Line Array Ruminations and Questions (long)

[cytokine]

I must be crazy.

I have been reading this forum for a while and have even built a pair of pretty nice two-ways using the Dayton 8” driver (295-310) and silk dome tweeter (275-070). I used Dave Tenney’s MT design and crossover. I like them a lot.

I became intrigued by the concept of the line array after reading about it and have done quite a bit of research before even considering posting here about it. I ruminated on the possibility of building a pair for a long time. I must have come to the conclusion that I thought it might be a good idea to build them because the parts arrive today via UPS from PartsExpress. This one is going to send the wife over the edge. I may be sleeping on the couch tonight even though the cost was not that significant ($76 including shipping).

I have ordered 40 of the NSB midrange drivers by Pioneer (269-568) and 50 of the Onkyo tweeters (269-702) with the intention of building a 16/24 array wired for an 8 ohm load.

I have read everything I can find on these drivers on this forum using as many search parameters as I could think of. I have a bunch of questions that I would like some help with if some of you are willing to stick with me on this project. I will try to parse them out into digestible chunks so that I don’t get confused.

First Question: What about a box instead of an open baffle? I’ve been messing with WinISD a little and it looks like if you make a ported box you can get reasonably good bass response that extends to 50 Hz. That should not be right because the Fs of the NSB is 106 Hz. It shouldn’t sound that good right? Additionally in modeling a ported box the port length and diameter are actually almost as big as the box itself. Has anybody done this and if so what are your results?

I’m inclined to go open baffle because it will be slightly easier and more of a challenge to build with a high WAF (should be fun if I’m already sleeping on the couch…). However if someone has the right box and it makes the crossover easier and cheaper, I would go with that.

It turns out that most of my questions relate to CROSSOVERS, BAFFLE STEP CORRECTION and BAFFLE SIZE

Second question: What about a crossover? I remember reading a thread in which sreten recommended a first order low pass at 600 Hz and a second order high pass at 5000 Hz. His recommendation was an economical one keeping in mind that you could spend a lot on the crossover but that doing so would defeat the purpose of the project which is to build something that is inexpensive and sounds good.

I will be removing the on-board crossover and the phase shield of the tweeter and from the frequency response graph provided by PartsExpress the 5000 Hz high pass makes sense.

I’m less comfortable with the low pass portion. Especially in the context of an open baffle configuration.

Third question: Will crossing first order at 600 Hz not create a noticeable dip in the upper midrange such that it won’t sound very good? 600 Hz seems a little high for baffle step correction, and a little low for achieving a reasonably flat frequency response.

Fourth question: Has anyone designed an inexpensive low pass for this driver that will work well with this tweeter and that has some measure of baffle step correction in it?

Fifth question: How do I build the circuit described in The Edge? I’ve used The Edge simulator to model the baffle size, driver placement, and the overall baffle step loss. I realize it’s not perfect, but it gives me pretty graphs to work with. One of the options is to design a circuit for “Open baffle bass drop compensation”. For those of you who have used The Edge and who know how to read a circuit, how do I wire the components up? The triangle is an amplifier, right? R2 and C are in Parallel with respect to each other and in series with R1. What is R1? Why are there two of them?

Sixth Question: What about an L-pad for the 16/24 configuration? Because these drivers are of virtually the same sensitivity won’t the tweeters need to be attenuated approximately 4 dB? Is that right?

OTHER TOPICS

Seventh Question: Has anyone tried coating the 4” Pioneer drivers with any kind of material to smooth out the response a little? There has been discussion of diluted Elmer’s glue and Dammar varnish that made some sense. The problem with those I have discovered after doing my homework is that both will dry brittle. I remember reading (and it makes sense intuitively) that the coating should provide stiffening but should remain flexible. Given all of this has anybody considered high quality shellac? I understand that shellac dries flexible and remains so. What do you think?


If you’ve stuck with me this far you know for sure that I’m nuts and probably ought to be locked up.

Thank you for your time.

[ScottG]

Generally..

Avoid ported enclosures..

Remember that midrange line will be dramatically boosted below 1kHz.

notice the difference here:
http://www.partsexpress.com/projects.../Kuze3201.html

Don't expect to do this with a reasonably flat freq. response without measuring equipment. Speaker workshop is free but you'll have to make your own jig and mic (which is the reason I purchased the speakerlab mic and software combo).

edit: oops.. thats loudspeakerlab - located here:
http://www.wavecapture.com/
(but it doesn't appear that they sell with a mic anymore..)

[mazeroth]

I, too, would LOVE an answer to these questions, or multiple answers

I'm in the same boat as you, cytokine. I have 96 of the Pioneers and 128 of the Onkyos just in case my brothers want a pair!

[cytokine]

Scott:

Thanks for reminding me of that Tang Band line. I had seen that and I still think it looks sweet. It got me thinking about Tangs for a later array if I can get this one to work well.

Mazeroth:

Let's do this together and see if we can add to the knowledge base of the forum on these critters.

I said I had searched a lot for info on these drivers and line arrays, and I do mean a lot. I'm pretty sure I've seen everything out there. Maybe not though. Today I learned that the NSB drivers might have been made for Sony TVs.

Here's my latest thinking on the low pass portion of the crossover. There will be no impedence EQ in the crossover circuit. Therefore, if I understand correctly, as frequency rises and impedence rises (see the impedence plot for details) the crossover will become less effective. Thus a first order low pass at around 600 Hz may have the effect of smoothing that midbass hump between 300 and 1000 Hz, while it won't take as much off of the curve starting at around 2 kHz.

I haven't committed to a circuit yet and I'm thinking that if I'm going to try to smooth that hump I might want to low pass a little lower, say at 300 Hz. Maybe not. Like I said, I haven't decided.

