8'' two-way any reason why not?

[robert]

Re: 8'' two-way any reason why not?

Yup - it works, but you have to tweak it a little bit to suit. Overall, I think the whole idea was very doable, and gives you a near FR speaker that isn't too complex to design.

You should hear Throwback now with the bass damped down to a reasonable level. I am going to try reducing BSC a bit, and also micro-tweak the tweeter to get max detail w/o any harshness.

FWIW, I am at app. 1600Hz LR4 and with a 6.75" cone, 2Khz is about where driver diameter = wavelength.

Cheers / Robert

[Wolf]

Re: 8'' two-way any reason why not?

Yup - it works, but you have to tweak it a little bit to suit. Overall, I think the whole idea was very doable, and gives you a near FR speaker that isn't too complex to design.

You should hear Throwback now with the bass damped down to a reasonable level. I am going to try reducing BSC a bit, and also micro-tweak the tweeter to get max detail w/o any harshness.

FWIW, I am at app. 1600Hz LR4 and with a 6.75" cone, 2Khz is about where driver diameter = wavelength.

Cheers / Robert

The topend on MAX could be tizzy to some, but detail might disappear if it's padded any further. That's why I left it alone.

I'm at 1.6 kHz LR8 with a 5.75" cone-only, but including the half-surround rule yields 6.5", and relatively the same pistonic end as yours.

Oh yeah- and these things are rockin' right now!!!
:D:D:D:D
Wolf

[DDF]

Re: 8'' two-way any reason why not?

If tweeter is radiating nearly 180 degrees at Fc (crossover freq) and woofer is radiating at some angle less than this, an on-axis only model will result in a tweeter that looks balanced in sim/measurements but due to wall reflections and off axis (or total power response) will sound brighter.

You have to dip/tilt down the power response a bit. I think you did the same thing on yours, and it sounds like Wolf did, too. My overall tweeter level averaged about 1dB or so down from what the woofer was doing 500-800Hz. Really surprised me, since it did this even after I fixed the top octave. Sounded right in the living room, but maybe a tad dull in the auditorium in Iowa.

I am in a hotel in VA right now, but when I get back, will place the off-axis plots and final for the XO - a pic or graph is worth 1K words.

Cheers/Robert

Try reducing the tweeter level only from ~ 3 to 5 kHz.

I find this gets rid of the hotness or imbalance due to "power response hole" through crossover. It sounds more balanced IME than reducing the entire tweeter level.

There are 3 main ways to do this. First is to increase the tweeter crossover frequency, but it just widens the hole, not recommended.

My preferred method is to use shallower initial crossover slopes (lower Q electrical filter corners), asking for a bit more output from the woofer around 3 kHz, and less from the tweeter, to better blend the changes in driver directivity from woofer to tweeter. Going to shallower Q still maintains the power handling advantage of a higher order crossover, but improves the off axis response through this better blending.

I also like to add a dB or two on axis dip around 3 to 5 kHz. Here's the thing. There may be a couple dB dip here but if used with a crossover showing lower Q corners, the crossovers tend to be more in quadtrature than in-phase, and the power response is better balanced since off axis the woofer and tweeter will on average sum with a bit more output than an in-phase (eg LR) xover would, and will therefore better fill the hole.

Take a look at curt C's more recent frequency response targets on axis (attached for the Tritrix), they have a nice dip from 3 to 5 kHz and by all accounts his voicing is impeccable.

This is also how I tend to voice many of my designs, but I also allow more 3 to 5 kHz off axis power to be radiated, to partially compensate.

(PS the Statements are a special case with the TL mid, curt finding they needed a dip in the driver's range due to rear radiation).

Dave

[brianpowers27]

Re: 8'' two-way any reason why not?

After a few version tweaks with the northseas, I found this version to be the best so far.

[robert]

Re: 8'' two-way any reason why not?

Yes, you can shallow out the woofer a bit, but neither mine nor Wolf's driver (esp. his) will do this super well. 2.5K and 4.2K are fair irregularities in woofer response on the F8 and probably to be avoided.

Tanking the woofer coil fixed the big issue: dual peaks at app. 4.2K, and the XO rolloff pretty much buried the smaller 2.5K issue. The slope on the tweeter is relaxed quite a bit, and did definitely improve things. Smaller cap, shallower roll-off, and a bit of attenuation at 3-4Khz all helped. Also adds to output in the Fc area.

