Stupid Cheap Line Array

WR,

http://www.diyaudio.com/forums/multi-way/99960-my-line-array-project-log-7.html

Here is the page that will show you what 49 of the Apex Jrs do when wired 7 series/7 parallel. 95dB from 5KHz to 16KHz with a 5dB down point at 90dB. Considering one tweeter is 90 to 93dB across the 5KHz to 20KHz band, it is not a large rise... I'm going to lower the impedance of my lines to bump it up to around 97dB which will work for me.

Since the tweeter lines don't have deafening output, I don't think running one 12 feet tall will increase output much--the mids would overpower them.

Greg, your RTA chart show a gradually increasing output with your 14 3" array then compare it to the bumpy output of a single driver. This is making the large assumption that your RTA picture was in 5dB segments?

If you change the filters for the mids to 2nd order, does that hump still stay there?
 
WR,
Greg, your RTA chart show a gradually increasing output with your 14 3" array then compare it to the bumpy output of a single driver. This is making the large assumption that your RTA picture was in 5dB segments?

If you change the filters for the mids to 2nd order, does that hump still stay there?

yes, 5dB segments. The bump is increased by about 5 dB. Keep in mind that there's an L-pad in there, so where the peaks match, the shoulders are rounded like a bodybuilder's. The filter's curve is the darker, smoother line. Notice how it doesn't look like a table but a hill, and see how the peak between 2-3k is increased greatly by the filter. It sounds like the bump in the fr of that driver is electrically driven and the x-over designer is picking up on that. Again, I can work around it, but I'm having to use a 5k LR4 low pass on the mid and a 7k LR4 high pass on the tweeter in order to keep FR flat.

Greg
 
Quick update - all drivers are on hand. I whipped up a couple 8 unit micro-lines with the 16 ohm NS3s (cute little guys) nudged the EQ on the top end (DEQ2496) and let em run open-baffle. I gotta say, this is a pleasing, natural mid-range sound. I also brought in a few Eminence APT200s to play with. I'm thinking of centering it in the line of NS525s and seeing what happens, forgoing the line of Audax Jr.s, altogether.
 
Good to hear you like the sound of the NS3s

A box appeared under my tree and it smells like NS3's to me :hohoho:

So are you going for a single line of the 5.25's with a line of NS3's next to it then squeeze in your bullet tweeters as needed? A much, much easier build than messing with the little tweeter lines--that would get old wiring in 24 feet of them. :eek:
 
Hey congrats on that "new driver smell!"

It's not so much that I mind wiring and drilling for the tweeter line, it just seems like a compromise that doesn't need to happen with the 3s. I'm certainly going to give it a listen several ways. The little butt-cheek horns showed up busted, literally shattered, but man, do they ever match the scale of the NS3. So I gotta try it.

I'd like to think that 12 of the 525s and an APT200 would play nice together without needing to head up to the 3". Lots of options on the table. I'm making 3' test modules out of everything. Funny thing, the 3s already have started to sound very different than 4 hours ago. Much more relaxed. Just a good, solid sound. Fun stuff.
 
Good to hear the 3s are breaking in nicely,

Since it is winter in the garage--I'll make a cardboard box to lay the 3 inchers in
to break them in by groups of 8--hook em up, run some music through them and run back into the house.

Give the 5.25s a shot with the tweeters--ya never know. Always an option later to run the 5.25s with the bullets along the line somewhere with the 3s sitting next to them.

I've already drawn up the 3-way line box--they have swelled to 17.5" wide and I'll mount the Audax Jr 48 tweet line from the side of the box with hinges. If I need to move the things, just swing the tweet line forward and grab the handles on the side of the box. The boxes are 13" wide so that will work.

Might look weird when I move them but with 12 five inch woofers, 21 three-inch mids and 48 tweeters in each box--looks are rather bizarre anyway. :wchair:
 
FYI: published impedance specs are way out.
NS3 measure 13.6 /published 16.
NS525 measure 6.6 ohm /published 9.

