The making of: The Two Towers (a 25 driver Full Range line array)

If I had the time to do it right now I would build Ron some new drivers with much more excursion so he could have his cake and eat it to on the low end.

I think I would love that! Sort of like the Audience A3 mini-monster drivers with 12 mm x-max :eek:.
attachment.php


Too bad their distortion graphs don't really show they are worth their hefty price tag. I believe you can't buy them anymore anyway.
 

Attachments

  • AudienceA3.jpg
    AudienceA3.jpg
    164.3 KB · Views: 464
Peter,
I'm sure you are familiar with the Scaena line arrays.
attachment.php


Did you ever try to cross over like they did? If I recall correctly they don't have any crossover on the woofer line's top end and use a single cap for the tweeter line. At least that's what stuck in my mind. They do cross a lot higher than 2 KHz, they let the woofers top end roll of by itself (if I'm remembering that correctly). They do use processing on the big bass units.
 

Attachments

  • dsc_4218.jpg
    dsc_4218.jpg
    295.7 KB · Views: 315
My first conclusion; Yikes! That exit looks blocked!

My thoughts regarding allowing my bass cabinets to roll off at the top end on their own; Would that not create all kinds of interference patterns in the 150 to 1000 Hz range?

I am afraid to admit the some 'high level' sessions have resulted in having to raise my bass to mid crossover point to 150 Hz. The 18 inch RCF Bass drivers still have some significant output at 1 k. Now that I think about it, a steeper crossover would allow me to lower that crossover point. My mid drivers were relatively inexpensive; consequently, they have an X max of only 3.6 mm and 35 Watts RMS. The mids are being driven by a Phase Linear 700B. I have had to replace several drivers before raising my X-over point. Damage was a result of over-excursion, rather than thermal overload.
 

Attachments

  • DSCN0464.jpg
    DSCN0464.jpg
    449.8 KB · Views: 201
I would have thought that above a certain frequency, for a given cone spacing, the drivers begin to act as point sources, producing a narrowing beam with increased F. These beams would have spherical wave fronts and generate peaks and cancellations.
It is my understanding that when the driver separation is equal to, and less than one half wavelength, the drivers act together as a 'line source' and produce a cylindrical wave front.
It would be interesting to drive your line arrays with (8 kHz) sine wave to note any volume fluctuation along the vertical axis ( ie. slowly standing up). This was the case with my less expensive mid drivers. I am guessing that higher frequencies were mostly radiated from the centre of the cones and dust caps. To a certain degree, I thought this was a characteristic of most cone drivers.
 
Peter,
Your choice of crossover point of 2khz seem sensible to me. How the tweeters and cone drivers interact at that distance and with such different radiation patterns is a question I wouldn't try to guess at. You may be better off with a much steeper xo to limit any interaction between the two devices. As far as letting all those 5 1/2" cones to go without a low pass filter I don't think I agree with Ron on that point as I would think those cone drivers would have output to a fairly high frequency and the interference pattern would ruin the idea of a close spaced line array in my mind. The difference between a great driver and another driver can make such a significant change in the perceived sound quality that I would think of perhaps upgrading the cone drivers at some later point. I would just use a single unit in a properly designed box to compare different drivers for sound quality before I would purchase a whole slew of them.

Ron,
Yes I know I could build a 3" driver with the kind of excursion you are looking for. The test model I build many years ago now had a 3/4" excursion, I still have a box of voicecoils and a few of the radial magnet sections I used for those test speakers. I think you would forget about needing supplemental bass drivers with that much excursion. A larger amp would be needed but that seems to be a simple thing to change. I do remember how surprised I was with the actual bass output of that little speaker, it wasn't something I really expected. I was just looking at doing a great small desktop computer speaker at the time, it was just one of those things you least expect and then there is that sound. I couldn't stand the Bose speakers and the little Apple,JBL stick speakers that everyone was using at the time, I just couldn't even listen to a computer with those horrid sounding speakers. I never listen to my laptop with the built in Beats speakers, I have to get out my headphones if I have to listen on my computer.
 
