[Rudolf]

BTW: The 7" midrange driver in my own dipole speaker has 5,5 mm.

What is it?

http://www.monacor.de/typo3/index.php?id=84&L=1&artid=4084&spr=EN&typ=full
This driver is in the budget league - stamped steel frame, but very good vented and open to the back. At least it still is manufactured in Germany.

Thanks. Assuming I did the math right back when I built my speakers, I figured I'd be limited by the 3.5W output of my SET amp before I'd hit the excursion limit on my Audax, especially given my high-ish 500Hz XO frequency.

But I agree with you, and if I were starting from scratch I'd use a different midrange driver. I only used this one because I already had it.

To give a better comparable picture WRT Orion and NaO midrange I have to add that those drivers are crossed at 120 Hz. This certainly demands some excursion. One octave up (250 Hz) I still see the demand for more than 1 mm. Another octave higher (crossing at 500 Hz) and restricted to 3.5 W the excursion limit of the Audax could be ok. I guess you are listening at quite moderate volume levels.

Coming back to my initial argument: If the useful range of the Audax starts 2 octaves above the "competitors" I would expect it to be useful at least one octave higher than those competitors - which the Audax isn´t. It needs to be smaller in size to achieve that.

[ultrachrome]

I'm guessing that Saurav's baffle is 10" wide.  The spreadsheet predicts a max SPL of 120dB @ 500Hz.  But a quick reading of the accompanying text, John uses steep slopes to further protect the midrange.

I have a mid with 2.5mm of excursion but unfortunately it's a 10".

[Saurav]

I guess you are listening at quite moderate volume levels.

And in a fairly small room.

I'm guessing that Saurav's baffle is 10" wide.

Close I think it's 11"... the waveguide was a little over 10" across.

The spreadsheet predicts a max SPL of 120dB @ 500Hz.  But a quick reading of the accompanying text, John uses steep slopes to further protect the midrange.

I vaguely remember a number in that range too, and it'll take a lot more than 3.5W to get there. The active filters between the mid and woofer are LR4. So... not ideal, and like I said, if I started again from scratch (or decided to make some changes) the first thing I'd do is select a different midrange. But I already had the Audax, so I kinda tried my best to design the rest of the speaker around it.

[ultrachrome]

Looking at max SPL of my woofers, if I want to use a U frame up to 500Hz, I can reach 110dB as low as 60hz.

If I where to use my 10" mid, I could crossover at 200Hz and reach 110dB at 40Hz with a 30cm depth.

By those numbers alone, I'm leaning toward the 10" mid.

Currently I'm listening to Ellis 1801s with a 10" Velodyne servo sub.  The bass performance of the tapered H was much more to my liking but I have yet to try and model it for comparison purposes.

[pedroskova]

Looking at max SPL of my woofers, if I want to use a U frame up to 500Hz, I can reach 110dB as low as 60hz.

If I where to use my 10" mid, I could crossover at 200Hz and reach 110dB at 40Hz with a 30cm depth.

By those numbers alone, I'm leaning toward the 10" mid.

Currently I'm listening to Ellis 1801s with a 10" Velodyne servo sub.  The bass performance of the tapered H was much more to my liking but I have yet to try and model it for comparison purposes.

I take it that the spreadsheet being discussed is the Linkwitz maxspl...or somesuch.  If so, it should be noted it overestimates excursion because it doesn't take a ground plane into consideration in its calculations.  The presence of your floor gives you an additional ~6dB to work with.  On top of this, you need to remember that you are listening in stereo; therefore, you can add another 6dB of headroom.

If you are using a U-frame, where you only need to worry about one pipe resonance (in the back), you might think about using something like rockwool to help damp the resonance out.  Even then, I wouldn't recommend trying to reach 500 Hz. I could be wrong...I've never tried going that high.

[ultrachrome]

I take it that the spreadsheet being discussed is the Linkwitz maxspl...or somesuch.  If so, it should be noted it overestimates excursion because it doesn't take a ground plane into consideration in its calculations.  The presence of your floor gives you an additional ~6dB to work with.  On top of this, you need to remember that you are listening in stereo; therefore, you can add another 6dB of headroom.

If you are using a U-frame, where you only need to worry about one pipe resonance (in the back), you might think about using something like rockwool to help damp the resonance out.  Even then, I wouldn't recommend trying to reach 500 Hz. I could be wrong...I've never tried going that high.

I'm using John Kreskovsky's ABC Dipole.  I may try Linkwitz's to see how it compares.

John K's documentation does detail the damping of the u-frame which I'm planning to follow.

With a 500Hz XO, I'd use a 12cm deep U-frame which puts the first dipole peak above 700Hz.

Why do you feel 500Hz would be too high?

[Saurav]

I thought the Linkwitz spreadsheet had a comment saying something like "If you're in 2pi, add 6dB". Also, he's estimating the excursion/output of his midrange, which won't benefit from the ground plane as much as his woofer(s) would, I think.

Why do you feel 500Hz would be too high?

I was going to question using a U-frame up to 500Hz as well, only because I vaguely remember seeing comments about now using one if you had to go any higher than 2-300Hz. But maybe if the frame is shallow enough, this won't be a problem. I didn't really dig into U-frames very much.

[pedroskova]

I take it that the spreadsheet being discussed is the Linkwitz maxspl...or somesuch.  If so, it should be noted it overestimates excursion because it doesn't take a ground plane into consideration in its calculations.  The presence of your floor gives you an additional ~6dB to work with.  On top of this, you need to remember that you are listening in stereo; therefore, you can add another 6dB of headroom.

If you are using a U-frame, where you only need to worry about one pipe resonance (in the back), you might think about using something like rockwool to help damp the resonance out.  Even then, I wouldn't recommend trying to reach 500 Hz. I could be wrong...I've never tried going that high.