Raising the XO frequency further narrows the dispersion of the 8" woofer too, so you kind of end up in a closing circle, where you pick one issue to keep or another

I decided that keeping the tweeter just a touch down fell into the category of voicing over theory, and in my living room it is really good, though I am going to order a spread of resistors for the Zobel and amp-side tweeter resistance and tweak it up until it starts to get unruly.

Personal preferences are really interesting (and important) in audio. It's gotta sound good to you. I will quickly trade some detail for smoothness while many others won't hear of it, and some split the difference. "Hot" in a tweeter 6K and up will drive me crazy, while 2-6K just fatigues after a while. Probably why I hate mylar domes so bad :D

Do not know if I can make the upcoming TBD DIY locally but if I can get the speaks there, you can all comment/listen on revised bass and XO.

Thank you all for keeping Norfolk interesting, this has been a massively dull trip interspersed with moments of stark "missing-production-schedule" and "over-budget" terror. :eek:

Cheers / Robert

[Wolf]

Re: 8'' two-way any reason why not?

Hopefully dumb Q:
Do you just hit your targets and leave it in a design?
*OR*
Do you hit your targets, remove the targets and tweak?

It just looks like you optimized for your targets, and nothing beyond that in the graphs.
Curiousity...
Wolf

[robert]

Re: 8'' two-way any reason why not?

For me on this project: aim for targets, listen, remove targets, and tweak. When it was dead-on LR4, I didn't like the sound of it much. At Iowa, OK was good enough since it was 2AM the night before the drive

Cheers / Robert

[brianpowers27]

Re: 8'' two-way any reason why not?

Hopefully dumb Q:
Do you just hit your targets and leave it in a design?
*OR*
Do you hit your targets, remove the targets and tweak?

It just looks like you optimized for your targets, and nothing beyond that in the graphs.
Curiousity...
Wolf

I think I usually shoot for the targets. I find my ears are much less reliable from listening to listening than the graph. It turns out that the one I favored has similar characteristics as the one DDF described above. I find that if I am tweaking much of anything it is usually the padding of the tweeter.

RElative alignment of the acoustic centers likely facilitated the pureness of the model against the reality of a textbook formulation. My design is almost dead on LR4 at the tweeter axis.

[DDF]

Re: 8'' two-way any reason why not?

Yes, you can shallow out the woofer a bit, but neither mine nor Wolf's driver (esp. his) will do this super well. 2.5K and 4.2K are fair irregularities in woofer response on the F8 and probably to be avoided.

...

Raising the XO frequency further narrows the dispersion of the 8" woofer too, so you kind of end up in a closing circle, where you pick one issue to keep or another

Cheers / Robert

Robert,

Most if not all drivers will have some edge or cone resonances that make life interesting at 1 to 3 kHz. FWIW in my case I found that, in these sort of cases, ignoring the curve and theory and voicing it with a bit higher and shallower xover most often works better than driving the xover lower and steeper, even if there are some lumps on axis. YMMV!

It does push the woofer up higher into a region where it beams more, but that's exactly what you want: you want the total off axis output from 3 to 5 kHz to be a combination of the overly strong tweeter and the underly strong woofer, so it blends better and is smoother than if the off axis from 3 to 5 kHz was dominated by the overly strong tweeter.

By using these low Q xovers, the slight dip from 3 to 5 kHz and xovers that aren't in phase on axis, it won't sound like 6kHz and up is hot, because the room will contribute to fill in the slight 3 to 5 kHz depression.

FWIW, I really can't overstress this too much.

I think this is the "secret sauce" to making a two way with 6" or larger mid woofers sound good in a small to med room. You won't hear it as much in a DIY meet in a huge room, but you would in the real at-home installation.

It's not getting much uptake here, but I haven't heard anyone try it and report back, other than George at North Creek who raves about it.

I think you guys are really missing out. I know if I put it on a fancy web site with graphs etc it would get some traction, but I've put this particular idea out there on the page (since the late 90s!!), it doesn't need the eye candy.

Try it before discounting, I think you'll like it.

Dave

[brianpowers27]

Re: 8'' two-way any reason why not?

Robert,

Most if not all drivers will have some edge or cone resonances that make life interesting at 1 to 3 kHz. FWIW in my case I found that, in these sort of cases, ignoring the curve and theory and voicing it with a bit higher and shallower xover most often works better than driving the xover lower and steeper, even if there are some lumps on axis. YMMV!

It does push the woofer up higher into a region where it beams more, but that's exactly what you want: you want the total off axis output from 3 to 5 kHz to be a combination of the overly strong tweeter and the underly strong woofer, so it blends better and is smoother than if the off axis from 3 to 5 kHz was dominated by the overly strong tweeter.

