hi-vi b3s bargain line array

[gonar]

thinking it might be fun to make a line array of 8 Hi-vi b3s's wired to result in 4ohm nominal resistance. (series of 2 groups of 4 wired to result in 8ohms)

wondering if it might be possible to re-use much of Zaph's b3s network (but only one per group of 4 vs one per driver)

http://www.zaphaudio.com/audio-speaker18.html

assume I'm keeping the same baffle width and volume per driver.

I figure the pair would cost around $200 in parts...

EDIT:
wow, I got bitten by the bug. I'm finishing up my second sub this weekend, have a pair of Zaph BAMTMs and a pair of BAMTM derived MTs on my plate already and now I'm thinking of building a line array pair.....

[joecarrow]

Provided that the filter "sees" the same impedance as a single driver, and provided that the baffle width is the same- the idea has some merit. Unfortunately, there are some issues with running 3" drivers as a line array. You're likely to run into comb filtering, meaning that your high frequency extension will be limited.

The reason why I say it would be OK is that essentially you could just build 8 identical speakers and stack them on top of each other, wiring them series-parallel to get the proper impedance, or you could look at all of the parallel crossovers and come up with the equivalent circuit to feed all the little woofers.

Call it a fun experiment, and don't fret too much about sound quality. There are some really good texts out there on line array design, but the links escape me at the moment.

[aprilia88]

You will need at least one tweeter or alot of eq. If you only have one tweeter and not a line, you will need a way of adjusting the output to match the linesource at the desired listening distance. I have built a few different versions of b3s lines. I can tell you that routing out the backside of the holes is a major pain. But it is good practice to mess with simple contour filters to help eq the line. Mine with 16 per side needed to have negative baffle step because of the high qts of the b3s. While the b3s line is not going to be reference speaker it will show you the effects of a line source and imho the amazing imaging and dynamics. But I think most people would agree the b@g rd 75 is best choice for a line source speaker. good luck,

[clearwaterms]

I had at one time investigated the possability of using the tang band drivers to get a 8 driver line source...

I posted over at DIYaudio asking and got alot of information that it was in the end a bad idea.

Here is the thread over there

http://www.diyaudio.com/forums/showthread.php?s=&threadid=92048&highlight=

here is a project from parts express that uses a Tang band 3" full range in a 32 driver line array. It was designed by Dr. K, however it's an active cross over.

http://www.partsexpress.com/projectshowcase/Kuze3201/Kuze3201.html

aprilia (sweet bikes, by the way) reading that thread, would you mind commenting on the idea?

The idea was to share ideas from this center channel

http://www.partsexpress.com/projectshowcase/cynosure/index.html

but to use them in a FFFFTFFFF arrangement (ala MTM, but line source wise) the cynosure is a 2.5.5 speaker.

also, what does the crossover look like to integrate a helper tweeter with a full range driver?

[aprilia88]

Thank you, I love it. I need to change my avatar to a pic of the mille. But anyway. Here is my theory, And please take it with a tiny grain of salt. Everyone knows that speaker design is about trade off. I like most people want the most engaging sound possible from my design's, The truth is that your ear will adjust to the sound of the speaker. I am not advocating bad sound, just what is more rewarding, a very big and focused soundstage with huge dynamics or the ultimate in resolution and flat frequency response. Or both. If money wasn't an object we would all have accuton and BG or some crazy something. But with a normal budget these are pretty cool. Until these my best speaker was a two way with a 5-880 and sd25lr2 at 2.2k. These lines destroy them in detail, soundstage, and dynamics. I haven't spent the time to try for a better freq response from the lines, which is not to bad really. These speakers just really engage me. They have incredible detail and an effortless sound. I am also sure that one of the rick craig designs would be much better. With a lot less quirks. These lines will work to about 4k before you have problems with combing. I have tried three different configurations. 16b3s with one nd20 on top, one with the nd20 in the middle of the array, and 16nd20 next to the b3s line with 1.5' spacing in between the tweets. I though the tweeter on top looked the best and sounded the same as the tweeter in the center which looked wierd to me. And the line of tweeter was not close enough to each other to make it worth it. The high frequency combing was very obvious. Oh ya, I have been using 2 hi-vi m12's as bass drivers with good results. No issues there, just big bass.
All one has to do to match the point source tweeter to the line source mids is use a l-pad network to set the levels to match at the listening spot. I would not recommend using any less than 16 per line. If you are not well into the middle of the line you will here a mess of problems. The other problem with using that many b3s drivers is that the qts is very high so high that I had to use negative baffle step. so I am going to try using the b2s instead to help the comb filter problem. Then the cabinet could be about 4'' wide. Which will look pretty cool.
If you build them and want better, use those as surrounds and build a dayton rs/bg combo. but then your a foot wide not 4 or 6 inches. Trade offs. Trade offs Trade offs.
and dayton gives good deals in quantity on these. just call. 50 might be cheaper than 32.

