New Home for the BG Ribbons - Part 27

[JonMarsh]

One thing that comes to mind about the RD50/75 vs. an array of Neo8s is the sensitivity. The RDs aren't all that sensitive but you should be able to juggle your series parallel arrangements if all three lines were arrays so the sensitivities all matched up -- assuming you wanted to build a passive crossover and avoid the emperor's wrath.

Well, I think both Hank and Chuck have decided to go active. I hear the Emperor is a bit old fashion, even if he lives in a technology center of the galaxy. A lot comes down to what kind of target transfer function and overall design is employed, obviously- an 88 db sensitive midrange could be part of a system design pushing close to 90 dB overall, which isn't chump change, considering what a lot of box systems come in at. Sometimes barely over 80dB/watt!

[chasw98]

Chuck, if you've got that kind of money, we need to be friends :^x

...oh no, I just realized what a stupid question THAT was on THIS forum :stupid2: :banghead:
Admin: you might want to sticky that question as THE question to not ask around here ops:

I am your friend. I am the only one that hasn't jumped you for your question! :T

Well, I think both Hank and Chuck have decided to go active. I hear the Emperor is a bit old fashion, even if he lives in a technology center of the galaxy. A lot comes down to what kind of target transfer function and overall design is employed, obviously- an 88 db sensitive midrange could be part of a system design pushing close to 90 dB overall, which isn't chump change, considering what a lot of box systems come in at. Sometimes barely over 80dB/watt!

I am looking into active on the Peerless drivers and then passive between the RD75/Neo8 and Neo3 drivers. OTOH, modern science makes it so easy to perform 'what if' scenarios quite easily in an active environment to see if pursuing the option in a passive build is worth it that I will probably set it up as fully active for a little while. Fully active makes it easy to measure without a lot of wiring swaps also.

[chasw98]

Besides, ask ANY woman, and she'll tell you a line array is ipso facto proof of going excessive....

ixnay on the line array. She knows I am building something and it is tall but I haven't quite explained it all to her yet. Just a little bit at a time so only small anger is spread over a long period of time. :E

[JonMarsh]

Smart man! :T

[Hank]

Chuck, you have a plan and I hope it works. How many men have had "plans" that have been seen through as though they were glass? (rhetorical question).
BTW, I'm not planning on a 3-way at this point. Why stick a super tweet ribbon on top of a true line array?
I"m outta here - going to a wedding in Fort Worth. Might outta take my friends Mr. Smith and Mr. Wesson just in case...
And I promise I won't drink any premium tequila :W

[JonMarsh]

And I promise I won't drink any premium tequila :W

You're going to drink the cheap stuff? Say it isn't so, Hank!

[Amphiprion]

I"m outta here - going to a wedding in Fort Worth. Might outta take my friends Mr. Smith and Mr. Wesson just in case...

Hank, the proper firearm for a wedding here in Texas is a double-barreled coach gun. Surely you know this? I think Stoeger makes some reasonably priced models

And why any animosity on here towards active crossovers? They are so easy compared to dealing with the limitations of passive ones. Are people so afraid of soldering tiny little things together?

[JonMarsh]

Some of us have soldered many tiny little things together, and listened to them, and developed phobic reactions to opamps... there are some fairly good ones, of course.

But anything with an active crossover I can't just plug into my Ayre integrated ampflifier and enjoy. And can you imagine the cost of three pairs of Cardas Golden reference speaker cables? :W

Seriously, there are many paths to nirvana... of course, whether you'll find any of them on this forum is a matter open to question.


Have fund this weekend, Hank! 8x)

[Amphiprion]

developed phobic reactions to opamps

I will take an NE5532 or OPA2134 over passive crossovers any day of the week And I know many will cringe at me mentioning the 5532 in that sentence.

[chasw98]

Some of us have soldered many tiny little things together, and listened to them, and developed phobic reactions to opamps... there are some fairly good ones, of course.

But anything with an active crossover I can't just plug into my Ayre integrated ampflifier and enjoy. And can you imagine the cost of three pairs of Cardas Golden reference speaker cables? :W

Seriously, there are many paths to nirvana... of course, whether you'll find any of them on this forum is a matter open to question.


Have fund this weekend, Hank! 8x)

Which is why us practical people will develop a crossover with discrete components a la Bryston 10B and only biamplify the lower portion of the frequency region. Then I will only need to use 2 pairs of my Cardas Golden reference cables and can use the other long pair for my rear speakers! :B

[Johnloudb]

I was pleasantly suprised to find that my Beige Bag Spice software will design Tube crossovers with it's wizard function. Took about 30sec to design.