If you are reading it with an ohm meter, the DC resistance (RE) will be lower than the AC impedance (Z)

The NS3 does not have an Re rating but the NS525 does at 7.4 ohms. Generally they are rated +/- 10% so it is close to spec.

To get a wild guess at the impedance of a speaker, I just take the resistance with an ohm meter and multiply it by 1.2 or 1.25 to get close. 6.6 x 1.25 = 8.25 ohms so it is close.

Ohm meters put out basically a 0 Hz DC signal, impedance is AC and varies by frequency so that is why you get readings that you mentioned. The main thing to remember is not the ohm reading but the difference in ohm readings between each speaker.

If you want to see the resistance fluctuate....lightly press on the cone with the ohm meter on....it will jump. This also generates electricity so don't go crazy with it but something to notice.
 
Now the fun starts!

Passive crossovers are sort of a dark art--a million ways to do them wrong and only a few ways to get them correct. They operate on impedance at a specific frequency which varies on speakers, the first step is to figure out the impedance at whatever crossover frequency you choose.

There are several ways to figure out impedance, one is to use the charts that Aurasound publishes that show what they do at specific frequencies. The second way is to use an impedance meter and run test tones into the speakers already wired. The third way is what I do... I made a "jig" which is a piece of speaker wire, a 4 ohm 10 watt resistor and a 10 watt variable resistor that spans 0 to 100 ohms.

This is how it works, I connect the speakers to the amp and run a test tone through the speakers and get a reading from my volt meter and adjust the volume to say around 2.000 volts. I then connect the jig to the amp and run the same test tone at the same volume...I adjust the variable resistor until the meter again reads 2.000 volts. After it does, I disconnect the jig from the amp then ohm out the resistor/variable resistor and get a reading... say 6.92 ohms. That is the impedance of the speaker at that specific frequency.

The reason I have a 4 ohm resistor wired in series with the variable resistor is so if I turn the variable one all the way down to 0 ohms or shorted, the 4 ohm resistor is series gives a minimum of 4 ohms to the amp to prevent damage.

After you get the impedance of say... 6.92 ohms at 4 KHz for instance, then do the same thing to the tweeter at 4 KHz and get, for instance, 9.56 ohms.
You can calculate the crossover with online crossover calculators to get the proper capacitors and inductors.

Another thing that is very helpful is to have a noise dosimeter (sound pressure meter) It will give you the output of the speaker at the crossover point so you know how loud the woofer line and tweeter line play. If the tweeter line is 6dB louder, then you can "pad" it down 5 or 6dB with power resistors to make it smooth.

Line arrays are a mess when it comes to impedance and actual output so the best way is to wire them up, test the impedance and test output. You're going to see some really crazy stuff--I did so am going with the 3-way array with 3" mids to solve the issues.

On paper, I'm very close with 6 ohm line woofers crossing to 7 ohm line mids and 4.5 ohm line tweeters...but the sound pressure meter will give me the truth. If the mids are too high in output, I can cut it a bit by changing the impedance higher to 12 ohms to bring it down.

That is why you do the impedance testing last--first have to determine the output of your lines at various frequencies--use at least third octave sweeps and watch the sound output meter to see if it jumps with a very wide swing. I'd do single tones at 1/3rd octave, write them down and do gradual sweeps to see if it rings at certain points. If it does, take note of that and generally avoid it by crossing lower (or higher) to avoid those spikes.

Looking at the frequency response charts on the NS525... it would be OK to around 1 KHz but rises to a peak at 1.6KHz then falls back down before having huge swings of 8 dB at around 3KHz.

The NS3 is much smoother, the steady increase in output over 1KHz is cut by the 3 dB per octave drop inherent with arrays with a 3.2" center to center spacing. The NS3 has a 7dB rise from 1 KHz to 5KHz or 2.33 octaves. The array effect will cut output by 7dB as the frequency rises from 1KHz to 5KHz (in theory) and as tested in reality, the output was flat from 400Hz to 2.5 KHz then rise about 3dB to 5KHz. Not an issue in reality.

