Author Topic: Line Arrays  (Read 3388 times)

Offline tmazz

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 7371
  • Just basking in the glow of my tubes.....
Re: Line Arrays
« Reply #30 on: May 11, 2017, 07:53:42 AM »
How crazy is too crazy to know I'm crazy?

Rich we all passed that point a long time ago.  :rofl:
Remember, it's all about the music........

Nola Boxers
Sunfire True SW Super Jr (2)
Quicksilver Mid Monos
ARC SP-9
VPI HW-19 Mk IV/SDS/SME IV/Sumiko Blue Pt. Special
EE Minmax DAC/Bluesound Node/Denon 2910
TWL Power Cords/MIT Cables

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #31 on: May 11, 2017, 07:54:15 AM »
If I may add something to the conversation that might help.

I have recently been testing the effects of wire gauge on "sound" of my test speakers. But it has not been one solid wire, but multiple wires of 18 gauge in parallel. The reason will become apparent.

What I have found, which correlates with theory, is that there appears to be an optimum gauge which produces the most accurate/natural sound over the entire audio spectrum.

With my speakers, increasing the gauge by paralleling more 18 gauge wires was beneficial to a certain point. Beyond that point is questionable. Depending upon the speaker type, the optimum gauge will depend upon the speaker type.

Several points are accomplished by paralleling. First, the size is increased, less resistance, so damping factor is increased. A little can mean a lot. The voices will also flesh out due to the lower midrange/upper bass minimum impedance being less important  (cone types with varying impedance), but up to a point.

Besides the low end being addressed, the inductance of the wire at high frequencies is substantially reduced, thus a little more highs. In my case, the inductance is reduced by approximately 10 times.

In conclusion, with too few wires in parallel, the highs can drop off, and the bass damping factor less. With too large of wire, the bass damping and lower midrange may become excessive, causing masking, and more highs.  I have found the right gauge maximizes flatness in response in my speakers.

Cheers and I hope this possibly helps with adjusting your speakers. If it does not, no sweat Rich.

Steve
« Last Edit: May 11, 2017, 08:01:38 AM by steve »
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline richidoo

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 10441
Re: Line Arrays
« Reply #32 on: May 11, 2017, 11:01:18 AM »
Nice post Steve, thanks.

I have 100ft of silver plated copper teflon insulation to use for hookup wire in these arrays. I think it is either 14 or 16 ga.

Each driver has it's own sealed chamber, so wiring passes through pre drilled holes in the chamber separator panels, so no opportunity to makes changes after initial assembly. Attached pic is the wiring diagram and a study to figure out how many holes to drill in each chamber panel. I'll go for a squeeze fit to avoid caulking.

There is a lot of series wired drivers to increase resistance, and a lot of voice coil inductance but it's "the load" so that doesn't seem to bother the sound according to those who have gone before. It will be interesting to measure the array.

Offline richidoo

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 10441
Re: Line Arrays
« Reply #33 on: May 11, 2017, 11:20:36 AM »
Might as well post the final version of the plans as it changed a little.

The tall front view is bottom half of one array. The top half is identical and sits on top, but inverted. Both screwed to wall and wired together.

I changed the plan to reduce the width of the front baffle, to reduce the distance from driver to the change in wall angle, which when too wide affects the sound noticeably, according to those who have gone before. I guess it's like having a concave kink inside a horn, causes a defraction or something. The edge of diffusion "BAD panels" will line up with the baffle edges to eliminate the convex speaker edges. 

This shape allows me to try it in my current listening room also.

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #34 on: May 11, 2017, 04:20:06 PM »
Nice post Steve, thanks.

I have 100ft of silver plated copper teflon insulation to use for hookup wire in these arrays. I think it is either 14 or 16 ga.
................

There is a lot of series wired drivers to increase resistance, and a lot of voice coil inductance but it's "the load" so that doesn't seem to bother the sound according to those who have gone before. It will be interesting to measure the array.

I am not sure that the series wires will be a huge problem Rich.
In my latest experiment, I am using some 3 feet of single 18 gauge wire between xover and mid/tweeters (I don't wish to disassemble the speakers, at least not right now),

and 6 feet from amps to speaker posts, which is what I am paralleling. Testing with just the 6 foot length made a sonic difference with my speakers.

It would be neat to see if paralleling two or more wires makes a noticeable difference verses just one. (I know I am being a little selfish, but I always love an experiment.) (I also wish the modify would last longer so I can correct my English.)

