In Reply to: My my posted by AJinFLA on October 26, 2006 at 18:05:21:
The driver in the RR design looks to be to be the Tang Band W3-881SD. The description and photos look to me to be the same or very similar. They retail at Parts Express for around $16 each so you could build this speaker for about $1100 a pair including 50 drivers, 2 sheets of 3/4" plywood, and a 31 band two channel equalizer. Parts Express closed out a similar model in both 4 and 8 ohm versions earlier this year which originally retailed for $10 for 49 cents each. I wonder if RR got started by buying them out. One nice thing about arrays Dr. Bose pointed out, each driver contributes so little to the overall output, normal individual variations from one driver to another are inaudible in the overall output. This is a very good driver for its type and purpose. It shold have better HF propagation than the 4" CTS driver Bose originally used. Internal enclosure volume looks to be about 2 to 3 cubic feet. One problem with getting accurate bass out of this kind of array is that the small drivers in a small enclosure often result in a fairly high F3. I'd judge the original 901 to be around 250hz to 500hz with a Q well above 1. This creates a peak which must be compensated for with a suitable notch filter. To make power requirements managable for its era, he only provided a 6 db/octave rise in the bass but an AS system falls off at 12 db per octave below F3, so if you want flat bass to 20 or 30 hz, you can double his power requirements. I'd say a good figure of merit for 901/I or II is about 500 to 1000 wpc and because of Xmax limitations, multiple such arrays are usually preferable. The other problem therefore is that while the combined radiating surface area is equivalent to a 15" woofer, the Xmax is only about 1/20 or less of a long throw model, in the vacinity of 0.5 mm. This means that an array this size will be no match for a single high quality acoustic suspension woofer such as NHT 1259. At the other extreme, a 3 1/2" driver will not have very good HF lateral dispersion even by comparison to a 1" tweeter (which IMO is inadequate as well which is why I use arrays.) One reviewer at the manufacturer's website said toeing them in improves their sound. I'd guess the best placement would be in the corners of a room which is around 14' wide by anywhere from 14' to 30' long and aimed in at 45 degrees. This will make the most of both their bass and treble and is what PQ recommends for his speakers for much the same reason. Therefore the design could be improved considerably by using anywhere from half a dozen to a dozen or more such arrays in a semicylidrical or cylindrical arrangement which would mitigate both problems. Is this what you had in mind? Another characteristic of arrays is that they often contain perfect square of integer number of drivers (9 in the 901, 16 in the e/e sweet 16 article and 25 here) which means that because the speakers are wired in parallel units of series strings to manage overall impedence, the effective amplifier damping factor across each driver is less than 1 (in this case 0.25 or less.) That makes performance independet of both amplifier damping factor and wire gage or type (not counting deliberately high shunt capacitance types.) Another nice thing is that with drivers this small, you can probably cut the baffle holes with a hole saw drill bit, no need for a saber saw and Home Depot or Lowes will of course cut the plywood for you if you want them to, often for free. This doesn't look like a particularly difficult DIY project as speaker projects go. Personally, I'd wait to hear them first and then for a markdown at Parts Express or some other supplier.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: My my - Soundmind 05:18:53 10/27/06 (0)