By using these low Q xovers, the slight dip from 3 to 5 kHz and xovers that aren't in phase on axis, it won't sound like 6kHz and up is hot, because the room will contribute to fill in the slight 3 to 5 kHz depression.

FWIW, I really can't overstress this too much.

I think this is the "secret sauce" to making a two way with 6" or larger mid woofers sound good in a small to med room. You won't hear it as much in a DIY meet in a huge room, but you would in the real at-home installation.

It's not getting much uptake here, but I haven't heard anyone try it and report back, other than George at North Creek who raves about it.

I think you guys are really missing out. I know if I put it on a fancy web site with graphs etc it would get some traction, but I've put this particular idea out there on the page (since the late 90s!!), it doesn't need the eye candy.

Try it before discounting, I think you'll like it.

Dave

Dave,

I finally got this 8" 2way project back from my friend. This winter I plan to try out the method you are describing. I plan to disconnect the drivers from the passive xo and fashion 2 or 3 different models and apply them to different dcx2496 presets. I then plan to switch back and fourth a few times to see what I am missing.

I will likely try something like 2500hz 2nd order and or a 1st order assymptoting to a 3rd order... per your email.

[DDF]

Re: 8'' two-way any reason why not?

Dave,

I finally got this 8" 2way project back from my friend. This winter I plan to try out the method you are describing. I plan to disconnect the drivers from the passive xo and fashion 2 or 3 different models and apply them to different dcx2496 presets. I then plan to switch back and fourth a few times to see what I am missing.

I will likely try something like 2500hz 2nd order and or a 1st order assymptoting to a 3rd order... per your email.

Brian,

I look forward to your opinion!

I've tried my method in a couple different environments. I have my preconceptions about where it excels, and when it wouldn't, but would rather have them challenged, or reinforced independently. Not only that but I'm hoping it works for you, and results in more accurate and enjoyable speakers!

As mentioned in past email, for the room you'll be doing your listening in I recommend measuring the drivers at 0, 15, 30 and 60 degs plus at the floor notch bounce angle, at standing up angles and at side wall bounce angles and give them strong consideration when designing the xover.

Dave

[brianpowers27]

Re: 8'' two-way any reason why not?

Brian,

I look forward to your opinion!

I've tried my method in a couple different environments. I have my preconceptions about where it excels, and when it wouldn't, but would rather have them challenged, or reinforced independently. Not only that but I'm hoping it works for you, and results in more accurate and enjoyable speakers!

As mentioned in past email, for the room you'll be doing your listening in I recommend measuring the drivers at 0, 15, 30 and 60 degs plus at the floor notch bounce angle, at standing up angles and at side wall bounce angles and give them strong consideration when designing the xover.

Dave

I will honestly admit that is quite a bit more measurement than I usually do. I would really like to create a turntable for this project. WHen I measure, I will start a thread to combine all of the info so others can see what I am doing. I imagine I might have time in a few months. Right now I am in the midst of a few non-speaker projects.

[DDF]

Re: 8'' two-way any reason why not?

I will honestly admit that is quite a bit more measurement than I usually do. I would really like to create a turntable for this project. WHen I measure, I will start a thread to combine all of the info so others can see what I am doing. I imagine I might have time in a few months. Right now I am in the midst of a few non-speaker projects.

I know the feeling.

The additional measurements are quite easy, once you have the set up its just a few more angles.

For the vertical measures, lay the speaker on its side and turn them into horizontal measures.

They don't have to be at the exact angle unless you plan on going back later and verifying the work, but even then close is good enough. We used a turntable at work but I personally never felt it would help with hobby measures.

I put together a very simple excel spreadsheet that inputs the driver layout and the mic distance and then it spits out the mic height or position for various angles, and calculates the position of the first boundary reflection in the impulse. I found it useful for knocking measures off quickly. Send me a PM if you'd like a copy.

In the first stage driver characterization of any design, I'll very often take up to 50 measurements, free field over many angles along with distortion, near field, near field distortion. I find it very helpful during later voicing exercises.

Dave

[ctbowman]

Re: 8'' two-way any reason why not?

Yes, yes, yes... AUDAX!!!
That's what I'm looking for now... Audax [and Vifa] 6.5" and 8" midwoofers!!!

GC

check out zalytron.com for audax drivers on sale.

[ctbowman]

Re: 8'' two-way any reason why not?

If you were on acid.....
:rolleyes:
Wolf

Or if you suffer/benifit from synesthesia.