Hmm, seems we need a refresher course in pi space loading. 😉
If some of us are actually 'obsessed', it's due to the preponderance of folks over time wanting to know why their sims and/or builds don't perform up to par based on a wide range of differing 'opinions' and/or misunderstandings of the physics of the situation.
You can try loading the measured frequency response into your XSim simulation, see how well the measured response matched your simulation curve. If it does not match well, most likely your mod delay is not set correctly. From your posting, it seems they don't quite match up.
Edit : On closer look, it seems you did not even adjust the mod delay of the woofer in your simulation.
Edit : On closer look, it seems you did not even adjust the mod delay of the woofer in your simulation.
Thanks, Richard! If I have a 3 or 5 ohm resistor I can add that, right? I'm just double-checking there is nothing specific about your recommendation for adding a two-ohm. No matter how much I now know, I don't know what I don't know ;P
Hi James,
I think this is my first time hearing "mod delay".... It looks like adjusting the distance of the speaker in XSim so that the center of each driver is the same distance away from the point of listening? Like if the magnet of the woofer sits 3 inches further away and 5 inches down you'd find that hypotenuse to the "microphone" kind of idea? Or... am I off....
Thanks for the link, GM. It looks like another article to add to the "read now, fully understand later" pile ;P With the knowledge, there are still decisions to make about which pi space loading to design around, and the best approach to designing around it, right? The article is info and not answer... or is there (for the first time in diy audio 😀 ) a consensus on best practice? For example, some manufacturer's specify a distance down to the inch/cm for how close to be to a wall on whatever side. In that case some pi space loading is utilized and considered when designing. Other cases you want to treat the room to "ignore" as many walls as possible and have the speaker pulled out as far as possible, etc.Hmm, seems we need a refresher course in pi space loading. 😉
If some of us are actually 'obsessed', it's due to the preponderance of folks over time wanting to know why their sims and/or builds don't perform up to par based on a wide range of differing 'opinions' and/or misunderstandings of the physics of the situation.
Correct me if I'm wrong, but your main point is the importance of awareness and consideration of these factors while designing? Not particularly that x way is best?
If you are serious about building speaker, you should start googling the net or YouTube on the proper way of speaker simulation using XSim. For XSim, when you input the FRD and ZMA into the speaker property, you will see a "mod delay" box for you to key in the value.
People tend to test speakers on a small table up 1.5 to 2 m off the floor, in the middle of an anechoic room, outside in the middle of a field, or use gating of software to kill the room reflections. This open position leads to a 6 db bass loss as the speaker drives the air behind it, and the air under it to the floor. The test microphone or your ears do not hear bass behind the speaker or from your navel to the floor. Baffle step compensation corrects for this. 3 or 6 db bass boost. IMHO for a speaker capable of 100 W or more, corrected passively in the crossover, this imposes a $25 charge for each speaker for a decent bass booster. Maybe $50 each for a capacitor and an air core inductor.
I intend to test my speakers backed up to the cinderblock wall of my garage, projecting on to a 2 acre area. Similarly reflective as the 1' thick plaster wall I have in my listening room. I will put speakers up on a stand as highs in my listening room need to project over the organs, pianos, and furniture. So I will have a -3 db loss of half space, not 6 db loss of open space, called by techies "4 pi space". If you put the speakers on the floor against a hard wall you get quarter space and the best bass you can get without buying bass boost components. But the highs run right into your couch or tables.
Since you rolled your speaker response off at 100 hz, I presume you intend to handle the bass with a subwoofer, or not have any. Hence no $50-$100 passive baffle step compensation parts. DSP are all the rage now for tweaking response curve between source & power amplifier. The cost of DSP is that cellphone & PC op systems change yearly, and after a certain time, 3-6 years probably, drivers for your DSP are no longer available. Then if you want to change anything you have to buy another DSP.
If I design a speaker for my summer trailer, I'll have to do something about boosting bass, as those walls are 1/8" wood panel backed up by aluminum siding. Ie, not very reflective. Cheaper than baffle step compensation crossover components, and more long lived than DSP, is the humble analog graphic equalizer. Available everywhere used under $100 for a stereo unit are they are way out of fashion. Pro touring bands use DSP now. They don't mind replacing everything every 6 years, the popularity of the band is shorter than that.
If you live in an apartment, cheaper than baffle step compensation is 2 sheets of 1/2" drywall installed behind your speaker. Even better, granite sheet or masonry block. With the speaker backed up to that wall. Additional benefit, bass is reflected into your room at your ears, instead of leaking through the wall to annoy the neighbors.
The advantage of subwoofers, they can sit on the floor against the wall, and get the quarter space bass boost without having to worry about the loss of any highs in the furniture.
Last edited:
You're welcome!
Correct, so in answer I can only say that one should design based on the needs of the app and since *basic speaker box design is for a bounded space its acoustics dominates be it a tiny box or outdoors.
*Note that T/S theory peters out at the driver's effective upper [Fhm], lower [Flc] mass corners, so any bandwidth response plot beyond these points are strictly due to [inputted] inductance [mH] and/or mass roll off, with on-line freeware generally just flatlining it:
Fhm = 2*Fs/Qts' end of acceleration BW
Flc = Fs*Qts'/2 end of roll off BW [normally never used]
Qts' = 2*Fs/Fhm
Maybe more easily understood from a horn loading POV along with all the extra math, pg. 7: http://www.xlrtechs.com/dbkeele.com...Preprint) - LF Horn Design Using TS Paras.pdf