This is a high pass LR4 with a 1kHz Fc. Generic tubes! This basically a rough circuit design, and one would need to re-adjust the bias for the particular tube. And design a power supply.

Image not available

This probably won't interest anyone, but if any of you wants a rough tube design, I'd be happy to Wizard one for you. I don't think I'm doing tubes since I've never built any tube gear. But this speaker project will probably take me most the summer, and I may change my mind. Or I may do something like this using JFETs. Or I might use opamps.

Jon,

If you decide to go with the RS270, I've got two of them like new. I tried them in our old IMF Monitors to use as transmition line subs. They worked well but wasn't quite as clean as the original KEF B139 in the lows. I suspect this was due to not being broken in. But I've settled on the RSS 10" subwoofer, which still wasn't as clean as the KEF B139 with it's 3.5mm of Xmax ... break-in? It had the windows rattling though.

Just PM me if/when you're interested and I'll give you a bargain.

[Amphiprion]

Maybe we should try and find some NOS K2-W vacuum tube op-amps

Image not available

[CraigJ]

Might outta take my friends Mr. Smith and Mr. Wesson just in case...
And I promise I won't drink any premium tequila :W

Me thinks Hank was already :nos: . Regarding digital crossovers, I'm 99% there, and having 101% of the fun=100%. :B I'm with Mark....solder what?

Craig

[Dennis H]

I'm with Jon, active is for wimps. Real men do passive crossovers.

Seriously though, there's an engineering elegance to one amp, one wire, one speaker -- IF you've got the chops to massage those pesky drivers and passive components to do what needs to be done.

[Johnloudb]

Maybe we should try and find some NOS K2-W vacuum tube op-amps

Image not available

Cool, I've never seen that before. 8O I have nothing against op amps - our current power amp, an old Boulder 250AE is just a big discrete op amp. It's a great sounding amp too - best we've had in our system. I had planned on using OPA2134 or National Semi op amps for my crossover. I'm kind of thinking discrete JFETs now just for the mid/high frequencies, though I have no idea which would sound better.

Mark, what do you think of Praxis? I been looking for a reasonable audio analysis tool for awhile and I downloaded the software. Do you have the full version? I may get the full version if it's capable of measuring very low distortion. They made it sound like you have to use their hardware interface, and I didn't see any specs on that. Or was that just for the speaker testing? I do have a low distortion sound card. But, it be good to hear your comments about the software.

[Dennis H]

Mark, what do you think of Praxis?

I'm not Mark but, if you want to save some money, check out the ARTA suite (includes STEPS and LIMP). It probably does 90% of what Praxis will do and a few things it won't. The free download will do everything except save files. It's something like $140 for the full version.

http://www.fesb.hr/~mateljan/arta/index.htm

[Amphiprion]

Praxis is buggy and likes to crash a lot. Other than that you can do some cool stuff with it. I bought it years ago, now there are other options I would explore first before laying down the cash.

[Johnloudb]

Thanks Dennis, Mark, ... I check out Arta. :T

[Amphiprion]

Is there an ARTA thread on here or another forum? I downloaded it and am giving it a spin, if for no other reason to have two tools to double-check myself.

[Dennis H]

There's an ARTA yahoo group but it never really took off. The pdf manuals are pretty good, similar to the help files but a bit more complete.

[Amphiprion]

Well I bought all the stuff to make an ARTA box. Should be here Tuesday or Wednesday. I'll give it a spin.

[Dennis H]

I didn't know there was an ARTA box. He just shows some voltage-divider probes in the manual (which I like -- KISS).

[Amphiprion]

I built an input box for LAUD back in the day (sound card protection and 10x/1x switching). This looks pretty similar (ie, easy).

[Hank]

Have fund this weekend, Hank!

Well I didn't get any fund(s), but it was a good trip and I got to dance with my daughters. I was only kidding about the primo stuff, Jon - I only drink premium tequila 8x)

Mark, boy, it looks like you have unearthed a polarized situation here: active versus passive. You young guys are trouble makers As I'm mulling over the aspects of this project, I've also thought about a combo of active and passive xovers. Next time you're in town, we'll sketch up some items over tequila.

[Amphiprion]

We will just have to remember the Ballmer peak...

http://xkcd.com/323/

[chasw98]

Got to keep this thread alive if I am going to get help and keep you all interested.... Hank!

I cut the slits in the test baffle today and drilled, tapped, and mounted the aluminum struts. As I went along I just did some basic 'shake' tests to see how the panel stiffened up. It stiffened up quite a bit. Now when you place it on blocks at either end when mounting the drivers it does not sag. Much stronger. I started out being lazy and just used drywall screws to hold the aluminum to the baffle, but it was only halfway there. When I drilled and tapped, the baffle really stiffened up. Over this weekend I am going to try and get some basic measurements in the 'lab' (my porch) and see what it does. I did hook it up to the JBL horn/Selenium D220ti just so there would be some top end and fiddled with the DCX2496. Not too shabby for just goofing. Did need some low end EQ.

