Audio / DIY / Speakers / QNPS-2 - The project: structure and geometry The structure is conditioned by:
- the "needs" of the various transducers in terms of a dipole panel
- NPS geometry, where feasible
- the shape of the panel to minimize risonnaze, diffraction et c.
- Tweeter: given the projected range, whatever the shape of the panel, this would be "seen" by the transducer as a virtually infinite level (relative to the wavelength). Potential effects of diffraction concern insignificant fraction of signal due to the high directivity, I do not think there are particular effects on the overall geometry, apart from the usual adroitness pairing "flat" to the panel.
- mid-high: that is what we saw for the tweeter, subject to any care to a minimum of fitting on the edges of the panels. Given the wavelengths involved siorienterebbe us to a single transducer, ideally, a ribbon of about 10 cm.
- mid-range, the expected distance (two feet) does not present any particular restrictions on the structure. Are probably still present diffraction effects: wanting them to consider the lowest frequency (320 Hz), should be rounded ends with a radius of at least 1 / 8 of a wavelength that is longer than 10 cm!
This is probably the most critical project.
-woofer: cutting back on a regular five-way speakers should cover from 80 to 320 Hz, with a distance of about 2.1 m. The problem is that a dipole classic (albeit with wings) would be very bulky in width to 80 Hz (about 1.3 m) and then you should use some variant of H-frame.
An interesting alternative would be to exploit the presence of the sub-woofer to limit the bandwidth to about 160 Hz, the frequency at which you could still make a dipole efficiently without extra unsightly structures. Nothing prevents naturally compensate instead the roll-off "brute force", especially considering the good results in this sense of QNPS-2.
The problem may be rather to allocate part of the band "voice" to the sub-woofer, with completely different shapes and colorings from the woofer. In addition, a band so important to one eighth compressed may cause the critical calibration. It may then be paying a corresponding "slip" up one eighth of all the higher cutoff frequencies.
Initial size of the panel for a cutoff frequency of 160 Hz, perhaps with a very steep slope, then reserving the right to lower it to the extent allowed by the DSP correction. All this would naturally strong
implications on the whole, would drop the assumption that the "cut-up" to about 1.5 m, freeing up space for the vertical sub-woofer, electronics and reducing the overall development required.
sub-woofer: common ground that would be a single cone, and that should be placed at the bottom (to qualify for the "reflection" of the floor) it would seem a more rational version of the "Ripoli. This would fix the speaker down by limiting the overall size and allowing the positioning of the woofer below.
From a structural standpoint is to realize the subwoofer as a base (perhaps independent) on which to mount the panel.
The geometry must be such as to the issue of frequencies up to 300 Hz in order to take account of "smearing" in the cross-over, and the possibility of having to use the loudspeaker above the frequency "sub" classic.