Are we back on the effects of comb filtering due to driver spacing?
I won't deny the effects of comb filtering, it will happen in a straight line array.
Look at this graph:
Designarray.jpg


Once we move closer, the difference in distance (driver to ear) to each separate unit will increase and you'll be able to hear the fluctuations you speak off. Move further away and the frequency where you'll notice these effects will go up. Ra7's thread shows us some data on that too.

I don't know how much of this thread you've read, one subject has been discussed at length, even a perfect stereo setup suffers from comb filtering at the ear. In this case you could notice it if you test a variety of sine waves, while listening in the sweet spot. Start at 1 KHz and move up slowly and you will hear fluctuations without moving out of your sweet spot.

Now take your arrays with 2 distinct columns of drivers. There's bound to be a difference if you move your head side to side around the crossover frequency.

How much of this really is a problem is up for debate. I knew this going into this project and based part of my guesses on the work of David Smith when he tested tweeter line arrays at McIntosh (long ago). The paper he wrote on that subject is attached to at least one of his posts here on diyma. Here's his user profile: diyAudio - View Profile: speaker dave

These papers convinced me to try the array as is, while I did have all kinds of backup plans with super tweeters, power tapering, varying crossover schemes based on driver position etc. I never followed up on any of those backup plans due to the success of the sound I was hearing.

I never tested with sine waves. I did do that in my car extensively. You don't want to know what a mess that is! Still you can get a more than decent Stereo performance even in a small space like a Car. That experience led me to believe this project just might work.

I'll think about running the test, but I'm not saying I will commit to it. I pretty much know what to expect. Remove all content below 8 KHz with a brick wall crossover and listen to your favourite songs. Tell me how you like that! :D

I've listened to tweeters only in my car, also Mids only with a band pass. Just to learn.

I would not want to add a crossover to a line of planar tweeters to make up for the possible problem you mention. For it will create new and in my opinion bigger problems when moving side to side.
 
Last edited:
Hi Ron;
Thanks for your patience, and pardon me, I have not read through this extensive thread, but I intend to!
At the risk of testing your patience further, the above diagram showing the distance to the drivers increasing, above and below the horizontal, would this only be the case (radiation path) when drivers are operating as individual point sources? I would have thought that in line array mode, the wave front propagates as one cylindrical front and therefore cancellation between drivers at centre and extremes is not an issue.
Yes, I am concerned that a crossover (point) particularly at the critical 2 kHz range is introducing a number of issues. In use, I have not pin-pointed any issues with nodes on the side to side plane. Admittedly, they may be masked by bigger issues. This is another motivation to consider a steeper crossover.

I think I will relax my "cancellation" issues before I burn any more bridges! You are obviously a huge line array knowledge resource, so again, I apologized for the jagged leading edge of my learning curve,
Thanks, Peter
 
Don't get me wrong Peter, I welcome line array discussions and even other type of speakers. I'd like to see more results of a project like yours. I'm just not really knowledgeable on arrays utilising a crossover as I haven't build any, nor have I ever heard one, except for some PA solutions.

I wouldn't know what to expect from your current lines and definitely would like to see measurements from it. Do you have any?

I have never denied combing to be a factor in line arrays. I knew that going in. What's more important is how does it manifest itself. Is it an issue? I'd be curious to hear Martin's focussed arrays, to be able to compare notes. Looking at his measurements it does seem like a fun project! Even if it wouldn't fit my particular needs. I do hope we'll see more of that too.

I'd also like to encourage you to show us what you get at the listening spot, and maybe more.

I've read all kinds of preconceptions on line arrays like: Listening to Line arrays gives you a larger than life singer, it's like listening to someone with huge lips... and all kind of other horror stories.
This isn't mimicking my experience so far. I may have done things a bit different than most with FIR filtering, in my room a small performance stays just like that, a small performance. Huge orchestra's do sound huge, but not out of proportions.
I'm still struck by the dynamics and the sense of being there with a lot of songs.

Can you describe what your arrays do? How do they sound? What do you experience in imaging, depth etc. And if possible, lets see the measurements, like the IR, FR, hopefully an APL_TDA plot :).

There obviously are different solutions to every problem. Without trying them all we won't know what does what. I've filled these pages with measurement after measurement. Talked extensively about my experiences. Just to give people an idea about what it can do. I've found two people willing to do a review and hope to be able to add more in the future. I welcome people to come and listen.