[gonar]

decided to put this idea on hold till I finish my 3.5 pending projects (finish daughter's sub, build 2 MTs derived from Zaph's BAMTM (Zaph will _not_ be held liable if they suck, I'm going "off the reservation" with these), build 2 Zaph hi-vi b3s, build BAMTMs

sooner or later, my buddy with the awesome set of tools is gonna throw me out of his workshop and I'll have to buy my own tablesaw.

[Xmax BR]

Dears,

http://www.audioroundtable.com/GroupBuild/messages/2253.html

http://www.audioroundtable.com/GroupBuild/messages/2249.html

And for inspiration...

http://home.pacbell.net/lordpk/speaker/

This is mine (USD 250,00):

http://www.mrweb.com.br/fvm/laf2/

Tks,

Renato

[aprilia88]

Gonar, sounds like you have a pretty long list of projects. I think that you will have a very strong group of speakers to judge the line against. Don't judge the b3's single zaph design with to much weigh, other than detail, because when you have 32 of those little babies going, they sound very big. I know that you can get the b3s's for way under 300$.
I think any one who wants to experience a line source should start with something like these because the cost to performance ratio goes up very steep after these. I built my just as an experiment and know I am stuck in line source world. Just ask Jim Holtz. He has posted a beautiful speaker on this forum. That he really seems to like.
sorry guys my camera SUCK'S!!!!!!!!!!!!!!!

[gonar]

aprilia- care to post plans for that? which tweeter and what did the xover look like?

[clearwaterms]

aprilia- care to post plans for that? which tweeter and what did the xover look like?

I would be interested as well, if your offering

[aprilia88]

these were all built using the beringer 3400 for modeling. sorry I don't have any values for a passive x-over. But I started using just a 6db at 4k. right out of the pe catalog charts. I liked them the best at 3k and 100hz with the beringer.
I just gave the beringer away with an open baffle project to someone in my family. After I get my new dongle I'll do some measurements and work something up. these are so smooth up to the breakup at 8k you have a ton of options with the x-over. I'll try LR2 at 3k with a neo3 pdr in the center of the line.

[clearwaterms]

these were all built using the beringer 3400 for modeling. sorry I don't have any values for a passive x-over. But I started using just a 6db at 4k. right out of the pe catalog charts. I liked them the best at 3k and 100hz with the beringer.
I just gave the beringer away with an open baffle project to someone in my family. After I get my new dongle I'll do some measurements and work something up. these are so smooth up to the breakup at 8k you have a ton of options with the x-over. I'll try LR2 at 3k with a neo3 pdr in the center of the line.

so what your saying is that your passive up to the tweeter, but your active down to a subwoofer? Were you running a subwoofer line array? or just a single sealed subwoofer? also at 100Hz did you have any issues with the little 3" drivers? what about localization?

[aprilia88]

Active at 3kand 100hz with a simple capacitor resistor to compensate for the qts induced resonse hump from 200-500hz. with two hi-vi m12's vented and tuned to 28hz. They integrate really well, and get pretty loud. This setup is very far from finished. I have a 18month old little boy and a 1 yr old buisness. Both require my full time attention. So progress has been slow. Maybe I just need more coffee.

[gonar]

how would I calculate the sensitivity of a line of these wired for 8 Ohms? (tweeter matching)

[Dennis H]

how would I calculate the sensitivity of a line of these wired for 8 Ohms? (tweeter matching)

According to Zaph's measurements, these are 8-ohm nominal drivers with an efficiency/sensitivity of 79dB. So the sensitivity of the line would be:

79 + 10*log(Number Drivers)

The longer version for calculating any number of drivers with any nominal impedance is as follows:

Base efficiency (1 driver, 1W, 1m) = Base sensitivity (1 driver, 2.83V, 1m) + 10*log(nominal impedance / 8 ohms)

The effective efficiency and voltage sensitivity gains from running multiple drivers can be calculated as:

Efficiency gain = 10*log(Number Drivers)

Sensitivity gain = 10*log(8 ohms / nominal impedance of series-parallel combo)

Net efficiency of system = base efficiency + efficiency gain

Net sensitivity of system = base efficiency + efficiency gain + sensitivity gain

[gonar]

10 * log(16) = 12.0411998
79+12 =91, most tweeters are ~92 so fairly close, but I understand that line arrays behave differently in terms of volume attenuation over distance vs. point sources.

if I were to build a model like the one above with a single tweeter at the top of a line of 16, I would need to match volume at distance (8-10 ft).

