Line array problem

[Spartacus]

I just finished a Bose type line array system using Celestian AN2775 speakers. They sound dead, no high end. I used acoustistuff loosely stuffed in the cabinets, could it be causing the highend loss? If not does anyone have any idea's? The frequency response on the 2 3/4" Celestians is 160 to 19k. Thanks

[Jeff B.]

Re: Line array problem

I just finished a Bose type line array system using Celestian AN2775 speakers. They sound dead, no high end. I used acoustistuff loosely stuffed in the cabinets, could it be causing the highend loss? If not does anyone have any idea's? The frequency response on the 2 3/4" Celestians is 160 to 19k. Thanks

Acoustistuff should not impact the driver's high frequency response much. More than likely the drop at high frequencies is due to driver directivity and comb filtering effects between drivers. These can combine to shelve down the top couple of octaves.

[Spartacus]

Re: Line array problem

Hi Jeff, Would you please explain what comb filtering is and what can I do to remedy the problem? Thanks
John

[Sydney]

Re: Line array problem

I just finished a Bose type line array system using Celestian AN2775 speakers. They sound dead, no high end. ...

I worked with a Bose L1 array very recently - be aware that it makes extensive use of EQ.

[r-carpenter]

Re: Line array problem

Are you using any sort of EQ? How many drivers and what's the size of the baffle?

[r-carpenter]

Re: Line array problem

Acoustistuff should not impact the driver's high frequency response much. More than likely the drop at high frequencies is due to driver directivity and comb filtering effects between drivers. These can combine to shelve down the top couple of octaves.

Driver he is using has a 10db bump in top octave (at list according to the published FR). Shouldn't sound "dead" as he describing but there may be diffraction related drop in upper midrange.

[hongrn]

Re: Line array problem

John, take some time to read and digest this article from Dr. Griffin, and you'll gain some pretty good understanding of line array designs:

http://www.audioroundtable.com/misc/nflawp.pdf

[rickcraig]

Re: Line array problem

I just finished a Bose type line array system using Celestian AN2775 speakers. They sound dead, no high end. I used acoustistuff loosely stuffed in the cabinets, could it be causing the highend loss? If not does anyone have any idea's? The frequency response on the 2 3/4" Celestians is 160 to 19k. Thanks

If it's an unshaded array (all drivers equal in output) then the output isn't going to be uniform in the higher octaves, especially as the wavelengths become shorter than the center-to-center distance of the drivers (comb filtering). Another issue is the off-axis response of a driver that size is limited (compared to say a 1" dome tweeter) so the room response will have more loss from the narrrow dispersion in the higher frequencies.

We have a large CBT array at our church using similar size drivers and equalization is needed to smooth out the response. Even at that it still is limited by the horizontal coverage.

[Dwight]

Re: Line array problem

I strongly suspect a wiring or crossover problem (if you used a crossover). These should NOT sound dead or flat. How many drivers per side and did you wire them series- parallel? Array wiring is tricky and there is not much info on the web that actually diagrams how to wire various numbers of drivers (I know of one source). One easy thing to do, although this will not affect high frequency much if at all, is to connect a single battery (AAAA,AAA,AA,C, D) to your speaker wires just to make sure all cones either move in or out together, you should not have some moving in while others are moving out, you could also measure resistance across your speaker input terminals, I would think somewhere between 3 and 15 ohms would be expected. I'm sure others will chime in, the more we know the more we can help.

[Spartacus]

Re: Line array problem

No eq as of yet, I've got two columns stacked with 18 Celestians total. The columns are 4" wide, 5" deep and 30" high.

[rickcraig]

Re: Line array problem

No eq as of yet, I've got two columns stacked with 18 Celestians total. The columns are 4" wide, 5" deep and 30" high.

You really need more drivers than that to get decent performance. I have built arrays about 36" long but with power tapering and 2-ways for surrounds in a large home theater.

[Spartacus]

Re: Line array problem

They are wired in series in groups of three and then those groups in parallel to result in an 8 ohm load for each of the two columns. There is a 99 micro farad capacitor installed inline on the positive wire at the input jack.

[Spartacus]

Re: Line array problem

Thank you I will.

[Chris Roemer]

Re: Line array problem

Those should start "beaming" around 5k-6kHz, but the highs even sound weak "on-axis"?

Not that it's part of your issue, but that single cap can't do much to high pass those near 150-160Hz due to your impedance peak (of roughly 4-5 times nominal) in that range.

[philthien]

Re: Line array problem

This is a fascinating thread to me because I had an unfinished project that used two of those per side and I ended-up selling it off.

Now having completed some speakers with single Faital 3FE22's, I have a different perspective on these small, full-range drivers.

In the case of the 3FE22, the response curve makes it seem like I'd have a pretty peaky top-end when measured on-axis. I can say that is absolutely not the case, they roll-off pretty quickly above maybe 14-16k. I'm still not clear if the difference in high-end response is due to the baffle size or something else.

When I was going to use the two Celestion drivers per side, I was told here about comb filtering and I didn't want to have to experiment with crossovers so I "got out."

But seeing as you've gotten considerably further, what I'd do in YOUR case is I'd try disconnecting everything except a single driver and take a listen and see if it still sounds "dead." If not, then the crowd here is likely correct and you're likely experiencing the comb filtering they're talking about.

Oh and one more thing, I originally started with 3fe25 drivers in different enclosures, and then switched to a new project with the 3fe22. And in doing so, I thought my newer project sounded pretty lifeless. But breaking-in the drivers made a world of difference. You'd think that a new, stiff driver would be brighter but I actually found the opposite to be true.