As Earl said, to avoid guessing requires measurement.
I measured an Eminence Alpha 6 in a small box on axis and at various angles off axis.
At 2.5 kHz, there is virtually no difference between on axis and 50 degrees off ("waistbanding"), while at 4 kHz around 18 dB difference.
Above about 1800 Hz, the speaker has no "constant directivity", no waveguide will match this speaker's polar response.
The box baffle size and edge radius also affect the direct radiating speaker's polar response.
Attachments
Last edited:
I agree with you.
Better to try some different size speaker - maybe a 10" or 12"- seems like I have seen that done somewhere before.
And correct....it's on a wide, 12" baffle now as a B&C 6MD38. No phase plug to help but it's cone isn't 'conical' and I suspect (guess) its going to hold pattern to 3khz or so. The baffle being as wide as it is, I could create a shallow guide but I doubt there'd be any real Improvement in directivity at only 12". But the 6" driver allows me more options for a smaller waveguide above....maybe even a horn loaded ribbon, my first plan.
The system is designed modular so it leaves me lots of options with the mid driver in its own enclosure on top of a bass bin loaded with a 15" woofer. I suspect that being to angle the top, HF module will allow me the flexibility to steer the forward lobe and mitigate some of the combing associated with the higher crossover point. That's the plan anyways.
As to measuring the mids directivity, I have far too much ambient noise in my backyard to do this, and no effective in door space either. I'd have to head up to the country and hope the birds aren't in love!
Read Arts post above http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-605.html#post3476425 - you have your answer.
Maybe it has been here before, nevertheless here is the exact equation for OS contour for a non-zero entry angle (phi):
I've derived it for myself to avoid the cut & try approach but it may be usefull for others as well. It really simplifies the process of generating the coordinates. Other variables are as usual.
I've derived it for myself to avoid the cut & try approach but it may be usefull for others as well. It really simplifies the process of generating the coordinates. Other variables are as usual.
Last edited:
More Formula
Here are some additional formula that may be helpful as well.
[1] http://www.diyaudio.com/forums/mult...n-compression-driver-os-horn.html#post2327385
[2] http://www.diyaudio.com/forums/mult...n-compression-driver-os-horn.html#post2340990
Regards,
WHG
Maybe it has been here before, nevertheless here is the exact equation for OS contour for a non-zero entry angle (phi):
I've derived it for myself to avoid the cut & try approach but it may be usefull for others as well. It really simplifies the process of generating the coordinates. Other variables are as usual.
Here are some additional formula that may be helpful as well.
[1] http://www.diyaudio.com/forums/mult...n-compression-driver-os-horn.html#post2327385
[2] http://www.diyaudio.com/forums/mult...n-compression-driver-os-horn.html#post2340990
Regards,
WHG
Hi mabat,
Rest assured - your formula is correct.
I have been using the same expression in Hornresp for years
.Kind regards,
David
Attachments
There may be some truth to the length issue. The longer the horn the lower in frequency the internal standing waves - along the axis - are. HOMs do not travel along the axis.
The HOMs depend on the flare of the device and the compression drivers match, not on its length, so there will be the same level of HOM's in both devices, BUT ... There is a difference in the mouth diffraction and the waveguides ability to control the sound radiation. The larger device has less mouth diffraction and it occurs at a lower frequency so it tends to be less of a problem. The really big difference is in pattern control. The larger waveguide controls to more than an octave lower than the smaller one. This has big effects on the crossover design and frequency location. The larger device just tends to have all of the performance advantages over the smaller one, but is, of course, larger and harder to package.
There are a lot of things going on in a horn, far more than most people understand or acknowledge and all of them seem to be important from a perception point of view. HOMs are one of those things, but hardly the only aspect that matters.
The HOMs depend on the flare of the device and the compression drivers match, not on its length, so there will be the same level of HOM's in both devices, BUT ... There is a difference in the mouth diffraction and the waveguides ability to control the sound radiation. The larger device has less mouth diffraction and it occurs at a lower frequency so it tends to be less of a problem. The really big difference is in pattern control. The larger waveguide controls to more than an octave lower than the smaller one. This has big effects on the crossover design and frequency location. The larger device just tends to have all of the performance advantages over the smaller one, but is, of course, larger and harder to package.
There are a lot of things going on in a horn, far more than most people understand or acknowledge and all of them seem to be important from a perception point of view. HOMs are one of those things, but hardly the only aspect that matters.
I have found it almost impossible to get satisfactory wide bandwidth out of horns. For example, if I could get 100Hz~20KHz with a direct radiating driver, trying to get the similar bandwidth with a horn to have higher SPL capability and a more unified wavefront across the band seems impossible. I have decided to go for a larger wide band driver instead, seems more achievable with a new cone design. I would probably go for horns in a 3 way design.