Adason tested his 14 speaker array using the NS3 paper cones, it was very even up to 10 to 12KHz. The best match would be a line of NS525's then cross them at around 500Hz to the NS3s. You have an option to use a horn tweeter at 5KHz or bullet tweeters at around 8KHz or so.

For PA purposes, I'd run the 3s with the 5.25 inch woofers--they will blend quite nicely at 500Hz. I'd then run a horn tweeter with very wide coverage at around 8 feet up--angle it properly then throw another array on top of it. Adjust the crossover to what sounds best and most even be it 5KHz, 8KHz or whatever works.

Sounds like a lot of work, it is! But once everything gets along, they will sound decent, have huge dynamics, low distortion, be very loud and look very, very different from typical sound systems.

Will it all be worth it? That depends on you! I'm moving to revision #3, cutting more wood, doing more impedance testing, designing more crossovers and pondering a Zobel network on the mids (I have output to burn) Once complete, my buddies will come over--crack a cold one in the garage and give their complaints, questions or (hopefully) praise.

First song is The Siberian Orchestra--Wizards of Winter.... just to show the dynamics.

Good luck!
 
I'll go next then.

I'm starting to worry about the power handling of the 5.25's as a single line. My idea has been to run a simple WMT line, but considering WMTW for low end power handling. I've been abusing my test drivers and found that I don't like the level of distortion that these things produce when they push more than a few mm. They can move a lot, but they sound bad doing it. Also, when my left driver (not sure if it's a one driver issue or if it's a left channel is louder issue) hits excursion limits, it puts out a LOUD pop. My only guess is that the VC is bottoming out. If 16 drivers per side did that, windows would break, ear drums would rupture, no doubt. I do plan on doing the huge active x-over bass boost thing, which means that the drivers will really only be ~70 dB/W (-16 DB) except will be at high excursion.

I guess I could always build a 2nd line and enclosure later on and re-wire to try to get a different impedance... With some of you having tested at least partial lines of these guys (or similar speakers), does it sound like I'm underestimating them? If I ever did go passive, a 2nd line would be useless as I'd already have both lines ~101db. I'd need subs, but that's a fair tradeoff, and will work well if I build the lines to be 8ohms. I'd go 4 ohms, but I figure stupid cheap line arrays should be able to be pushed by stupid cheap amps too, right?

Greg
 
Greg,

I'm not running the NS5.25s but use 12 of those Sony 5" woofers with neo magnets in a sealed and stuffed box. Roamed out to the cold garage and fired up "Wizards in Winter" and cranked the old Pioneer to clipping on the meters (around 80 watts) and the woofers moved maybe about 2mm peak to peak. Well below their rated Xmax of 3mm or 6mm peak to peak and there were no issues with distortion. The Sony's are rated for 10 watts RMS and 20 watts music--I would say they rated them conservatively. Granted, they did not have thundering bass but the mid bass was smooth and then it rolled off going into the sub bass bands.

Just a weird thought, keep the WMT lines on the front panel but add the second line on the sides as a "baffle step correction" or 0.5 woofer? Roll off the things at around 200Hz or so with a coil so side mount won't create issues?

My NS3's are sitting in a box under the tree...I'll get to play around with them after Christmas and start to break them in. Then it is hit or miss with the weather when I build new bezels and see if I can get 48 tweeters, 21 mids and 12 woofers to play with each other. Figure I'll have it all built, tested, crossovers configured and ready to go by mid-February.

About how many watts go into the NS525s when they start getting weird? I'm assuming you're using 8 of them and they are rated for 20 watts each...
My grand plans involve the "loudness" switch to bump up the bass of the 3 way arrays, then build a proper tapped horn as the year wears on.

Nothing quick and easy with arrays--back to pealing the onion...;)
 
Greg,
...cranked the old Pioneer to clipping on the meters (around 80 watts) and the woofers moved maybe about 2mm peak to peak...The Sony's are rated for 10 watts RMS and 20 watts music

Just a weird thought, keep the WMT lines on the front panel but add the second line on the sides as a "baffle step correction" or 0.5 woofer? Roll off the things at around 200Hz or so with a coil so side mount won't create issues?