Anyway, looking forward to your results Rich. :)

Cheers
Steve
« Last Edit: May 11, 2017, 04:22:28 PM by steve »
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline richidoo

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 10441
Re: Line Arrays
« Reply #35 on: May 11, 2017, 05:09:59 PM »
I remember that I once paralleled different kind of wires for SCs. I think it was anticables solid transformer wire + 12ga zip cord ("RCA" brand chinese made clear vinyl insulated speaker wire from Home Depot.) It sounded pretty nasty. Either was better by itself, and that's really saying something for the zip cord. :)  I think I never did parallel two runs of the same wire.

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #36 on: May 11, 2017, 08:44:42 PM »
I remember that I once paralleled different kind of wires for SCs. I think it was anticables solid transformer wire + 12ga zip cord ("RCA" brand chinese made clear vinyl insulated speaker wire from Home Depot.) It sounded pretty nasty. Either was better by itself, and that's really saying something for the zip cord. :)  I think I never did parallel two runs of the same wire.


I found the same results when mixing stranded wire with solid wire. Just does not sound right. I parallel the same brand solid wire, much much better, up to a point.

For all DIYers. If one has a frequency counter, signal generator, and scope, or a friend who has the equipment, one can match the inductors to the nearest turn. 0,5% capacitance meter match the capacitors.

(One can order a frequency counter for $7.50, which includes shipping. But one must know soldering and able to identify and match parts to the board labeling. There is no schematic or instruction manual. Minimum signal for the counter is approx. 2vrms and maximum input is 7vp-p.)

http://www.ebay.com/itm/like/122251535734

Just some thoughts that might help.

Cheers and looking forward to your speakers completion and review.

Steve


My ending last post I meant looking forward to how your speakers sound, not necessarily the experiment.
« Last Edit: May 11, 2017, 08:53:04 PM by steve »
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline tmazz

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 7371
  • Just basking in the glow of my tubes.....
Re: Line Arrays
« Reply #37 on: May 12, 2017, 07:51:33 AM »
I never doubled up on any internal wiring, but I can tell you that about 6 years ago I upgraded my speaker cables from MIT MH-750 to MH-750 Shotgun cables. Both sets were late 80s vintage and unlike some of the later models with the word Shotgun in the name, these Shotgun cables were two pieces of the standard 750 cable wired together to form a single cable. I bought them not really because I wanted an upgrade, but rather because some system rearrangements lead to the need for shorter cables and I found these for a good price and grabbed them. And although the orgignal goal was to get shorter as opposed to better cables, I was very impressed with the amount of SQ increase that I got with the Shotguns.

So for what it's worth, at least sometimes, parallel runs of wire can be beneficial, but like everything else in this hobby, YMMV.
Remember, it's all about the music........

Nola Boxers
Sunfire True SW Super Jr (2)
Quicksilver Mid Monos
ARC SP-9
VPI HW-19 Mk IV/SDS/SME IV/Sumiko Blue Pt. Special
EE Minmax DAC/Bluesound Node/Denon 2910
TWL Power Cords/MIT Cables

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #38 on: May 13, 2017, 07:50:57 PM »
So for what it's worth, at least sometimes, parallel runs of wire can be beneficial, but like everything else in this hobby, YMMV.

Agreed T. Another example; if the system were a shade thin/lean, I would try one large gauge wire. This would provide a touch of inductive reactance to lower the signal as it rose in frequency, thus tending to negate the thinness/leanness. It would also allow for maximum bass and lower midrange, adding a little more flesh to the voices.

I know it is kind of a tone control, but with no adjustments on a system, it could be helpful.

Cheers
Steve
« Last Edit: May 13, 2017, 07:59:19 PM by steve »
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline Bob in St. Louis

  • Audio Neurotic
  • *****
  • Posts: 1032
Re: Line Arrays
« Reply #39 on: May 13, 2017, 08:32:05 PM »
So for what it's worth, at least sometimes, parallel runs of wire can be beneficial, but like everything else in this hobby, YMMV.

Agreed T. Another example; if the system were a shade thin/lean, I would try one large gauge wire. This would provide a touch of inductive reactance to lower the signal as it rose in frequency, thus tending to negate the thinness/leanness. It would also allow for maximum bass and lower midrange, adding a little more flesh to the voices.

I know it is kind of a tone control, but with no adjustments on a system, it could be helpful.

Cheers
Steve
This is an awesome post.
An array is on my bucket list, but it won't be happening anytime soon.
Some day, I'll think back to this thread, this post, and won't be able to recall what you said.  :duh
Not a professional speaker builder, just a guy who likes building things.
Thanks for the trumpet Rich!

Offline Tinnitus Towers

  • Newbie
  • *
  • Posts: 3
Re: Line Arrays
« Reply #40 on: May 14, 2017, 04:55:24 PM »
I have recently been testing the effects of wire gauge on "sound" of my test speakers. But it has not been one solid wire, but multiple wires of 18 gauge in parallel.

