designing dipole baffle, how wide in relation to midrange XO point?

[thadman]

I'm having a bit of difficulty coming to a conclusion in regards to my current baffle design. With excursion being greatest at the bottom of a drivers passband (assuming IB operation), one would assume designing the baffle width so that the boost occurs around the 300hz XO point (assuming 6dB boost occurs one octave below fundamental wavelength of the baffle) would be a good idea...but that discounts the influence of the crossover. I've got a DCX2496, so any slope <48dB is achievable, but lets generalize and assume I'm going to use a 4th order (24dB) Linkwitz-Riley slope, which should give us -6dB at the crossover point and alleviate excursion requirements by 50%. Where would I want that 6dB boost to be so I can design the baffle width around it?

Also, Linkwitz' mentions phase distortion (byproduct of XO) is much more noticeable at lower frequencies and inaudible at higher frequencies. Would I gain anything by going with a lower order filter at the 300hz crossover point?

[Doug Lockwood]

), one would assume designing the baffle width so that the boost occurs around the 300hz XO point (assuming 6dB boost occurs one octave below fundamental wavelength of the baffle) would be a good idea...but that discounts the influence of the crossover.

I have a Similar speaker running. Dual RS270, Seas M14 Mid-range, NEO3 with a DCX2496. My design answer is to cross 1 octave above F equal, if the crossover is a LR4. Doing this will reduce the excursion requirements of the mid-range on the baffle.
However, the answer is: What are the trade-offs?

By the way, I have crossed as low as 200 Hz and as high as 500 Hz LR4. I really am hard pressed to tell the difference. I tried LR2 @ 320 and didn't like the result.

My advice: Create Sawdust. Report Back. Enjoy.

HTH

Doug

[Dennis H]

Suggest you download The Edge and play with it. Be sure to check the 'open baffle' box.

http://www.tolvan.com/edge/

[thadman]

I already have "edge", but it doesnt include the influence of the crossover on excursion requirements. I will be taking advantage of it, once I conclude where the boost is most appreciated.

[Dennis H]

Oh, man. You know, there comes a point where you just have to build something. How many months/years have you been asking questions you could figure out for yourself with a bit of effort? I've tried to be patient but I'm joining the others and getting the feeling that the questions will never end and nothing will ever get built. Analysis Paralysis.

[augerpro]

thadman do you have any modeling software?

From an excursion standpoint the lower crossover point for the mid depends first on what it is physically capable of, and second on just how little excursion you want it to actually do as it has some effect on SQ. What I can say is you will need real excursion starting around 200Hz. So if it was REALLY important to keep it to minimum I would cross at around 400Hz LR4. That's just a guesstimate. If you want an EXACT answer you really need to just build a mockup with the actual drivers adn do some measurements.

From what little experience I have though I would say there are much more important factors that influence that crossover point. Excursion would only be an important consideration for me if I had mid that really had a tiny Xmax, like some PA drivers do. Otherwise ctc distance, floor bounce, phase integration, tonal qualities, and baffle step would be the factors I would use to determine that crossover point.

[thadman]

Oh, man. You know, there comes a point where you just have to build something. How many months/years have you been asking questions you could figure out for yourself with a bit of effort? I've tried to be patient but I'm joining the others and getting the feeling that the questions will never end and nothing will ever get built. Analysis Paralysis.

I know and I'm itching to get started, but there isn't a cheaper design of the Isiris available yet, so I'm going to have to spend some time designing it myself instead of making a fatal error while I'm building it. My questions have recently shifted from driver choice to baffle...I'm almost there. I have decided on the 831882 + RSS390HF on flat baffles to keep things simple.

From what little experience I have though I would say there are much more important factors that influence that crossover point. Excursion would only be an important consideration for me if I had mid that really had a tiny Xmax, like some PA drivers do. Otherwise ctc distance, floor bounce, phase integration, tonal qualities, and baffle step would be the factors I would use to determine that crossover point.

I have already determined my crossover point (300hz) and am designing my baffle around the crossover point to position the phase boost (1 octave below the fundamental wavelength of the baffle) where its most needed.

[augerpro]

If you're reffering to baffle step then I would try advise against placing where you have to implement BSC to both the woofer and the mid. Either/or. IMO

Generally on dipole I think the main consideration is bass response in determining the width. And fine tune that so that it works with other, secondary, considerations. The dipole peak is one of those, I don't have a real answer there. I'm doing my first dipole also. Given that the bass response dictates the width and thus the peak you don't have alot leeway. You'll have to incorporate it into the XO network so whatever is easiest to deal with I guess...

[thadman]

If you're reffering to baffle step then I would try advise against placing where you have to implement BSC to both the woofer and the mid. Either/or. IMO

Generally on dipole I think the main consideration is bass response in determining the width. And fine tune that so that it works with other, secondary, considerations. The dipole peak is one of those, I don't have a real answer there. I'm doing my first dipole also. Given that the bass response dictates the width and thus the peak you don't have alot leeway. You'll have to incorporate it into the XO network so whatever is easiest to deal with I guess...

I'm going active with a DCX, so BSC can be easily compensated with equalization. I'd like to place the dipole boost where it will lower distortion the most. Also the baffle will not be a rectangular shape (narrower at top, wider at bottom).

[augerpro]

Well technically I think the dipole peak *is* distortion.

[thadman]

How do I model peculiar shapes in edge so I can get an idea on what baffle shapes I should build/test/measure?

Is there anyway I can get around the rectangular shape requirement?

[herm]

Quoting Dennis:

How many months/years have you been asking questions you could figure out for yourself with a bit of effort?

[augerpro]

Click and drag the corners to make trapazoid type shapes. I don't think you can do complex or curved shapes. For those i would suggest buidling and measuring the results to see what they do.

[thadman]

Quoting Dennis:

How many months/years have you been asking questions you could figure out for yourself with a bit of effort?

calm down man, I've been getting this a lot lately and it isn't getting me anywhere. All of my recent questions have been related to the construction of a single specific design, not vague questions referring to different theories/concepts. I still have specific questions related to building and dipole baffle design and those need to be addressed before I can begin construction. I'm going to begin cutting up some test baffles between next week and the end of this month, if time permits

[thadman]

Does it matter how tall the baffle is in the simulator, or only where the driver is in comparison to the mic? By virtue of the design, it is going to require a rather tall baffle (maybe >4') as the top of the 2nd RSS390HF will be at least 38-40" off the ground.

[augerpro]

Well it has to represent the actual baffle that would be used adn the mic positioned representative of the real life listening/design axis. Or maybe I've misunderstood your question?

[thadman]

nvm, I figured it out

[Rudolf]

How do I model peculiar shapes in edge so I can get an idea on what baffle shapes I should build/test/measure?
Is there anyway I can get around the rectangular shape requirement?

There is a "# of corners" input field in EDGE. You should have seen it already. Shapes can only be built with straight sections, so put in a sufficient number of corners first.
All corner points will first be concentrated in the lower left corner of the original rectangle. You have to pull them out one by one. So you better design your own shape up from there.

[JonMarsh]

Click and drag the corners to make trapazoid type shapes. I don't think you can do complex or curved shapes. For those i would suggest buidling and measuring the results to see what they do.

You can approximate nearly anything in EDGE by adding vertex's and using short straight line segments to approximate the shape; at frequency of interest (below 1 kHz for the most part) this works just fine...