Large midrange for OB??? Scott G ?

MBK · 2007-02-09 9:29 am

BTW so I did an IMD comparison between 15ND930 and SS8543. I didn't quite know what to choose but thought, 120 Hz is more useful than the usual 60 Hz because it is present both as hum and as music. Else I went with default, i.e. the interference signal is a full 12 dB louder than the test signal (400, 800 or 1000 Hz). I think it's tough test anyway. Result is again, no replacement for displacement. At 90 dB/1m interference signal and 78 dB for the test signal, the 15ND930 was slightly better in all cases, both hovering at 0.9-1.2% IMD or so. But at any higher levels, IMD for the smaller woofer could easily jump to 5%.

Here is the SS8543 at 800 Hz.

MBK · 2007-02-09 9:30 am

And here the 15ND930 at the same.

Paul W · 2007-02-09 1:59 pm

HOM supposedly occurs only at high levels, so the pulse comparison is a fairly stressful test because pulse is so loud (probably lethal for dome tweeter). For a while, microphone was a concern. Dunno SPL, but it is louder than driver is ever used for listening :xeye:

. There seems to be correlation between waveform distortion and energy storage which is one of the things I need to explore further.

Have you compared 2-tone distortion for the 15ND930 and SS8543 with both signals near XO frequency for 2-way with waveguide?

MBK · 2007-02-09 4:20 pm

Yes, I have. The nearfield FR 15ND930 starts showing aberrations starting at 1.2k so I kept the highest test at 1000 Hz, attached.

(The farfield is OK until ca. 1.5k when the breakup takes off and I would think 1.2-1.3k XO is feasible, but the testing is harder to do there because farfield introduces a lot of "special effects" so I kept it at 1k).

MBK · 2007-02-09 4:24 pm

SS8543.

Bottom line is that IMD just like THD for both drivers is fairly constant over a wide range, and depends mainly on level. And here the 8543 suffers first, and in IMD more severely, than the 15ND930.

BTW I fiddled with CSD but I don't think it is feasible unless I use a large, special baffle. THe usual time slices are in the ms range, and that is exactly the time frame it takes for the dipole rear wave to come around the front baffle. So I get practically no decay at all (unless, quite possible, that I did a major blunder).

nullspace · 2007-02-12 9:19 pm

Really, REALLY great thread...

As far as combining a 12" or 15" midrange with a compression driver, what are some feasible options for horns/ waveguides which will afford a smooth transition in directivity in the 1.2k to 2khz crossover range?

Obviously, the DDS waveguide seems reasonable, but how about one of the ubiquitious round Tractrix horns? John Sheerin posted some polar response maps, and generally characterized them here: "Based on a few simulations, it looks like tractrix horns with no mouth termination (like Bruce Edgar makes) have a slightly bumpy power response going from ~60 degrees at relatively high frequencies (10x the flair frequency) to 100-120 degrees about 1 octave above the flair frequency. Around ~1.5x the flair frequency, the pattern narrows a bit and then opens way up to be around 280 degrees at the flair frequency - almost omni-directional."

So, any thoughts on whether or not one could crossover from a 12" midrange to a 1" compression driver on, say, a 800hz tractrix at 1.6khz and have some continuity in directivity?

I think that probably the poor power response at higher frequencies with tractrix horns might be a problem, but again I'm not sure. Any thoughts?

Regards,
John

MBK · 2007-02-13 2:37 am

Interesting point and good link info...

There are two ways going about this:

- trying to match the theoretical dipole pattern, for maximum power response flatness. This means, crossing anywhere up to the point where the woofer/mid loses its -6dB @ 60 degrees pattern. Say for my 15" this would be ca. 1200 Hz. The Tractrix seems to have too narrow a pattern, but the Le Cleac'h might do the trick

- accepting to take the woofer even higher, and match radiation patterns in a range where the woofer itself becomes directional. Geddes does this in the Summa, in fact he uses a large woofer specifically for polar control in a monopole setup, and thus getting directionality at a lower 900 Hz. For my 15" dipole setup however this would not be a good option, because it holds 60 degree off axis pattern right into 1.3-1.4k when the breakup modes set in. In Paul's case, the 12" being better behaved, he can pick and choose which radiation pattern he wants to at HF.

But other things should be considered as well. The main advantage of low loading, shallow waveguides (and specifically, either Geddes' oblate spheroid or Peavey's quadratic) apparently generate fewer "higher order modes" (HOM), i.e., internal reflections, which Geddes blames for generating hornish sound and reduced openness. The trade-off of lower loading should be acceptable for a domestic setting.

Paul W · 2007-02-13 3:24 pm

Thanks for the links!

When simulating Tractrix with Hornresp, my conclusion was that they beamed too much at HF...of course many people use them so they can't be too bad.

Here are a few shots from this weekend comparing the DDS Eng 90 with the 12"/$7 MSC waveguides modified for the BMS 4552ND bolt-on drivers. Don't know what profile the "cheapies" represent, but it seems like an exponential flare in a round format. (Note that the throat is actually about 1 1/8" so there could be a throat mismatch with 1" screw-on drivers.)

You can draw your own conclusions, but directivity for a 1.2-2k XO seems adequate.

First, the cheapie 12" at 0-15-30-45-60 degrees, 0.5M from throat.