The main limitation that I can see with the current design is that the sum and difference drivers share one cabinet volume. Due to the nature of most stereo recordings (bass is shared evenly over L & R) the difference drivers mainly replay mid & high frequencies. If I had the space i'd install a pipe between the rear of the two difference drivers filled with damping material. This would isolate the two equal but opposite drivers from the main cabinet volume. Then it would be much simpler to optimise the bass performance as the two sum drivers would be working into a defined volume with a single port.
I finally got round to introducing a limited amount of isolation between the sum and difference drivers, using some handy carpet samples. This has definitely reduced the mid and HF leaking from one channel to the other (mid high mono content now isn't audible via the difference drivers).
I did one other change that has been very worthwhile. I changed the two sum drivers from operating in parallel at all frequencies to being a 1.5 way pair using a 4.7mH inductor on the lower driver. I did some very rough calculations and figured that the lower driver would roll in +6dB at the baffle step (600-700Hz from memory, but a guess given the angled cabinet baffle). This has added lower mid and bass weight and also significantly improved the HF dispersion, which was compromised with two full range drivers covering HF.
This is definitely a trial and error effort (many estimates, not fully modelled) but it sounds good
nice way of listening to BBC Radio 4.
I did one other change that has been very worthwhile. I changed the two sum drivers from operating in parallel at all frequencies to being a 1.5 way pair using a 4.7mH inductor on the lower driver. I did some very rough calculations and figured that the lower driver would roll in +6dB at the baffle step (600-700Hz from memory, but a guess given the angled cabinet baffle). This has added lower mid and bass weight and also significantly improved the HF dispersion, which was compromised with two full range drivers covering HF.
This is definitely a trial and error effort (many estimates, not fully modelled) but it sounds good
nice way of listening to BBC Radio 4.Attachments
I don't use Dropbox or Photobucket. My stuff is here:
http://www.diyaudio.com/forums/full...board-speaker-enclosures-239.html#post3961261
A quick web search turned this up - I believe it's the same one :
The diagram must have been a link to one of Elias' images which may have gone bad. I don't have my own diagram. Perhaps Elias can help?
More here:
http://elias.altervista.org/html/SingleSpeakerStereo.html
Last edited:
A friend gave me a pair of cheap paper-cone full range speakers. I figured they'd make a nice trio with my other cheap paper-cone full range speaker. I turned to this thread for inspiration and made some quick drawings. My quick calculations show that this makes a sealed enclosure of approx.10 liters. Design is for 4" drivers, adjust the hole for whatever driver you're using. Hope this is useful. Any comments or correction please feel free.
I turned to this thread for inspiration and made some quick drawings. My quick calculations show that this makes a sealed enclosure of approx.10 liters. Design is for 4" drivers, adjust the hole for whatever driver you're using. Hope this is useful. Any comments or correction please feel free.Attachments
What about a 2 driver version with the side channel set up as an open baffle dipole? The center driver could still be in a box and the side driver behind it. In the wiring image for the 3 driver version in post #1, just delete the right driver (but keep the wire). Would this require a "psychoacoustic" circuit, or would it be fine just wired up with straight copper? Sorry, I really don't understand circuitry very well.
Did anyone have a conclusion on the correct class of amplifier to use for this arrangement.
I have the suggested x=0.5 wiring set up on a cheap 3116 class d amp. It's a 2.1 so I have built in a bass unit, with 3 Visaton FR 8 2" full rangers (which are tightly sealed 1-2ltrs inside a 12ltr box). It sounds great. I think I need to dampen the L+R centre as I have a mono addition with the sub, but it does sound great.
As mentioned here (years ago) I do get some buzzing (particularly on piano), which I'm guessing is the amp straining. XRK suggested something like this; TDA7297 2*15W Digital Stereo Audio Amplifier Dual Channel 9-15V Module AMP G3Z2 | eBay
...but is that particular chip universally class AB? In some listings it says it is a class D amp. I am not an Amp expert! Just wondering if I could get some experiences and/recommendations?
Secondly, if I want to apply some second order high-pass relief to the full range matrix, I'm guessing I would need to do this twice in series with the +L and +R?
I have the suggested x=0.5 wiring set up on a cheap 3116 class d amp. It's a 2.1 so I have built in a bass unit, with 3 Visaton FR 8 2" full rangers (which are tightly sealed 1-2ltrs inside a 12ltr box). It sounds great. I think I need to dampen the L+R centre as I have a mono addition with the sub, but it does sound great.
As mentioned here (years ago) I do get some buzzing (particularly on piano), which I'm guessing is the amp straining. XRK suggested something like this; TDA7297 2*15W Digital Stereo Audio Amplifier Dual Channel 9-15V Module AMP G3Z2 | eBay
...but is that particular chip universally class AB? In some listings it says it is a class D amp. I am not an Amp expert! Just wondering if I could get some experiences and/recommendations?
Secondly, if I want to apply some second order high-pass relief to the full range matrix, I'm guessing I would need to do this twice in series with the +L and +R?
I can’t help anyone shopping for an amp, but there’s one very easy test you can do on any amp to see if it will work with matrixed stereo.
Measure the resistance between the two (-) speaker output terminals on any given amp. If it’s around zero, you’re good to go. If the resistance is high (or infinity) ohms, it’s only suitable for traditional 2-channel stereo.
Measure the resistance between the two (-) speaker output terminals on any given amp. If it’s around zero, you’re good to go. If the resistance is high (or infinity) ohms, it’s only suitable for traditional 2-channel stereo.
Last edited: