Since class D amps are so cheap, maybe the way to do this is to use two power supply isolated class D amps each in mono. Use separate non-grounded SMPS bricks for each one. You could use each amp in regular mono or mod to be PBTL in order to fully utilize all circuits and get max power. Then use regular passive matrix wiring per Elias' or Nagaoka circuits. This would be simple to try out. I think it should work because the negative on each amp can be tied together now with the negative of another amp since the power rails are independent. This is certainly a loss less expensive than any multichannel sound card. The amp to try is the new SMAKN TPA3116D2 with built in optional solder bridges for PBTL.
What's the matter with the simple passive matrix ?
Best things in life are simple
.
There's nothing wrong with the speaker-level connection.
However, I don't have 3 identical speakers (or 3 identical drivers) suitable for this. Using different speakers for such 3-ch, it'd be a mess by their different impedances and efficiencies.
I wish it can be simple, too, but reality isn't that way.
And I'm too greedy to let go of the active big system.
Simple passive matrix...
Hi Elias,
Came across your site a while back and am ready willing and unable to implement it.
I am familiar with crossovers, series & parallel having designed both 2 & 3 way types but your matrix is beyond me.
Where and how do I start. So any help would be appreciated.
Thanks
Fred P.
Hi Elias,
Came across your site a while back and am ready willing and unable to implement it.
I am familiar with crossovers, series & parallel having designed both 2 & 3 way types but your matrix is beyond me.
Where and how do I start. So any help would be appreciated.
Thanks
Fred P.
Hi Elias,
Came across your site a while back and am ready willing and unable to implement it.
I am familiar with crossovers, series & parallel having designed both 2 & 3 way types but your matrix is beyond me.
Where and how do I start. So any help would be appreciated.
Thanks
Fred P.
I just followed the diagram that Elias posted on his site. Here is my implementation - I used a small protoboard and did point to point wiring. You need a board as there are actually more connections than you think.
http://www.diyaudio.com/forums/full-range/223313-foam-core-board-speaker-enclosures-239.html#post3961261
The circuit is this:
An externally hosted image should be here but it was not working when we last tested it.
When built, looks like this:
You have to play with the R & C values to get the psychoacoustic filter you need - it is basically a high pass shunt to provide a low pass filter for the central driver.
Final box looks like this:
Note that Elias' circuit is actually simpler than Nagaoka's - a good place to start.
Basically for stereo difference there is a push/pull action across the loop between the left and right drivers and the mono part just goes straight through all drivers.
That way you can steer the sound by canceling out with a "mirror version" of the stereo information.
So it's kind of like M/S stereo with a mic, only of course emitting sound, and with a low pass filter for the middle driver.
The common is just connected to both negative terminals, though Nagaoka seems to prefer just connecting to one of them.
That way you can steer the sound by canceling out with a "mirror version" of the stereo information.
So it's kind of like M/S stereo with a mic, only of course emitting sound, and with a low pass filter for the middle driver.
The common is just connected to both negative terminals, though Nagaoka seems to prefer just connecting to one of them.
I just discovered this thread, and didn't read the whole thing, so I hope I'm not being redundant here. I have a few thoughts.
Each driver gets a different signal, so they must be in separate enclosures (isolated from one another internally) or they will modulate each other and become more mono, especially at low frequencies.
The other thing is that I don't think you want to have L-R and R-L signals sent to the side firing drivers, since that amounts to a mono signal when you exclude the center speaker. Yes when added acoustically to the center speakers I believe you will still get a significant good effect, but I can't help wondering if the effect might be better if the side driver signals were instead L-XR and R-XL, where X might be 0.5 - 0.7 rather that 1.0, so they by themselves would still be perceived as stereo (with significant cancellation of the L+R signal). I would think this would have an even better stereo effect.
Each driver gets a different signal, so they must be in separate enclosures (isolated from one another internally) or they will modulate each other and become more mono, especially at low frequencies.
The other thing is that I don't think you want to have L-R and R-L signals sent to the side firing drivers, since that amounts to a mono signal when you exclude the center speaker. Yes when added acoustically to the center speakers I believe you will still get a significant good effect, but I can't help wondering if the effect might be better if the side driver signals were instead L-XR and R-XL, where X might be 0.5 - 0.7 rather that 1.0, so they by themselves would still be perceived as stereo (with significant cancellation of the L+R signal). I would think this would have an even better stereo effect.
Last edited:
The modulation isn't a problem at all for moderate volumes and small drivers in a fairly large box. You just have to try it and see for yourself. The math seems strange but works.
Most of the time the bass is sympathetic on both channels. All modern mini Bluetooth speakers use common volume and common passive radiator. Those boxes would benefit from this SSS for better stereo.
Most of the time the bass is sympathetic on both channels. All modern mini Bluetooth speakers use common volume and common passive radiator. Those boxes would benefit from this SSS for better stereo.
xrk971, remind me again what the cap is doing there in post 170 ?
I see the resistor, and it is knocking down the signal to the center driver.
I like the passive summed mono.
I remember having the car audio magazine with an article on it.
How To - Car Stereo - How To Install A Center Channel
L-pad around mono speaker works well as long as impedance is fairly flat.
My buddy swears by it.
"If you run them with a center channel on the long wall, they image just like live music.
Most stereo is a 'tunnel' effect. If you're not in the right spot, or if you turn your head, the illusion goes away.
