different animal. And that animal is very well characterized by several software programs and is the basis for a few million commercial speakers.

[pinkmouse]

Rodd,

I think you have put your finger on the problem in the real world as well, there should be no difference between the loading of two ports and the loading of one bigger port. I'm confused. It worked when I tried it, but now I'm thinking about it properly I can see no reason why it did!

Nope, nothing to do with aperiodic, yet...

[phase_accurate]

In the latest issue of the German mag "Klang & Ton" there is a proposal for an active system comprising of an MSW and a GIA woofer (sporting a cone made of a thing called "Chitin" in German, i.e. the material that insect's "armours" are made of).
They use a reflex enclosure on this woofer having four ports of different lenght.
They claim that this arrangement had a single resonance with a broader peak than usual.
If that's true it looks as if the resonance has a lower Q. So the effect could be the same as a "lossy port".

I didn't think about this one when I answered in the lossy port thread (one only gets older !)

Regards

Charles

[roddyama]

Originally posted by SY
However, there's really no (or negligable) resistance if the two ports are firing into the same enclosure.

In the realized speaker. My point exactly.

My question is, How can this "resistor" be implemented to work on only one of multiple ports? My thoughts keep going to the 4 port example from Charles. The "Lossy" port system.

Al, are we there yet?

Rodd Yamashita