I'm using John Kreskovsky's ABC Dipole.  I may try Linkwitz's to see how it compares.

John K's documentation does detail the damping of the u-frame which I'm planning to follow.

With a 500Hz XO, I'd use a 12cm deep U-frame which puts the first dipole peak above 700Hz.

Why do you feel 500Hz would be too high?

Sorry about that, my leathering brain can't remember what I've read the day before.  A 5" U-frame is nothing: if you want a 500Hz x-over, go for it.  I was stuck in my own little world, where a U-frame is considerably deeper.   As far as excursion limits go, I do believe that people lose too much sleep over it.  Of course, you have to design within the constraints that physics throws at you. However, once you come close to meeting them, forget about it.  110 - 120 dB limits seem to be in vogue these days...which is kinda sad. 120 dB may cause permanent hearing loss with short exposure (measured in minutes). 110 dB is just a little louder than a jackhammer at 1m.  Can I produce 110 dB at my listening position while listening (from a realistic front row perspective) to a classical recording? Yeah...but its duration is measured in milliseconds.  If it happens to be a 30 Hz tam-tam walloped by the Babe Ruth of tam-tam players, my woofers may be in peril, but mostly, that transient is going to be much higher in frequency, and the biggest danger will come from a (solid state ) amp melting down.

Sorry if I led you astray...I'll try to reread the thread before posting my diatribes again.

I thought the Linkwitz spreadsheet had a comment saying something like "If you're in 2pi, add 6dB". Also, he's estimating the excursion/output of his midrange, which won't benefit from the ground plane as much as his woofer(s) would, I think.

You may be right, but I've seen dozens of examples of people fretting over whether or not their driver(bass) could possibly work... based on his spreadsheet.

[ultrachrome]

Without any appreciable design experience (I've built speakers designed by others), I'm pulling some requirements out of thin air.

Obviously, I won't be listening at that level.  John K.'s accompanying documentation sets 110dB as a goal in his examples.

I figured that it would be a safe number to use when trying to determine if I'd be running into excursion limits of my woofers.

I should spend the next couple weeks putting some test baffles together while waiting for my mic to get back from calibration.

[pedroskova]

John K.'s accompanying documentation sets 110dB as a goal in his examples.

I figured that it would be a safe number to use when trying to determine if I'd be running into excursion limits of my woofers.

It's not a bad goal, especially if you're into headbanging electronica...whatever that is.    With other music, it gives you enough headroom to stay away from power compression: the biggest fault of the hoity-toity  Laura Ashley boutique speakers of the highend. 

[ultrachrome]

Last week I found some time to make some test baffles to test out two different midranges.  I'll be pitting the PR170M0 against the JBL 2123H.  The JBL has more xmax and a lower Fs but it's Sd is > 2x that of the PR170.  I had them so I thought I'd give them a whirl.

I've been spending the last couple of days following the ABC Dipole guide and mastering SoundsEasy MLS Lab (pun intended).  I've discovered all sorts of unexpected behavior so I have been measuring and remeasuring until I've gotten to the point that I can connect a driver, sit down and whip out sane looking impedance and FR plots.

The ABC Dipole method has you take a gated 1m FR measurement and then a nearfield FR measurement.  After spl matching and sprinkling with some delay you end up with a predicted dipole response.  You then export this as an FRD file and import it into the spreadsheet.  I can wiggle the values to bend my FRD to match the target response fairly well.  However, you then need to recreate the EQ in SoundEasy's CAD Schematic Editor.  This is the part I'm currently mired in.  It's not very clear how John translated his excel filter values into SoundEasy.  I can get what appears to be reasonably close but then there's no direction of how the Frequency / Time Plot Control works.

I'm going to beat on it some more.

Now some data...

Here's some exported SPL data for my neopro5i, pr170m0, and the 2123h.  I think the relative SPLs are accurate but I could be wrong.  The absolute SPL is not correct.  Each of these drivers is in the vicinity 100dB/W.



For fun, I just measured my old tapered h-frame woofers.  I'm not sure why it doesn't roll off below 50Hz and this was my first attempt at measuring them per ABC Dipole method.  This is with the mic positioned between the woofers at the mouth of the h-frame.  At that height, the h-frame depth / C = .94ms which was added to the rear FR measurement.  I reversed the polarity of the rear FR and summed it with the front.

If you look in the top right corner you can see how the front and rear FR's looked prior to summing.  The rear measurement looks more dipole-like than the front.  Again, this was my first attempt of this measurement so perhaps something was missed.

[John_E_Janowitz]

hi ultrachrome.  Have you heard anything about Larry Selmer lately?  I hadn't heard from him forever so was thinking something must have happened to him but never had it confirmed. 

I like the looks of the original system yet.  What could be really cool would be to do similar, but make the baffle similar in shape to the new Salk speaker with the open baffle mid.  Sort of an arched shape.  Then make the sides out of multiple layers of bending plywood or masonite.  Similar to what you have now but curved in both directions. 

John

[ultrachrome]

Hey John, I haven't heard anything from Larry in about 3 years.  Hope he's doing okay.

This is a sketchup of my current test baffles.  This particular one is mounted with a JBL 2123H.  The second one has a slightly shorter midrange panel due to the smaller PR170M0.

Compared to my original baffles, the bass section is straight-sided and the wings do not rise past it.  The wings are conceptual.  I haven't done much simulation to compare them to a traditional H-frame layout.  Ideally, I'll test u-frame, h-frame, and tapered variations but right now I'm focused on learning soundeasy and designing a passive crossover for the mid/tweeter.  After that I'll do try to test some different woofer alignments and see what looks best.

btw, really looking forward to your new 6.5's!