Doomed project? (Cardoid attempt)

[Ktacos]

Hello, I'm working on a small speaker at the moment similar in dimension and driver layout to the Kii Seven in hopes of playing around with cardoid dispersion. I'm keeping the drivers cheap as I'm not sure I'll stick with the idea. Currently the box is router for a Dayton Sig150+ND25FW, and two Dayton DCS165's on the sides. I assume this setup will be SPL limited as there is not much volume for the drivers and EQ will be needed to get OK bass extension.

I'm trying to figure out some info on how the side woofers are driven. As I understand it a delay and phase manipulation needs to occur for this to all work. I have 6 channels of DSP and amplification to go around, but unsure of the requirements for driving the side woofers. I imagine both woofers need to be operating in opposition to each other, so they both push/move outwards from the cabinet. Am I correct in assuming that there will be issues attempting to run all the woofers off of two channels? If one woofer is wired out of phase with the other, how will the phase alterations apply themselves? I'm wary that I would need to add another pair of channels to control one of the pairs of woofers. I'm overall finding it tough to find the info I want on this kind of speaker and would appreciate any advice.

 

[kemmler3D]

Am I correct in assuming that there will be issues attempting to run all the woofers off of two channels? If one woofer is wired out of phase with the other, how will the phase alterations apply themselves?

As far as I know, you're on the right track as far as cardioid depending on cancellation via out-of-phase waves. This can be done acoustically or via additional drivers... but I am not sure on exactly how much phase cancellation and filtering you need. I guess for starters, you need to find out how much bass the front driver is contributing at the BACK of the speaker, then use the side drivers to cancel that as exactly as you can, without cancelling stuff you want in the front.

I think some experimentation and measurement is called for. I do think that just wiring the side woofers out of phase is not going to work particularly well.

Definitely interested in seeing what results you get though!

 

[Ktacos]

I know it will take quite a bit of trial and error on the dsp end but that parts easy. I think I may have bit off more than I can chew though, after some thought it's become apparent that I will require 4 channels per speaker for this to work, and I just don't know if I can really afford that, this type of speaker really seems better suited for custom dsp and amplification for just the speaker.

 

[kemmler3D]

I know it will take quite a bit of trial and error on the dsp end but that parts easy. I think I may have bit off more than I can chew though, after some thought it's become apparent that I will require 4 channels per speaker for this to work, and I just don't know if I can really afford that, this type of speaker really seems better suited for custom dsp and amplification for just the speaker.

Re-reading your post, I think you can get away with 3 channels per speaker:
1: Tweeter
2: Front woofer
3: Side woofers

The side woofers should probably run in phase with each other as long as they're placed symmetrically relative to the front woofer and the baffle is likewise symmetrical... but not having built a speaker before I'm talking purely out of my ear, here.

 

[Ktacos]

Re-reading your post, I think you can get away with 3 channels per speaker:
1: Tweeter
2: Front woofer
3: Side woofers

The side woofers should probably run in phase with each other as long as they're placed symmetrically relative to the front woofer and the baffle is likewise symmetrical... but not having built a speaker before I'm talking purely out of my ear, here.

I don't know either and I've made tons, but nothing like this. The Mesanovic CDM65 on erins audio corner kinda shows what I may be able to expect, even uses the same dayton sub drivers. Of course it seems pretty compression limited.

I think this also depends on desired cardoid range, I guess I need to accept that this is very much a prototype and it may just end up in the bin.

 

[Ktacos]

So it does appear I would need to have another pair of channels for the second woofers for cardoid to work right, guess I'll have to scrap this one, 8 channels is a bit much for me and I'd largely lose the ability to use my subs. Oh well not the first project I've had to scrap.

 

[jaakkopetteri]

Sounds like a wonderful project. What do you mean by channels here - amplification or DSP? In either case, you could also go hybrid and cross the mid and tweeter passively, then crossing the mid and the subs actively.

VituixCAD is a great tool for playing around with cardioid concepts. You can create ideal drivers with a nominally flat response, simulate the responses in your baffle and "align" them on the enclosure to get a qualitatively accurate playground. As the CMD65 measurements show, it's difficult to get a perfectly cardioid response, but it shouldn't be difficult to get most of the benefit.