About how many watts go into the NS525s when they start getting weird? I'm assuming you're using 8 of them and they are rated for 20 watts each...
My grand plans involve the "loudness" switch to bump up the bass of the 3 way arrays, then build a proper tapped horn as the year wears on.
Nothing quick and easy with arrays--back to pealing the onion...;)

Hmm.. These are rated 20W, and I'd say it's maybe about right. I think the killer is that bass boost, but again, I'm in this about 70% for the smooth bass integration (30% for watching Inception). The excursion looks like they're jumping out at you, but again, I get maybe +/- 2-3 mm at the MOST before they start sounding distorted. That said, they look like they're going to jump off the baffle just before they start popping. And no, I'm only running one per side (see my fancy shmantzy test rig attached). I've since replaced the Focal 6.5" with the Aura 5.25. I like it. The Focal is better, but 4 times the price and is port/large enclosure hungry. I'm currently running an old Carver 125W/channel power amp and the NAD pre-amp can't be turned more than 1/3 before the woofers clip (those were some good thrift scores).

I really like the 3.5 way, especially considering the bass boost will be applied actively but the 0.5 way passively with only a coil. No major passive x-over issues there. Then again, if I'm going to have 10.5" wide worth of woofers, why not just get some of those MCM 8 or 10" woofers with much more excursion, a lower natural frequency, and overall cheaper? This way I won't have to cut 15-20dB off of the meat of the curve, maybe 5-10.

Greg
 

Attachments

  • 2011-12-07_20-24-53_456.jpg
    2011-12-07_20-24-53_456.jpg
    315.1 KB · Views: 808
Nice rig, Greg

My 1988 Carver M1.5t bit the dust last month--great amp and after the butt kicking it received (pro sound and sub amp duties) I'm surprised it lasted over 23 years.

The 0.5 woofer would be a good trick if you had a box of the woofs laying around, WR has a ton of the things. My array boxes are built complete with flip out side handles, changing the woofers is not an option. I'll cut out the tweeter line and attach it to hinges on the side of the box so I can flip them forward to access the side handles when moving the things for BBQs etc. The mids will take up the space with the 5" woofers so that was my last option.

It would look interesting to have a 12" woofer line array to match the 3 way line array-- a dumpster diving special version of the 1980 Infinity IRS V ???
 
18,

Wow, I had seen the IRS V's once before, not realizing what they were. I like the IRS Betas. Probably not as good, but cool as heck looking. You're right though, bargin basement IRS series.

I think I'm going back to modular. I can always switch out or add boxes later on and rewire. I was hoping to do some sort of cool corner enclosure that touches the back and side walls has an angled baffle, but oh well. I need to remember that line arrays this horrifically big aren't really going to last long in my house regardless (as it will ward away women, and because vinyl siding isn't particularly sound-deadening) and is more of a learning tool. If I decide to pick up some 8's, 10's or 12's (none of which would comb below 1kHz), then I can always plug those in. I'll make sure to make everything 8 ohms on its own, and WORST case I'll end up with 4 ohm or 16 ohm pairs of towers.

Greg
 
Greg,

Never actually seen the IRS V but there was a huge audio place in Germany I visited. They had what we called "The you can't afford it room" with the best from Infinity, Polk Audio, Klipsch, JBL, KEF, B&W and others. Went in, put our blindfolds on and each of us would pick out what we liked. My picks were #1 the Infinity Beta with #2 a tie between KEF 107.2 and Polk Audio SDA SRS. That day proved I had very expensive tastes and like "big sound".

Basically, your array consists of Onkyo tweeters, NS3 for mids and NS5.25 for bass? Now to find a 12" woofer that will go low and be loud enough to keep up with the other lines. I'm not going with a line array of woofers, although it would be cool to see--I'll go with a tapped horn 12 that is almost the same height as the arrays...close enough! :cheers:

I think I'll call my 3-ways IRS V but not for Infinity Reference System... more like Idiot Recycling Speakers revision 5. :rolleyes: