Clive wrote on Thu, 26 August 2004 09:11 |
I would assume that with having 11 labs in one row you would get nulls by virtue of phase differences.
|
If you place the boxes in a continuous line (i.e. a "line array") other boxes in the array fill in these holes. The interference patterns that create nulls added to places where two or more sources combine to sum positive and averaged together is why a line array is as even as it is and why it drops off on the edges so drastically. (i.e. toward the center of a line array you have more summation and toward the ends more cancellation)
Clive wrote on Thu, 26 August 2004 09:11 |
I would also hazard a guess that this would have the unfortunate consequence of dropping the effective impedance on the drivers.
|
Hmm, I see your point, but I can't quite get my mind all the way around this. When you stack the boxes next to each other you get an impedance change. Just how far apart does this hold? Do you get further significant impedance change after the sources are more then1/3 wavelength apart?
Clive wrote on Thu, 26 August 2004 09:11 |
The fact that the sound improved as cabinets dropped out seems to lend credence to to this speculation.
|
Well I figured one of two things. Either some drivers had their polarity flipped or it was an interference pattern. I wanted him to check polarity on everything because the root of the problem is not obvious. It may have nothing to do with his driver failures.
Clive wrote on Thu, 26 August 2004 09:11 |
An interesting experiment you could try would be to place four labs in a row with a defined amount of space between each cabinet, and then feed them a pure tone. For arguments sake, lets say we use 60Hz, then space the cabinets (center to center) 2.87m apart. Effectively, the cabinets will be 180 degrees out of phase with their immediate neighbors, and in phase (but lagging a full 360 degrees) with any cabinet 5.73m away from it. ie) cabinet A <--2.87m--> B <--2.87m--> C <--2.87m--> D Apply the tone to cabinets A & C only, and measure the current flowing to the speaker coils. Now switch on the power to B & D and measure again. I expect that there will be an obvious difference to the sound as well as a change in the effective impedance of the speakers.
|
This does look worth doing on its own, but considering he had a line array with no gaps I don't believe this will he help him find his problem.
Clive wrote on Thu, 26 August 2004 09:11 |
fwiw) I suggested 60Hz, so you can just use mains voltage (120 Volts), and it goes without saying to wire them up in series. While this experiment may seem a little silly, it may save you time chasing non-existant problems elsewhere. Hope this helps, Regards, Clive.
|
Your experiment is worthy, but not something I would try to find his problem.
Checking polarity on every line and driver is accepted trouble shooting tech, while your experiment is something you might look at after eliminating many more simple issues. Note I do not claim my guess that somewhere he may have something with the polarity flipped is the cause of his driver failure. It's just one of the first obvious things to check. You can not assume anything is correct till you check it no matter how slim the chance when the cause is not obvious. An Eminence employee swapping the red and black terminal on some drivers is a real long shot, but after what happened I would not leave it to chance.
Here is one of my FAVORITES!
I have a large industrial tablesaw. Years ago I bought a Biesemeyer T-Square fence system for it. Cost about $550.00 or so more then twenty years ago. It has a measuring tape pasted to the square tubing the fence rides on. I would cut a test piece, adjust the gauge till it was correct and run some stock to recheck. Usually I would do it with a piece 3" to 6" wide. The trouble was that when I would cut longer pieces the cut would be a just a little longer then it should be. It wasn't much. Perhaps 1/32" to 1/16" or so. I figured the fence gave a little when I cut bigger pieces.
One day I happened to have my scale (markings down to 1/64") laying on the tube and I lined things up to see how well it agreed with the tape. The first 12" of the measuring tape has 1/32" markings and the rest of the tape (to 101") has 1/16" markings.
The first 12" of the tape was WRONG! It slowly changed from the beginning to by the time you were at 12" it was almost 1/16" long. The tape after 12" agreed exactly with the scale. Now that was something I had assumed must be correct for YEARS and it was literally the LAST place I would check for the problem, but there it was.
Too "You can assume all you want till something goes wrong!" Tall