Here are the pics....

A look at the struts.



Strut closeup.



Wide view of struts.



The beginning of 'Stairway to Heaven?'



Front view with JBL Horn



Room view.



Just so you know.... Debbie noticed the tall array being set up and wired today. She said,"So tell me what this tall board with a lot of little speakers on it is?". I answered her,"Do you want to know technically what a line array is?". She said,"No." I said,"It wil sound great when you are watching a 60 inch flatscreen between the speakers on your exercise bike!". She replied,"Great!". Am I lucky or what? :T

[ThomasW]

I'm in shock...... 8O 8O 8O :B

[chasw98]

I have taken some measurements and calculated some gains according to the formulas on Jim Griffin's paper.

Using 12 Peerless 830875 drivers the efficiency gain is 10.79 dB, sensitivity gain is 1.23 dB and system sensitivity will be 100.62 dB. The wiring of the 12 drivers in 4 series groups of 3 each which equals 6 ohms when all is wired up.

Using 8 B&G Neo8-PDR drivers the efficiency gain is 9.03 dB, sensitivity gain is 3.01 dB and system sensitivity will be 104.54 dB. The wiring of the 8 drivers in 2 series groups of 4 each which equals 8 ohms when all is wired up.

Using 16 B&G Neo3-PDR drivers the efficiency gain is 12.04 dB, sensitivity gain is 0 dB and system sensitivity will be 102.54 dB. The wiring of the 16 drivers in 4 series groups of 4 each which equals 4 ohms when all is wired up.

It seems that 100 watts per channel will get you well above 110 dB 3 meters away. Got headroom? :T

[chasw98]

Here are some measurements. These are of the Peerless 830875 drivers in an array. The first set was taken at both 1 cm and 1 m at 90 dB SPL using justMLS included in LspCAD.

1 cm 90 dB unsmoothed



1 cm 90 dB 1/6 smoothing



1 m 90 dB unsmoothed



1 m 90 dB 1/6 smoothing



impedance

[chasw98]

Here are some REW sweeps at 1 m and 1 cm, smoothed and unsmoothed. These are of the Peerless 830875 drivers in an array. These were all done at 75 dB SPL.

1 cm unsmoothed



1 cm 1/6 smoothing



1 m unsmoothed



1 m 1/6 smoothing

[Johnloudb]

Chuck, I'm glad you're pushing ahead here. I need guidance, a leader to follow. I'm thinking if those aluminum strips really stiffened up the board that much, I may try putting two pieces of 1.5" aluminum angle back to back for my final design. Something like this:

Baffle = XXXXX


Front                    Angle             Back
                 ______________ 
                 XXXXX||
                 XXXXX||
                 XXXXX||
                 XXXXX||
                 XXXXX
                 XXXXX
                 XXXXX

Or This:

         ||
         ||      Angle
         || 
         ||
   _______________
   XXXXXXXXXXXXXXXXXXXXXXXXXX
   XXXXXXXXXXXXXXXXXXXXXXXXXX
   XXXXXXXXXXXXXXXXXXXXXXXXXX
   XXXXXXXXXXXXXXXXXXXXXXXXXX
                        
                 Front

It would also be easy to mound some kind of support structure to the angles. Do you find the support stiff enough, or are you planning on something more, like a spine?

[chasw98]

I'm thinking if those aluminum strips really stiffened up the board that much, I may try putting two pieces of 1.5" aluminum angle back to back for my final design. Something like this:

It would also be easy to mound some kind of support structure to the angles. Do you find the support stiff enough, or are you planning on something more, like a spine?

Well, the idea is great and works, the execution leaves a little bit to be desired. I should have listened to Dennis H and tried 1/4" angle instead of 1/8" angle. The 1/8" angle really does stiffen it up quite a bit and it does provide a pick off point that is very sturdy and machineable. I am pretty sure I will be using the struts in some fashion for rigidity and also possibly as a frame for the B&G drivers maybe. Still kickin' that idea around in my head. I have come to the conclusion that to do a spline with anything other than wood will cost a lot of money and not gain much in the way of fidelity. A really cool 1/2" aluminum strut powder coated black and CNC machined with holes in it would be great, but...... the cost is out there compared to a 1 X 8 piece of 3/4" oak from the box store.

Keep coming up with ideas, one of us is bound to find the 'way' to do this effectively.