[ciberproton]

Hello, I'm Trying to built my first array speaker, I was thinking in a desing using full range drivers, my option was 20 x B3S per chanel, modeled by a ultracurve pro 2492.
Reading your opinions above I found one problem: the B3S high qts.
Zaph has another test for full range drivers in its blog:

http://www.zaphaudio.com/blog.html

I think the A3S is very interesting considering its Qts value an smoth response,

http://www.zaphaudio.com/temp/A3N-TS.gif

http://www.zaphaudio.com/temp/A3N-FR.gif

The distortion was very good too:

http://www.zaphaudio.com/temp/A3N-HD.gif

the Spl was not high but using 20 per chanel I think won't be a problem.
What do you think about?
Regrads
Fran

[PoorboyMike]

I was also thinking about making PE an offer on about 32-40 or the Hi-Vi buy outs. Not really sure if I want the 3S or the 3N though. Any opinions on which one would be a better option?

[ciberproton]

For an array I think is better 3S cause the square shape of the basket.
If you make an offer to PE I'll be happy if I can match with you (to make double items): I'm from Spain you'd have to send the items overseas. what do you think?
Regards
Fran

[JonP]

For an array I think is better 3S cause the square shape of the basket.
If you make an offer to PE I'll be happy if I can match with you (to make double items): I'm from Spain you'd have to send the items overseas. what do you think?
Regards
Fran

Not exactly sure, but I think the spacing would be the same. If so, the B3N looks nicer, especially flush mounted. I've used both in a couple of pairs of Zaph's B3 mini speaker. No, I'm not trying to torture you with 40 or so extra rabbet cuts... :twisted: Really I'm not.... The B3N supposedly had lower Qts, but I recall little difference between them. Of course if there's a significant minimum space difference... the lower one wins.

If the A3xx buyout that had the rather low Qts really is that low... it would be nice.

[ciberproton]

Thanks JonP for your awnser:
If you look closely to the manufactures expecifications you can see taht 3N takes 83mm for the diameter and the 3S has78 mm per side of its square faceplat (taking 20 units per side you get 10 cms more for the same cuantity drivers). And you dont need to put the 3S flush mounted (this is very important).
As you can see in the links above Zhap has mesurements of the A3S and the Qts is (more or less) as the manufactures says.
But I have a doubt the A3S has a hole in the frecuencie response around 15000 Hz Will it be posible to arrange that with equalization?
Regards

[PoorboyMike]

For an array I think is better 3S cause the square shape of the basket.
If you make an offer to PE I'll be happy if I can match with you (to make double items): I'm from Spain you'd have to send the items overseas. what do you think?
Regards
Fran

What do you think it would cost to ship that many drivers from the US to Spain? The package would be quite heavy!

[ciberproton]

Asking for the 3Bs (they don't have the A3S) in madisound shipping cost is 153 $.
I've bought lots of drivers to guys in the states, and the perfect shipping (time/cost ratio) was always USPS.
Regards

[JonP]

Thanks JonP for your awnser:
If you look closely to the manufactures expecifications you can see taht 3N takes 83mm for the diameter and the 3S has78 mm per side of its square faceplat (taking 20 units per side you get 10 cms more for the same cuantity drivers). And you dont need to put the 3S flush mounted (this is very important).
As you can see in the links above Zhap has mesurements of the A3S and the Qts is (more or less) as the manufactures says.
But I have a doubt the A3S has a hole in the frecuencie response around 15000 Hz Will it be posible to arrange that with equalization?
Regards

I made an edit to my comment, because I checked Zaph's data, but it didn't make it onto the board, somehow.

Yep, the A3S does look much better, with a measured Qts of .49 and a Fs of 100hz... as well as the closer spacing.

As for the rolloff at 15Khz... how good are your ears? Just kidding.... Really, why aren't you thinking about a separate tweeter(s) to handle the high end? Then any problem up there goes away....

[ciberproton]

Thans JonP:
I have an old proyect in my head and papers: a 3 ways array with 8xRS125 and 4xplanar tuiters (maybe Dayton too) in open baffle-dipolo disposition, and a bass module with and RS315 in closed box. But I want to built and easy array first, in order to know the array sound and to learn some tips. then I'd preffer one simple proyect to start.
In other way, with only one driver for all the frecuencies, you'll have less problems for xover and phase and, I have to admit it, is more "solid concept" more "rotundo" more conceptually simple than a 3 ways.
Regards