I wish that the content of this thread was a bit easier to follow but I got lost in it. My old National ghetto blaster back in 89 had a stereo/mon button. It made a clear difference when switched to a wider and more alive sound. Very pleasing when compared to the mono setting. I wonder how they did it
Always thought that the word "loudspeaker" meant the whole single cab and "speaker driver" meant the transducer
Always thought that the word "loudspeaker" meant the whole single cab and "speaker driver" meant the transducer
Multiplying by 1/k and the formulae simplify a bit.
What if L and R are mounted on the same open-baffle board facing/firing left and right respectively? Highest frequency won't fire backward, is that a real problem? (Inspired by @acmn)
(k=1/2) L-R | L+R | R-L can be approximated using 4 fullrange drivers: OB left-facing L, right-facing R; non-OB front speakers L, R.
(k=1/3) 2L-R | L+R | 2R-L can be approximated by adding to the above, 2 more non-OB speakers left-facing L, right-facing R.
What if L and R are mounted on the same open-baffle board facing/firing left and right respectively? Highest frequency won't fire backward, is that a real problem? (Inspired by @acmn)
(k=1/2) L-R | L+R | R-L can be approximated using 4 fullrange drivers: OB left-facing L, right-facing R; non-OB front speakers L, R.
(k=1/3) 2L-R | L+R | 2R-L can be approximated by adding to the above, 2 more non-OB speakers left-facing L, right-facing R.
Last edited:
(edited)
Multiplying by 1/k and the formulae might simplify a bit....
What if L and R are mounted on the same open-baffle board (possibly H-frame) facing/firing left and right respectively? Highest frequency won't fire backward, is that a real problem? (Inspired by @acmn) (edit: doh! msg#462)
L-R | L+R | R-L can be approximated using 4 fullrange drivers: OB left-facing L, right-facing R; non-OB front speakers L, R.
2L-R | L+R | 2R-L can be approximated by adding to the above, 2 more non-OB speakers left-facing L, right-facing R.
3L-R | L+R | 3R-L = (2L-2R)+(L+R) | L+R | (2R-2L)+(L+R) i.e. OB L->left R->right plus OMNI volume-controlled L+R.
Truth be told, I find the math back-and-forth pretty confusing (despite being a mathematician). I would like to make it work using the following: multiple drivers and integrated amps, and an adder box if absolutely necessary.
Multiplying by 1/k and the formulae might simplify a bit....
What if L and R are mounted on the same open-baffle board (possibly H-frame) facing/firing left and right respectively? Highest frequency won't fire backward, is that a real problem? (Inspired by @acmn) (edit: doh! msg#462)
L-R | L+R | R-L can be approximated using 4 fullrange drivers: OB left-facing L, right-facing R; non-OB front speakers L, R.
2L-R | L+R | 2R-L can be approximated by adding to the above, 2 more non-OB speakers left-facing L, right-facing R.
3L-R | L+R | 3R-L = (2L-2R)+(L+R) | L+R | (2R-2L)+(L+R) i.e. OB L->left R->right plus OMNI volume-controlled L+R.
Truth be told, I find the math back-and-forth pretty confusing (despite being a mathematician). I would like to make it work using the following: multiple drivers and integrated amps, and an adder box if absolutely necessary.
Last edited:
This patent from msg#462: matrix L+R to both front-firing closed speakers, L-R and R-L to side-firing dipole speakers i.e. 2(L-R) | 2(L+R) | 2(R-L) is that right? Why not un-matrix the speakers L, R front and L, R sides.
Using the formula (L-R)+2kR | L+R | (R-L)+2kL, simply an extra integrated amp to feed 2kR, 2kL to extra side speakers (6 total, of which 2 are dipole).
And if not using this dipole trick: (L+R)-2(1-k)R | L+R | ... i.e. adder output and one int amp for the 2(1-k) coefficient (5 speakers total unless L+R is OMNI then 3).
Last edited:
I started out trying to get "foot-of-bed" stereo then got hooked on up-firing omni mono/stereo, even in "LX" configuration:
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7656303
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7675238
(Pictured below: stacked-LX 4" high-gloss piano finish 4.5L each.)
Earlier I had tried the Hafler Differential hookup with inconsistent or befuddling results:
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7605036
All these ideas will come together somehow ;-)
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7656303
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7675238
(Pictured below: stacked-LX 4" high-gloss piano finish 4.5L each.)
Earlier I had tried the Hafler Differential hookup with inconsistent or befuddling results:
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7605036
All these ideas will come together somehow ;-)
Last edited by a moderator:
OK here's what I've got working, and I should just declare success, clean up and go to bed (Beijing time).
I simply used a 3.5mm Y-adaptor to two sets of RCAs, to connect my hi-fi phone to two TPA3221-on-steroids integrated amps. One amp drives a pair of up-firing omni 4" speakers (touching), the other amp drives a pair of nude 4" drivers (dipole) firing to the sides (toeing helped a lot). Through tweaking I settled on blocking the nude drivers (else lose floating effect); covering the nude drivers (else lose depth); adding ad hoc reflector on the right (wall on left worked much better); first turn up the omni pair to get decent bass (but not quite enough) and the floating image, then turn up the dipole pair to spread out the image; finally tweak the overall volume so the floating 3D sound stage distance is consistent with the loudness (quite important but my phone volume control isn't granular enough). I tried 1-5 meters and they all kind of worked, with caveats noted above. Whew!