[ThomasW]

That wide/deep valley centered at 900Hz is confusing. My measurements (stored on a different PC so I can't attach them) don't show anything like that.

[JonMarsh]

Hey Chuck,

Couple of questions- where in the space was the line array relative to nearby boundaries, and what sort of window/gating on the received signal were you using? Some of those look like really long time gates, possibly both pre and post- mostly with rectangular windowing.

[chasw98]

Hey Chuck,

Couple of questions- where in the space was the line array relative to nearby boundaries, and what sort of window/gating on the received signal were you using? Some of those look like really long time gates, possibly both pre and post- mostly with rectangular windowing.

Refer to the picture in Post 206 of the room and the attached array standing up next to the woofer cabinet.

Sample rate was 48,000
MLS Length was 32,768
Offset was close to measured distance from speaker to mic, i.e. 100 cm and got set to 104 cm

Length on nearfield was set to 10 ms, farfield was set to 30 ms.

It is very easy to reproduce so give me suggestions as to what settings to change and try along with physical positioning.

[Jed]

Farfield is usually set around 6-8ms... nearfield you can set it much longer.

[Amphiprion]

I'll bet you $1 if you move the mic up and down the array, the dip will move too.

[chasw98]

Here is 1 meter, 90 dB, 44 inches height, 7 ms window.

[chasw98]

The other measurements were taken at 44 inches tall. The exact center of the array. Here are 2 more taken at 41.5 and 54 inches tall which is pointing directly at the center of a driver.

41.5 inches tall



54 inches tall

[augerpro]

Cool stuff Chuck. I don't know much about LA's but I think you might get the best idea of the response pulling the mic back to the listening position, or at least a distance equal to that. Shortens the gate on an MLS a lot when you do that, so maybe use a combination of MLS and swept sines.

EDIT: I think these LA's demand a number of measurement "perspectives". Swept sine and MLS, short gate and long gate, maybe even one in room and one outside.

[JonMarsh]

well, I'm at something of a handicap advising you, as I don't use MLS or Room EQ Wizard, so I'm not so familiar with your software.

As Jed mentions, a near field measurement can usually use a long window, because the levels nearfield dominate over the room reflections. I'd suggest measuring at about 4" so that the output from the cone can combine from across the cone area.

The crux is, how far from nearby boundaries/walls are you, because even with the short window (shorter than I usually use), you've got a lot of noise pickup- reflections within the window? So for starters, I might suggest measuring with them 6 feet or more from any wall boundaries; I'd also suggest using a Half Hamming window, or Half Blackman-Harris, depending on what your software offers. Measure far field at 18"; this is enough for the baffle conditions to develop adequately, and lessen the impact of the reflected waves.

[Jed]

Mark K dragged his outside. Might be the only solution to get really clean measurements for an array this tall.

[chasw98]

well, I'm at something of a handicap advising you, as I don't use MLS or Room EQ Wizard, so I'm not so familiar with your software.

As Jed mentions, a near field measurement can usually use a long window, because the levels nearfield dominate over the room reflections. I'd suggest measuring at about 4" so that the output from the cone can combine from across the cone area.

The crux is, how far from nearby boundaries/walls are you, because even with the short window (shorter than I usually use), you've got a lot of noise pickup- reflections within the window? So for starters, I might suggest measuring with them 6 feet or more from any wall boundaries; I'd also suggest using a Half Hamming window, or Half Blackman-Harris, depending on what your software offers. Measure far field at 18"; this is enough for the baffle conditions to develop adequately, and lessen the impact of the reflected waves.

justMLS is included in LspCAD. I thought you used to use that... Although it is very simple without many, if any, variables to tweak besides sample rate, gate, mls length, and length of flight time. Certainly nothing as you are mentioning. But, yes, the next step will be to move the array around and see what effect that has.

Just so you are aware one wall of the porch is totally exposed to the outside. Might be the place to back it up against, hmmmmmmm. 8O

Mark K dragged his outside. Might be the only solution to get really clean measurements for an array this tall.

I think you may have a point there. Unfortunantely it is the rainy season down here in Florida and it has been pouring on and off all weekend. I do think moving it outside will ultimately need to be done to get measurements to work with rather than just second guessing at incomplete or flawed measurements.

[Paul W]

Maybe try an outdoor GP measurement with the array laid on one side.

[chasw98]

OK, last of the day. Here is a picture of the array in another location in the room. The rear is totally exposed to the outdoors.

New location.



And here are measurements at 41.5 inches, 90 dB, 1 meter.



And here are measurements at 44 inches, 90 dB, 1 meter.



And here are measurements at 54 inches, 90 dB, 1 meter.



It would appear that both array location and mic location play a significant role in results.