I'm also open to hear from you, I've probably read most of the same papers you did... Tell us what you like, and don't like about your sound. Or have some others who's opinion you trust describe it. That's what I did, because I'm always going to be biased. I build them after all. It might be possible that I'm my own worst critic though :D.
Bringing in other people to hear what they say is always a bit scary. Most listeners have been nice though. Either they weren't honest to me or I'm at least doing something right. Who knows :).

Bring on your "problem". Lets see why you're thinking of using steeper filters. Do you use EQ and/or other DSP?
 
Last edited:
Ron,
Line arrays in your car, oh my! :D

You would have been weirded out by my monitor system in my car, only in the back deck, no front speakers at all. Reflected off the window except for a small horn loaded dome tweeter facing forward. Tri-amped.

I can't imagine what that would look like, Steven. :) Do you have any pictures?

Ron,
Line arrays in your car, oh my! :D

That thought has crossed my mind :eek:. It's next to impossible in my tiny car though. I was amazed what 'one' can accomplish in a car. That has been a wonderful journey all by itself. Its what started this project.
 
Ron it was long ago in a car I now have out in the garage in many pieces. A 1967 Firebird. The system had a wooden horn loaded 6 1/2" with rear chamber, horn loaded 1" dome tweeters in wood horns and 2-10" Radian Audio pro-audio cone drivers. I had two Monolithic brand car amplifiers and one that I can't remember the name of for driving the bass. The crossover where hand made on perf boards and were fourth order LR. I had a Monolithic parametric eq and a Sony head unit. It sounded like no other car I knew of at the time. It got loud but that wasn't at all my purpose, I just wanted nice clean sound and I hated most car systems with the speakers down by your knees in the doors and rear deck. I didn't want to deal with cutting up the dash and making the front and rear work together, it was simply a matter of getting use to having the sound generated from behind you. like turning your back on your arrays and listening. I wasn't looking to buzz my license plates so no subs in the trunk and nothing having to go through the seats, a stupid idea as far as I am concerned. When I finally get to put that car back together with many custom features and a modified body I think this time I will redesign the dash so I can have a two way system with my 6 1/2" speakers and a Be dome tweeter and perhaps something again in the rear deck. I'm a bit out there I guess in my designs but that makes things interesting.
 
Hi Ron
Just to confirm suspicions of my skill level, I have to confess that I have only really utilized the FR aspect of the REW program (tone sweep and pink noise RTA) Before that, I was using a DOD 1/3 octave RTA.
I have to say that the clarity of vocals (Bette Middler in one case) has had me on the verge of tears! The satisfaction of coming up with something so breathtakingly real after years of speaker swapping and experimenting. I can't take credit for the Dayton Planar tweeters but WOW! Sixteen of these units was a purchase I had to dance around with the wife! I truly believe that huge issue is having a crossover point in the sensitive 2kHz point. As mentioned, a higher slope might help any possible problems. I have built (copied) a JBL theatre sub-woofer cabinet, the JBL driver was 99 db/w @ 1 meter, but cost prohibitive. I chose RCF 18 inch drivers with almost identical t/s parameters but only 97 db/W/meter. This difference force the bass amp (another Phase Linear 700) to run to close to clipping for comfort. This required the addition of a Crest CA-18 for bass duty (1000 W/ch into 8 ohms) This is another on-going saga to modify the amp for home use. I have started a rather extensive thread on this ( Toroidal Transformer Noise ). I have to say that if there is a loud explosion in a movie scene, my only concern now is whether the windows will stay intact!
At one point, I purchased a rather expensive 'digital' equalizer (Rane DEQ-60L). After someone pointed out that a digital source (CD) was going from D to A to D and then A again, I retired this unit and went back to an older analogue EQ.
I am considering the miniDSP 4x10 HD, but have not researched it. At first glance, it seems like the ideal solution.
 
I have to say that the clarity of vocals (Bette Middler in one case) has had me on the verge of tears! The satisfaction of coming up with something so breathtakingly real after years of speaker swapping and experimenting. I can't take credit for the Dayton Planar tweeters but WOW!

THIS is why we do it, right?