Steve,

Do you have any measurement results that you could share that could help us quantify the increase in quality that you are seeing with your addition of paralleling 18 AWG wires? Not sure of the magnitude in distortion/frequency changes in your findings.

Basically, wanting to know if it's worth re-wiring 46 drivers in my arrays...

Thanks!

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #41 on: May 15, 2017, 07:54:31 PM »
I have recently been testing the effects of wire gauge on "sound" of my test speakers. But it has not been one solid wire, but multiple wires of 18 gauge in parallel.

Steve,

Do you have any measurement results that you could share that could help us quantify the increase in quality that you are seeing with your addition of paralleling 18 AWG wires? Not sure of the magnitude in distortion/frequency changes in your findings.

Basically, wanting to know if it's worth re-wiring 46 drivers in my arrays...

Thanks!

The only thing I can do is to give some examples of the inductive reactance. I don't have any meters etc that will work with any accuracy.

1 length of 18 gauge solid straight wire, 6 feet long, at 20khz, has an inductive reactance of approximately 0,368 ohms.

1 length of 14 gauge solid straight wire, 6 feet long, at 20khz, has an inductive reactance of approx. 0,3472 ohms.

1 length of 8 gauge solid straight wire, 6 feet long, at 20khz,
has an inductive reactance of approx. 0,3155 ohms.

Notice not much difference when enlarging a single wire. Add one more wire, for two total wires, and reactance approximately halves. Three and 1/3 the inductive reactance. Change length to 1 foot, and inductive reactance is 1/6th. Halve the frequency, the inductive reactance is half.

I have not altered the wiring in my test speakers, approx. 3 feet, as my speakers seem to be very very nice. So it may or may not be worth it in yours. How long is the internal wiring of your speakers? Might try testing with other, simpler but excellent  speakers.

Cheers
Steve
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline richidoo

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 10441
Re: Line Arrays
« Reply #42 on: May 16, 2017, 06:03:00 AM »
Does cable geometry (twisting conductors) affect inductive reactance?

What affect does inductive reactance have on signal passing through the cable?

Offline steve

  • Obsessively Audiophilic
  • ****
  • Posts: 371
    • Reference components
Re: Line Arrays
« Reply #43 on: May 16, 2017, 05:59:10 PM »
Does cable geometry (twisting conductors) affect inductive reactance?

Yes, it will be lower but not zero because the centers of the conductors would have to occupy exactly the same space. Typically, the inductive reactance will be reduced. It depends upon how close the wires are physically.

The closer the wires to each other, the higher the capacitance. The twisted speaker cable is much larger than having the +/- wires separated. If global negative feedback is involved, the feedback signal will be phase shifted, distortion will rise as the frequency rises. How much is dependent upon the amp design. Could be very little, or somewhat.

Quote
What affect does inductive reactance have on signal passing through the cable?

Think of a voltage divider. The higher the inductive reactance of the speaker cable, the lower the signal voltage is to the speaker driver. Of course, phase relationships are also involved.

One thing in our favor is that usually the driver impedance rises along with the speaker cable inductance. Still from my testing, the results are noticeable.

I just reduced my parallel speaker cables from 10 parallel 18 gauge wires to 8 parallel 18 gauge wires. There was a difference.
From a DC resistance stand point, the gauge was reduced from approximately 8 total gauge to 9 total gauge.

From my above post, inductive reatance....

1 length of 18 gauge solid straight wire, 6 feet long, at 20khz, has an inductive reactance of approximately 0,368 ohms.

1 length of 14 gauge solid straight wire, 6 feet long, at 20khz, has an inductive reactance of approx. 0,3472 ohms.

1 length of 8 gauge solid straight wire, 6 feet long, at 20khz,
has an inductive reactance of approx. 0,3155 ohms.

8 parallel speaker wires, inductance is only 0,046 ohms inductive reactance at 20khz. It is more like a direct connection to the amplifier (assuming no xover). And very little, if any cable capacitance if the +/- wires are separated.

Cheers
Steve
« Last Edit: May 16, 2017, 06:17:34 PM by steve »
Pre/Amp/ICs are the only components that can be tested for accuracy/naturalness.

Steve Sammet
SAS Audio Labs
http://www.sasaudiolabs.com/pream11a.htm
sasaudio@omnilec.com
Ph: 309-263-0736
Retired.

Offline richidoo

  • Out Of My Speaker Cabinet
  • ******
  • Posts: 10441
Re: Line Arrays
« Reply #44 on: May 17, 2017, 07:24:26 AM »
Thanks steve.
So in a feedback loop wewant short untwisted conductors.  Phase compensation with cap would not help because it rotates phase in same direction as the twisted conductor capacitance? Usually the ground leg is separated anyway.