The split formula is (L+R)+f(L-R) | L+R | (L+R)+f(R-L). (f=2 then @acmn; f=3 then closer to @Elias)
I simply used a 3.5mm Y-adaptor to two sets of RCAs, to connect my hi-fi phone to two TPA3221-on-steroids integrated amps. One amp drives a pair of up-firing omni 4" speakers (touching), the other amp drives a pair of nude 4" drivers (dipole) firing to the sides (toeing helped a lot). Through tweaking I settled on blocking the nude drivers (else lose floating effect); covering the nude drivers (else lose depth); adding ad hoc reflector on the right (wall on left worked much better); first turn up the omni pair to get decent bass (but not quite enough) and the floating image, then turn up the dipole pair to spread out the image; finally tweak the overall volume so the floating 3D sound stage distance is consistent with the loudness (quite important but my phone volume control isn't granular enough). I tried 1-5 meters and they all kind of worked, with caveats noted above. Whew!
The split formula is (L+R)+f(L-R) | L+R | (L+R)+f(R-L). (f=2 then @acmn; f=3 then closer to @Elias)
OK here's what I've got working, and I should just declare success, clean up and go to bed (Beijing time).
I simply use a 3.5mm Y-adaptor to two sets of RCAs, to connect my hi-fi phone to two TPA3221-on-steroids integrated amps. One amp drives a pair of up-firing omni 4" speakers (touching), the other amp drives a pair of nude 4" drivers (dipole) firing to the sides (toeing helps a lot). Through tweaking I settled on blocking the nude drivers (else lose floating effect); covering the nude drivers (else lose depth); adding ad hoc reflector on the right (wall on left worked much better).. First turn up the omni pair to get decent bass (but not quite enough) and the floating image, then turn up the dipole pair to spread out the image; finally tweak the overall volume so the floating 3D sound stage distance is consistent with the loudness (quite important but my phone volume control isn't granular enough). I tried 1-5 meters and they all kind of worked, with caveats noted above. Whew!
The split formula is (L+R)+f(L-R) | L+R | (L+R)+f(R-L). (f=2 then @acmn; f=3 then closer to @Elias)
I simply use a 3.5mm Y-adaptor to two sets of RCAs, to connect my hi-fi phone to two TPA3221-on-steroids integrated amps. One amp drives a pair of up-firing omni 4" speakers (touching), the other amp drives a pair of nude 4" drivers (dipole) firing to the sides (toeing helps a lot). Through tweaking I settled on blocking the nude drivers (else lose floating effect); covering the nude drivers (else lose depth); adding ad hoc reflector on the right (wall on left worked much better).. First turn up the omni pair to get decent bass (but not quite enough) and the floating image, then turn up the dipole pair to spread out the image; finally tweak the overall volume so the floating 3D sound stage distance is consistent with the loudness (quite important but my phone volume control isn't granular enough). I tried 1-5 meters and they all kind of worked, with caveats noted above. Whew!
The split formula is (L+R)+f(L-R) | L+R | (L+R)+f(R-L). (f=2 then @acmn; f=3 then closer to @Elias)
Last edited:
p.p.s. If the "omni" speakers are backstopped (like I did with the green cutting mat), the front output will be greater than the sides, say double -- f=1.5 yields 0.5(L+R)+1.5(L-R) | L+R | ... i.e. Elias 2L-R | L+R | ....
So I'm not sure which split formula I got working haha. I might have to swap in my TA2024 amps with dB-stepped volume control.
So I'm not sure which split formula I got working haha. I might have to swap in my TA2024 amps with dB-stepped volume control.
Last edited:
English please
dave
diyAudio moderation team
J.R.R. Tolkien about the Finnish language:
It was like discovering a complete wine-cellar filled with bottles of an amazing wine of a kind and flavour never tasted before. It quite intoxicated me.
🤓
Tuntui siltä kuin olisin löytänyt kokonaisen viinikellarin täynnä upeita viinejä, sen kaltaisia ja makuisia, että en ollut koskaan kohdannut vastaanlaisia. Se kieli todella huumasi minut.
Last edited:
I think I got a little confused by the different formulae in msgs# 490=493 not= 497=513. Perhaps standardize on L+R coefficient 1 for the center would help a bit (disallow boundary case of no center).
Anyway, #513 looks kind of like a box missing sides "short-circuit" backwave -x(L+R) plus sealed side drivers L and R.
Anyway, #513 looks kind of like a box missing sides "short-circuit" backwave -x(L+R) plus sealed side drivers L and R.
Last edited:
I am referring to this star connection of the speaker elements. Msg #492 is another story.
(I am talking about false circuit analysis by Elias, sorry to say.)
Last edited:
Of course you can imagine what ever signals to be fed to those speaker elements, but if you use them as star wired, then the signals are:
to left: L-x(L+R)
to right: R-x(L+R)
from the middle: x(L+R)
with x running 0...0.5.
Of course these formulas are not the "real truth" because they are based on a model that sees speaker elements as resistors.