I come with data-
Last year @bikinpunk popped one of my cabinets on his measurement stand for his Klippel Near Field Scanner.
In short, this automated measurement machine gives anechoic level measurements with a frequency resolution of ~2Hz from 20Hz to past 20KHz, and take measurements in a full 360 degree arc, and uses math to give the readings at 2 metres.
Here is a 6.5" mid-woofer with a passive radiator, in a typical 2 way cabinet
Cabinet dimensions:
21cm W x 40cm H x 35cm D
Here's the polar measurement:
Zoomed in:
Now changing the axis to focus on what happens if the woofer was side mounted.
That means + 90 degrees would be facing the front (listener)
-90 degrees would be the rear (wall)
Viewing up to 300Hz:
Now, unfortunately now viewable to you, but moving my mouse cursor around- here is the 90 degree -1dB point- 147Hz
The 90 degree -2dB point: 172Hz
The 90 degree -3dB point: 290Hz
So my 1 line conclusion now:
Bass is omnidirectional up to 120Hz (confirms THX / Dolby standard for LFE channels)
It is best to crossover a side mounted woofer below 150Hz. This is consistent with real world findings of designers who made speaker with woofer around 200Hz who, when the woofers were turned to the side, felt there wasn't any appreciable difference. (eg. the late Jeff Bagby)
Based on my data, I now wouldn't cross a side firing woofer at 300-400Hz eg. KEF Blade 1/2.
Last edited:
The work around for WAF is to build the box and say you're just trying something out temporarily. I did this 15 years ago and the speakers have only gotten bigger since. I always switch out the speakers when no one else is home.
Like KEF would not know this. That is why they use woofers on both sides. Not only bringing force canceling but +6dB on axis of the speaker too, while off axis the sum is lower. Combined with the right encloure shape, woofer placement and crossover point the response would be pretty all right.
Here is some more data on a prototype of a twin dual opposed cabinet. (2 x 8" woofers on opposite sides of the cabinet- 4 in total)
Hor 0 is the blue line; Hor 180 is the red line:
If we rotate the cabinet 90 degrees we see the two other 8" woofers, and the cables on the back of the cabinet.
When the mic is placed here, and the cabinet is in this position, it is thus Hor 90-
Hor 0 to -90 in 10 degree steps:
Here are measurements in 30 degree steps, from 0 to 360 degrees.
Exported for VituixCAD2 for analysis:
Horizontal Directivity- normalized
Horizontal polar map; normalized:
Their challenge is to making the right compromises that will adhere to their goals eg. Single Apparent Source ideology, not running out of excursion for the Uni-Q LF element at the bottom of its passband near 400Hz… as well as making an attractive speaker mass production, AND will find customers.
On all these fronts I feel that the company made good decisions.
Crossing over around 200Hz would create other challenges that would require other solutions.
Hor 0 is the blue line; Hor 180 is the red line:
If we rotate the cabinet 90 degrees we see the two other 8" woofers, and the cables on the back of the cabinet.
When the mic is placed here, and the cabinet is in this position, it is thus Hor 90-
Hor 0 to -90 in 10 degree steps:
Here are measurements in 30 degree steps, from 0 to 360 degrees.
Exported for VituixCAD2 for analysis:
Horizontal Directivity- normalized
Horizontal polar map; normalized:
The engineers at KEF “knowing it” is not the difficult part.
Their challenge is to making the right compromises that will adhere to their goals eg. Single Apparent Source ideology, not running out of excursion for the Uni-Q LF element at the bottom of its passband near 400Hz… as well as making an attractive speaker mass production, AND will find customers.
On all these fronts I feel that the company made good decisions.
Crossing over around 200Hz would create other challenges that would require other solutions.
Attachments
Last edited:
Not only made good decisions, they were quite open about it. Imho a lot of others could follow the example, albeit KEF too shows only the merits, not the shortcomings.
@tktran303 some images are missing, or am I the only one not seeing response plots?
@tktran303 some images are missing, or am I the only one not seeing response plots?
Last edited:
decided to go front firing after all thanks for the input everyone
My driver choice now is
BMS 18n862
Ae td8m
I still have the tpl200h but now im wondering if I need the sensitivity of the H version
Bms is 95db
Td8m is 94.7db
Is there any benefit or cons to the H version over a normal tpl 150? To match to a 8” driver at roughly 2k
Thank
My driver choice now is
BMS 18n862
Ae td8m
I still have the tpl200h but now im wondering if I need the sensitivity of the H version
Bms is 95db
Td8m is 94.7db
Is there any benefit or cons to the H version over a normal tpl 150? To match to a 8” driver at roughly 2k
Thank
Sounds like an excellent project.
With DSP and multi amps no need to worry about the H version, in fact it might help off access response.
The BMS is also excellent , although I've never heard one or even know where I can get one in the USA. I have the TD15S for many years in a 160ltr BR tuned to 30 hz with very good results. I would consider a BMS as an upgrade to the TD15S but mostly for output and durability at the extreems.
The 8 or 10 would work well to the TPL's 1500 hz low end.
Good luck.
With DSP and multi amps no need to worry about the H version, in fact it might help off access response.
The BMS is also excellent , although I've never heard one or even know where I can get one in the USA. I have the TD15S for many years in a 160ltr BR tuned to 30 hz with very good results. I would consider a BMS as an upgrade to the TD15S but mostly for output and durability at the extreems.
The 8 or 10 would work well to the TPL's 1500 hz low end.
Good luck.
Hi Jonone
I just finished a large cabinet with a 15 inch woofer. I made cabinet tall (to get large volume) but the height was close to 1 metre which caused problems of cabinet resonance around 200hz. So if you can use smaller sealed cabinet you can avoid these issues. And if you plan to xo 15h to 10TDM in 200 to 300 hz region then it's important !.
I originally planned a ported box but had to stuff it and seal it to fix the resonance problem. I hope that helps a bit
Or simply factor that potential issue into it and sort it out in the design phase. It's pretty basic physics.
Good point. I have similar problems with my bass bins. Not sure if building them without parallel sides would help. Here's a handy tool: http://www.mh-audio.nl/Calculators/standingwaveinbox.html
After years of tweaking, I cross below 300hz and today I'm crossing at 100hz to the JBL 2206.
Hi the boxes will be sealed but around the height you said with probably a 250hz xo
I saw trols mention this in his damping guide
http://www.troelsgravesen.dk/cabinet-damping.htm
I would not rely blindly on T.G. In this department. I suggest looking for proof in his measurements (BD plots close range of the woofer in the affected area will do fine) Damping of a closed system is quite easy in fact: just fill’er up with acoustic foam or mineral wool 20-40kg/cubic meter.
Actually, I've just pulled half the stuffing back out of the enclosures and unsealed the ports. Tuning is at 40 hz. There is no problem with 200 hz resonance now. It may also be due to a lower Xo point of 220Hz, 2nd order BW only for W-M. I like how it sounds now. Sealed was for my tastes just too damped , even with active EQ. Now.it sounds more dynamic and I don't necessarily need my Sub for music now. So.... it can be done !
Did you do a close range burst decay plot of the woofer and the port when using the crossover? That shows problematic resonances quite easily (although the opinions about what is perceivable are spread widely…). I would try to counter any resonance that shows up 6-8 periods at -20 to -25dB.
I never thought about using the decay plots in REW. I still have a lot to learn about that program and it takes time and practice. Also the help from other people out there is much appreciated. I did put the mic right at the mouth of the port to get the port output peak spl, it is right at 40hz . I will want to read about the spectral decay measurements now🙂
This one shows the primary mode in an enclosure some more than a foot high internal.
The damping could obviously be bettered, but the level is quite low (below -25dB), so here I decided to give it a go. It hardly shows in the level/frequency response, but it’s there.
Same here, all I see is the photo and normalized horizontal directivity plot. All the others just show up as a file nae.