At a live performance if the trumpet player is on the left, and you turn your head towards him, he gets louder.
With a L-C-R set-up it images the same way.
At a live performance if you sit on the left, certain instruments are louder, if you sit on the right, other instruments are louder. No matter where you sit, you hear a smooth 'curtain' of sound across the stage.
An L-C-R set-up behaves in the same way."
"As we see, the left and right speaker grounds are tied to each other and return to amplifier ground through the center speaker. The L-pad controls the level of the center speaker from all the way off to about 3dB louder than it need be (assuming all three speakers are similar).
Rear surrounds (optional) receive the difference between left and right, mainly ambient type information. Again, the L-pad allows for level adjustment. Note that the two different L-pads are connected differently.
You will be surprised at how well these simple circuits work. The first car I put a center channel in with this circuit won all sorts of trophies for sound quality in the lower power classes, no one believed there were not more amplifiers."
I see the resistor, and it is knocking down the signal to the center driver.
I like the passive summed mono.
I remember having the car audio magazine with an article on it.
How To - Car Stereo - How To Install A Center Channel
L-pad around mono speaker works well as long as impedance is fairly flat.
My buddy swears by it.
"If you run them with a center channel on the long wall, they image just like live music.
Most stereo is a 'tunnel' effect. If you're not in the right spot, or if you turn your head, the illusion goes away.
At a live performance if the trumpet player is on the left, and you turn your head towards him, he gets louder.
With a L-C-R set-up it images the same way.
At a live performance if you sit on the left, certain instruments are louder, if you sit on the right, other instruments are louder. No matter where you sit, you hear a smooth 'curtain' of sound across the stage.
An L-C-R set-up behaves in the same way."
"As we see, the left and right speaker grounds are tied to each other and return to amplifier ground through the center speaker. The L-pad controls the level of the center speaker from all the way off to about 3dB louder than it need be (assuming all three speakers are similar).
Rear surrounds (optional) receive the difference between left and right, mainly ambient type information. Again, the L-pad allows for level adjustment. Note that the two different L-pads are connected differently.
You will be surprised at how well these simple circuits work. The first car I put a center channel in with this circuit won all sorts of trophies for sound quality in the lower power classes, no one believed there were not more amplifiers."
The cap provide a low pass filter for the center channel by shunting some of the HF's to ground. This allows you to adjust the "presence" of the center channel so that it is not too dominant from a psychoacoustic imaging stand point. Hence, called the "psychoacoustic filter". At least that is my understanding of it. Elias can probably explain better - that circuit came straight off his website.
A lot more info on SSS here on Elias' website:
http://elias.altervista.org/html/SingleSpeakerStereo.html
TPA3123 an unbridged class D amp
It outputs max 25W/channel into a 4-Ω Load from a 27-V supply. It can also be bridged and the output is then doubled.
I bought one of these boards (actually the 2.1 version) and established that the GRND (–) speaker terminals of the L and R stereo channels have a common ground.
Recently, I became interested in 3-channel stereo and found this thread.
Based on the scheme of xrk971 in post #170, I used the TPA3123 fed with a 19V DC laptop power supply and hooked it up to three small bookshelf speakers, each about 10'' apart.
My temporary circuit contains a 5Ω 10W resistor but not yet a capacitor.
Despite the shortcomings, my first impressions are that the spacious sound is simply amazing!
Indeed, one particular class D amp, the TPA3123 is in principle single ended (see TI's data sheet).
It outputs max 25W/channel into a 4-Ω Load from a 27-V supply. It can also be bridged and the output is then doubled.
I bought one of these boards (actually the 2.1 version) and established that the GRND (–) speaker terminals of the L and R stereo channels have a common ground.
Recently, I became interested in 3-channel stereo and found this thread.
Based on the scheme of xrk971 in post #170, I used the TPA3123 fed with a 19V DC laptop power supply and hooked it up to three small bookshelf speakers, each about 10'' apart.
My temporary circuit contains a 5Ω 10W resistor but not yet a capacitor.
Despite the shortcomings, my first impressions are that the spacious sound is simply amazing!
Kor010,
Wow - great to know that. I never knew the TPA3123 was unbridged. This is great news although it's getting tougher to find this amp offered as easily as the 3116 or 3118.
Thanks for letting us know.
Edit: a quick search shows these are still sold. Indeed the output pins are labeled GND and Right and GND and Left rather than +/-
http://m.aliexpress.com/item/32603867013.html
Wow - great to know that. I never knew the TPA3123 was unbridged. This is great news although it's getting tougher to find this amp offered as easily as the 3116 or 3118.
Thanks for letting us know.
Edit: a quick search shows these are still sold. Indeed the output pins are labeled GND and Right and GND and Left rather than +/-
http://m.aliexpress.com/item/32603867013.html
Last edited:
@xrk971
I bought the same board and I was not happy with it.
In the thread "TPA3116D2 Amp" post #9218, I described my experiences with a few TPA31xx boards
For my 3-channel matrix system I used this TPA3123 2.1 board instead
I bought the same board and I was not happy with it.
In the thread "TPA3116D2 Amp" post #9218, I described my experiences with a few TPA31xx boards
There may be of course better TPA3123 stereo boards available and I would love to know these.
For my 3-channel matrix system I used this TPA3123 2.1 board instead
Last edited: