Tweeter Selection for B&W Matrix 800

[dosik2]

It is difficult to measure large speakers, but you are primarily interested in the response of medium and high frequencies, and that part is not large. The tweeter and two mid drivers are relatively close. You can measure at 1m in the axis of the tweeter, at 1.5m, at the listening position, out of axis, etc. You have gating in the measurement to turn off everything below 200-300Hz to make measurement more realistic above 1kHz, etc. I say again, download Holmy Impulse, it's the easiest for the first time. REW is much more complicated because there are many options.

Thanks a lot for help. I will keep your instructions in mind.

 

[dosik2]

Third option: Improve the compatibility of your chosen amp and speakers. This could be achieved by an impedance correction in parallel to the speaker. Maybe the aforementioned RC circuit is already enough. Maybe not. Measurements and some learning may be needed.

Regards

Charles

I see. I will take note of your comments too.

 

[dosik2]

If the sensitivity of the mids is actually above 93 dB
If the specs account for peaks or trying for best published
numbers. Which is likely.

It would be minimal to no padding needed for the tweeters

Luckily the documentation is good from the pdf posted in post #91
The mids are already padded very slightly. So the factory crossover
is already padded and has component locations to fine tune any padding
needed.

Look as if one mid is crossed lower , which is good for vertical.
It could be somewhat not as painful to fix the high Q on the high pass.
and change the lowpass to blend with, new tweeter added.
Least it is not a complete shot in the dark, no reverse engineering is needed
the documentation is good.

View attachment 1460439

Thanks for your explanations.

To be honest, I cannot yet understand 100%
what you explained. But, very helpful for me.

 

[Lojzek]

This modification should reduce the tweeter level some 2dB assuming amplifier output impedance in the range of 2 ohms,
even more the greater the OI.

 

[Freedom666]

r1 could be bypassed with 0.47 or 1uF more linear above 15khz

 

[Freedom666]

maybe r1 could be 3.3 ohms instead of 2.7 ohms without much changing the filter curve.

 

[temp25]

This modification should reduce the tweeter level some 2dB assuming amplifier output impedance in the range of 2 ohms,
even more the greater the OI.

View attachment 1460458

Just out of curiosity, what would my estimate of a 4uf cap, and 12 ohm resistor look like? I was going for about 8 ohms up top.

 

[SBA_SLOB_GUY]

Hi there,

the thread is getting very interesting, i really like the original PDF from B & W for the affected speaker model

the fact that some other professional builders are offering modifications for thie speaker suggest that there is a pay off to try

my last suggestion for a more non technical person for this is to start to train you ears with really decent near field monitors on the desk, like the Genelec's or the K+H studio monitors

before i end up in buying measurement equipment i used this simple "trust your trained ears" approach for years with some small Genelec's at my desk

but on the other hand the measurement equipment opened to me a completely new more reliable approach

other idea is to borrow a pre amplifier that offers a good Dirac Full Range signal correction integration and to take the measurements with this tool

i used the miniDSP Flex 8 DL in a 6 way open baffle / multi horn proof of concept and was astonished to hear the result, i would be never able to tweak the system this way by hand

good luck!

- Stefano

 

[Lojzek]

Just out of curiosity, what would my estimate of a 4uf cap, and 12 ohm resistor look like?

 

[temp25]

Thanks for doing that.

 

[rabbitz]

I am generally of patient nature. So, I usually change
just a little bit and then hear the sound for weeks or
months to determine the results.

That's what I do as I make one change at a time and live with it for a while before proceeding to others.