After all my experiments with phase shuffling and XTC I decided to go back a bit and just have the mains + ambient on again. Renewed experience, it is clear that while I might have gained a bit on one side (mainly depth) I did lose some details with my digital cross talk experiments.
This basic setup is less confused and more clear, I cannot deny it.
Luckily it's just a flick of a few buttons. Maybe I need a Mohawk :D.

By the way, I may have mentioned it before... the track: Keith don't Go (live) from Nils Lofgren isn't an audiophile favourite by accident.
That track is quite mesmerising, especially the opening guitar part. All those harmonies floating in the air...
 
Last edited:
I love that Nils Lofgren track... one of my favorites to "wow" my friends when they come hear my setup. Another genre, but just as fun to listen to is Yello's Takla Makan

At one point, I purchased a rather expensive 'digital' equalizer (Rane DEQ-60L). After someone pointed out that a digital source (CD) was going from D to A to D and then A again, I retired this unit and went back to an older analogue EQ.
I am considering the miniDSP 4x10 HD, but have not researched it. At first glance, it seems like the ideal solution.

But don't the miniDSP products also go D/A/D/A as well?

Best way to avoid all those D/A conversion is to use a computer and an app like JRiver or Audirvana. Let's all those DSP filters apply on the digital side, before analogue conversion. Connect that to a nice DAC and it will sound great.
 
Last edited:
I am in way over my head here but from my interpretation of the signal path block diagram; Could I not connect my CD player (digital output) or my laptop (via usb) to the DSP. Please correct me but it appears to me to go through only one D to A conversion, right at the outputs.
Thanks, Peter
 
If you are using a laptop, you don't need the miniDSP. JRiver will do anything that the miniDSP will do (and much more). At $50, it's a better bargain than the $500 unit you are looking for.

I've ripped all my CDs over time and al my libraries are now stored on HDs. It is the simplest way to access all my library, without having to stand, find a CD, load it in the player, go back to the couch and press play! Now, just a couple clicks on the wireless mouse and I'm good.
 
Thanks for sharing your build Peter!

Nice system, Peter. I will be following your progress, as I am also curious to see what digital DSP can do to fully integrate your planar tweeter array to your mid-range line. I really like the time coherency and integration of full-range drivers, like Ronald's Towers, yet how close can we get that with a two-way system with the right hardware at the source?

I am in way over my head here but from my interpretation of the signal path block diagram; Could I not connect my CD player (digital output) or my laptop (via usb) to the DSP. Please correct me but it appears to me to go through only one D to A conversion, right at the outputs.

With the Mini DSP product You are considering, You can hook up your CD and DVD players using the SPDIF or Tolink (optical) inputs. If I recall, the USB input is for data only, it is what You use to upload your DSP filters (cross-overs, EQ, delay, etc) to the 172 MHz DSP processor. So You can not stream audio from your lap top to this USB input. But You can use your laptop to create your filters to upload to the miniDSP.

If You connect your CD player's digital ouptut, You would only have one D to A conversion at the outputs. :up:

I use JRiver in my PC, like Ronald and Perceval. The DSP capacity of this software is only limited by your computer's hardware. But there is a lot to learn and a lot of loop holes to jump thru if You want to watch DVD and Blu-rays. (I really hope You watch movies on your rig, You are really missing out if You don't!) Also You will need a multi-channel DAC for your rig. I will not overwhelm You anymore, unless there is in interest. But I can say, it is a good option to consider.

One thing to keep in mind with all these options is the DACs do make a difference. Considering your current options between your existing rig and the miniDSP: some of the differences You will hear will be the difference between the DAC in your CD player (going to your current analog cross-over and power amps) and the DACs in the miniDSP going straight into your power amps. If you want flexibility to choose output DACs there is the nanoDIGI 2x8 B:

https://www.minidsp.com/products/minidsp-in-a-box/nanodigi-2x8-b

I am thinking now that we are aware of your build, You should start a new thread. I do not get to chime in as much as I would like, but I will be following your build for sure :)

Also, You mentioned you did some measurements in REW. We would love to see them. Just be sure to let us know what position You took the measurements. It would be cool to see a drawing of your room also. No hurry, it is fun to dig in a little. :D I am no where near as knowledgable is Ronald and Kindhornman on measurements and material properties, yet, that is what this forum is all about.

Allen ;)