As someone who has already built these has said elsewhere, to paraphrase: don't dwell on it, just build them and see what happens. In the true spirit of DIY that is what I will do.

As for coating the drivers I'm going to try and find the time during the day tomorrow to get some shellac to coat one of the extra NSBs with to see how that affects the sound.

I don't have any formal measuring equipment but I may get a mic and try to get Speakerworks working for this project.

Oh yes, let me not forget to thank Svante Graqnvist for The Edge baffle simulator and for answering my e-mails this afternoon. He's been very helpful, even with questions that reveal my woeful ignorance of electronics and electricity.

Svante suggests that I need a baffle as large as possible for an open baffle and NO compensation.

Live update: I arrived home to find three boxes on my porch this evening after work. One was quite large and very heavy. I assumed this to be the NSBs. One was quite small and I assumed this was the 8 mids that I ordered over and above the case quantity of 32. The last was quite compact but still large and not as heavy. This should have been the tweeters, right?

Nope. The tweeters were in the largest box and they are very heavy. They look pretty nice for the price and I'll start taking them apart tomorrow. The NSBs are actually built fairly solidly and look pretty good from the front except for the yellow colored face plate. I might make a jig to cover the cone and paint the face plates flat black.

Final thought. It looks like I get to sleep in my own bed tonight. The wife is a wonderfully forebearing soul and was more surprised that the whole load only cost $75, than that I was making them. She's a peach.


Last thing. I think I may just go ahead and cross over as sreten suggested with the low pass at 600 Hz first order and see how it sounds.

I'll keep you posted. This will take some time.

[paulspencer]

Here's a project that might interest you if you haven't already seen it:

open baffle line array speakers with fullrange drivers



Open baffle does sound very promising - why not give it a go. If you do this, then I'd pick a point where you want to cross to subs or bi-amped woofers and make the baffle wide enough so that there is no dipole cancellation in the range you are using them. This is something you need to experiment with.

It's not a good idea to use drivers much below their fs.

I dont' believe vented boxes are evil! You can get a vented alignment to be more linear due to decreased cone excursion. You could tune a vented alignment at fs and roll off the bass with a steep slope where you cross to a sub. Trying to do this with a sealed alignment may mean you will have a lot of rolloff before 100 hz and hence a dip where you cross to your subs.

2 options that stand out to me:

1. open baffle crossed fairly high at say 300 Hz to bi-amped woofers (and possibly subs if you want to go lower than 40 Hz)

2. vented tuned to fs at which point you cross to a sub - the highpass filter is critical to prevent overexcursion

[cytokine]

Paul;

Gorgeous! I'm going to have a close look at the link you sent me as soon as I have a moment.

I appreciate the suggestion about a vented enclosure. The problem I ran into with WinISD modeling 16 drivers was that the recommended size of the port was almost as large as the enclosure and the enclosure was smaller than the volume of the drivers. I must have done something wrong. Any suggestions?

For now, I stopped at the hardware store this morning on the way to work and didn't find shellac. Some store. I did pick up some rubber cement to try on some paper to see if it will dry non-tacky and flexible.

I then went to a craft store and found this:

[IMG]Winsor and Newton Varnish[/IMG] Winsor and Newton Varnish

The Dammar didn't say anything about not cracking and this stuff did so I opted for this.

I appreciate all the help and I will be sure to share the results of my experiments because it seems like there are a lot of us who have yielded to the temptation of these little drivers.

Final thought: If coating the NSBs works and it takes out the 7 kHz on-axis peak then I may just run them full-range with no crossover and cross the tweeters as high as I can.

[planet10]

Originally posted by cytokine
50 of the Onkyo tweeters (269-702) with the intention of building a 16/24 array wired for an 8 ohm load.

Anyone looking for a budget tweeter shouldn't forget about the ApexJr tweeter. At 2 for a buck it certainly qualifies as a Frugal-phile(tm) kinda product. I have been very pleased with their performance... my 1st 50 went 1 pr at a time (mostly SuperTs for 40-1197), i may look at a LA for this batch of 50 i just got.

Tom (IIRC) used them in his LA (thread on this forum)

dave

[paulspencer]

Originally posted by cytokine

I appreciate the suggestion about a vented enclosure. The problem I ran into with WinISD modeling 16 drivers was that the recommended size of the port was almost as large as the enclosure and the enclosure was smaller than the volume of the drivers. I must have done something wrong. Any suggestions?

A vent tuned at 100 hz should be small. Tell us the volume, tuning point and diameter of the vent and we can confirm if the vent length is correct.

[cytokine]

Dave: Thanks for the suggestion about the ApexJr. tweeters. Have you heard the Onkyos? I would be interested in knowing whether one or the other sounds better and what the measurements on the ApexJr.s look like. The price is right.

Paul: I'll try to get back to WinISD and repeat my modeling tonight. I'm sure whatever error there is is probably mine. I'll let you know when I have some information for you.

Today's update is this:

I applied either one or two "coats" of varnish to the cones of two of the NSBs and likewise one or two "coats" of Elmer's rubber cement diluted 1:4 in mineral spirits. I put the word "coats" in quotes (sheer poetry...) because the paper of the cones is quite porous and the material absorbs immediately. This morning there is no obvious difference in the appearance of the four test drivers when compared to the control untreated driver.

I'm hoping to pop them into open baffles tonight to give them a listen.

It looks like I'm going to have to get a measuring microphone and fire up some testing software because the scientist in me (I really am a scientist) wants to know what the effect of these different treatments will be.

[swirv]

I have a few NSB's Ive coated with Mod Podge and damped the basket as well. One of them has a phase plug made from a wine cork, coated with Mod Podge, and painted.

Basket treatment