On paper experiment of a Bessel line array

[voodooless]

From what I can see, it's really that much worse than a dome tweeter:
SB Acoustics SB29SDAC-C000-4:

Tectonic TEBM46C20N-4B:

Note that the scale of the SB tweeter is a bit larger. It really doesn't look too bad.

 

[thewas]

Yes, its not too bad on absolute scale but personally I like to have it for wide radiating loudspeakers so I tend to combine it with a small (19mm) dome or ribbon tweeter. Also a good waveguide actually can make the directivity of a tweeter wider wider in the upper octave.

 

[voodooless]

Also a good waveguide actually can make the directivity of a tweeter wider wider in the upper octave.

the idea would be to rear mount them on the baffle so I can add a small waveguide made from the baffle material by adding a chamfer.

 

[BenB]

the idea would be to rear mount them on the baffle so I can add a small waveguide made from the baffle material by adding a chamfer.

Wont the waveguide force you to use a larger spacing between the drivers? The bigger the spacing, the more problematic the interference is at higher frequencies. When you run your sims, make sure use a listening distance that's realistic. I've seen others comment that the bessel arrays need distance to get everything to sum correctly, so it's not just me.

 

[voodooless]

Wont the waveguide force you to use a larger spacing between the drivers? The bigger the spacing, the more problematic the interference is at higher frequencies.

I’ll do some sims to see the influence. I can always make the spacing smaller/waveguide shorter.

When you run your sims, make sure use a listening distance that's realistic. I've seen others comment that the bessel arrays need distance to get everything to sum correctly, so it's not just me.

That could be. Actually I have no idea what the distance is in Boxsim, or if you can change it I would guess you’d need at least 3 or 4 times the length of the array as decent distance. Is 2m should be fine, which is roughly the distance that I would use the thing.

 

[bigjacko]

I am wondering is it possible to get many divers in series and still able to power by normal amp. From what I have know 2 driver in series does not add any sensitivity where parallel will add 6 dB sensitivity. Then we can add 5 drivers in series to get 20 ohm, it will still be driverable for the amp and sensitivity should not be lower than single driver. Is what I said correct?

 

[voodooless]

I am wondering is it possible to get many divers in series and still able to power by normal amp. From what I have know 2 driver in series does not add any sensitivity where parallel will add 6 dB sensitivity. Then we can add 5 drivers in series to get 20 ohm, it will still be driverable for the amp and sensitivity should not be lower than single driver. Is what I said correct?

In this setup, you can't have all drivers in series. Not sure what you are after?

 

[bigjacko]

In this setup, you can't have all drivers in series. Not sure what you are after?

I was thinking why you guys say 5 driver is hard to do crossover. The impedance and sensitivity does not have issue to me when you connect 5 drivers up. Or maybe because Bessel array will need a specific way to do crossover so need weird imedance for the drivers?

 

[voodooless]

I was thinking why you guys say 5 driver is hard to do crossover. The impedance and sensitivity does not have issue to me when you connect 5 drivers up. Or maybe because Bessel array will need a specific way to do crossover so need weird imedance for the drivers?

It's not hard because of impedance. It would be just as hard with a DSP. It's the interactions of the array that makes it hard to cross to a single tweeter.

 

[voodooless]

So this is the first attempt to put the array in VCAD. I'm sill trying to figure everything out I just traced the SPL and added the array setup:

This is at 2 meters distance. So now watch what happens at 1m:

So there is no proper merging of the driver units yet, and stuff get's messy. That basically rules this out as back channels in my case, because they will be 1 to 1.5 meters away. And even at 1.5m, it looks bad. Too bad really.

Funny thing: the suckout at 1m is exactly where the driver has a dip as well, which is usually measured at 1m distance as well.. Since these drivers are not pistonic they rely on resonances for HF dispersion, so could be that at other distances, the response would look different for a single driver as well. Interestingly the official datasheet that does not show the dip eludes to something:

It says: "corrected to 1W/1m". My guess would be that it was actually measured at 2m or more. If true, that means that these drivers are basically not usable for nearfield listening if used full-range, same as the Bessel array.

 

[ppataki]

I am wondering if anyone here actually tried building a Bessel Array using full-range drivers?
One of my friends has a larger living room and we are thinking about a 5-speaker Bessel using Markaudio CHR-120 drivers
Any suggestions/experience would be appreciated
Thank you

 

[voodooless]

I am wondering if anyone here actually tried building a Bessel Array using full-range drivers?
One of my friends has a larger living room and we are thinking about a 5-speaker Bessel using Markaudio CHR-120 drivers
Any suggestions/experience would be appreciated
Thank you

Well, from what I could see in my simulations, you should be at least 2m away from the array to have proper merging.

Specifically on the CHR: it has quite a large VAS, so with 5 drivers, you would need quite some volume. Then there is the beaming of the drivers at high frequency. So best to listen on-axis.

 

[ppataki]

Thanks @voodooless
He will be listening from 3.5 meters, cabinet volume will be 76 liters (his current box is 89 liters so no issues with that )
He always listens on-axis in an exact equilateral triangle
I guess then it shall all be fine, right?

 

[ppataki]

Actually one question came up: how shall we configure ear height? I mean shall the top-most speaker be at ear height? Or maybe the second one below?.....
Not sure about that

 

[voodooless]

Actually one question came up: how shall we configure ear height? I mean shall the top-most speaker be at ear height? Or maybe the second one below?.....
Not sure about that

Ear height should be the middle driver. For a few people listening I think the setup could work. A small CBT array might also be a consideration. Note that efficient wise, a Bessel array is obviously not the best, it's better than a single driver but not as good as 5 drivers.

 

[ppataki]

All clear, thank you @voodooless

 

[voodooless]

I would definitely build one sample and test it out before making any more.

 

[BenB]

Ear height should be the middle driver. For a few people listening I think the setup could work. A small CBT array might also be a consideration. Note that efficient wise, a Bessel array is obviously not the best, it's better than a single driver but not as good as 5 drivers.

If someone is building a 5 element bessel array, I would recommend that they raise it so that ear height is where I put the tweeter in this diagram:

https://www.audiosciencereview.com/forum/index.php?attachments/two-way_bessel5_locations_annotated-png.160610/

That's right between the two +1 drivers in the array.

 

[ppataki]

Any particular reason for that @BenB?
Can that be simulated maybe? @voodooless
Thank you

 

[Yevhen]

Hi guys, did somebody hear those large arrays in the average living room or even a bedroom ?

What is the max distance between the drivers (from a practical point of view)?