B&W Matrix Series 2 - " How to reduce brightness "

[arupg]

I've a pair of B&W Matrix 802 Series-2 (1988-1993) & they're somewhat brighter than the later version, the Series-3 (1994-1997). Further, the APOC Protection in the Series-2 was dropped in the Series-3 and the crossover simplified with separate PCB boards - HF/MF & LF whereas the Series-2 had a large single PCB for all the HF, MF & LF sections.

Below is the schematic of the Series-2 with Series-3 component values (marked in Green) where relevant.

My objective here is to somehow run the LF a wee-bit hotter than at present (primarily to reduce the brightness factor) &/OR reduce the MF/HF slightly (pad-up) without losing too much on the time alignment or onset of detachment on the whole. In other words, I don't want to wreck the beautiful sound of the speaker!

I would appreciate to have suggestions for you 'Gurus' but I'm not in favor to doing too many changes & would rather leave the originality intact as much as possible.

Your inputs will be much appreciated.



[IMG][/IMG]

[Cal Weldon]

Hi there. If you want to reduce the brightness you add an L-Pad. This can be either with resistors or something similar to a potentiometer. The resistor type can be hidden inside the cabinet along with the XO but is fixed (not variable). The variable type you will have to cut a hole in the cabinet to mount it. Look up L-Pad calculators, there's a few out there.

You can't really add anything to the woofer via the speaker, you can only take away from the mids and tweeter. If you are lacking bass, try relocating the speakers, perhaps closer to the corners or adding bass by using an equalizer.

[arupg]

Hi
Am already running the prescribed B&W Bass-alignment filter that runs the speaker to 6th Order from 4th Order default passive crossover. So, I've that front covered. As for relocation, the positioning was worked out with lots of work from RPG Diffusors Inc and I doubt they would mess up - with their reputation...

So - that leaves me to pad-up the MF/HF. Ok, you say L-pad. Any values to can recommend based on this circuit? Please look at the Blue box.

Many thanks in advance!

[soongsc]

I would recommend measure the frequency response and impedance first. Look at the CSD if possible. Most of the time improvement in significant enough with just flattening out the mid frequency impedance.

[Inductor]

Don't make a hole in the box. Test it outside, and later move it inside the box when (you're) done. One made of fixed resistors or the variable attenuator. A 30W or 50W would be more than enough, I guess, on the tweeter.
Boutique haut parleurs et audio DIY Atténuateurs
http://www.bwgroup-support.com/
B&W crossover rebuild
Note: I would not change the balance in the woofer/mid region if this is all you need. Is possible to change (make it louder) but it requires changing the components. Theres also another thing that might work for you, and did work for me (FOCAL's), before changing the xover. You can "implement/experiment" with black shoulder pads on the tweeter (not good for looks, but hey, easy to do).

[mwmkravchenko]

If I am remembering this correctly is this the one with the kevlar midrange?

I was working on fullrange driver design back then and we had tremendous trouble trying to get Kevlar to behave itself. We called it cone cry and I always remember listening to the B&W's back then and their evil kevlar cones screaming at me.

If you are stuck with them you may need a notch circuit. Looking quickly the next generation seems to have a zobel network of some kind in olay that the schematic in green does not.

A bit tired to really check it out but maybe tomorrow.

Mark

[arupg]

Originally Posted by Inductor

Don't make a hole in the box. Test it outside, and later move it inside the box when (you're) done. One made of fixed resistors or the variable attenuator. A 30W or 50W would be more than enough, I guess, on the tweeter.
Boutique haut parleurs et audio DIY Atténuateurs
B&W Group North America Service & Support Homepage
B&W crossover rebuild
Note: I would not change the balance in the woofer/mid region if this is all you need. Is possible to change (make it louder) but it requires changing the components. Theres also another thing that might work for you, and did work for me (FOCAL's), before changing the xover. You can "implement/experiment" with black shoulder pads on the tweeter (not good for looks, but hey, easy to do).

Thanks for your inputs.

BTW, in this loudspeaker, the LF has a set of binding posts while the Mid/Tweeter is on another set of posts.

Therefore, what I am looking to do is to cut the Mid/HF by max 1dB or 2 may be and that should create a balance where the LF will be louder than the Mid/HF.

Now, this is a powerful speaker system and wonder if the 50 watt L-pad on the combined Mid/Tweeter feed would hold or do I need a 100 watt attenuator?

[arupg]

Originally Posted by mwmkravchenko

If I am remembering this correctly is this the one with the kevlar midrange?

I was working on fullrange driver design back then and we had tremendous trouble trying to get Kevlar to behave itself. We called it cone cry and I always remember listening to the B&W's back then and their evil kevlar cones screaming at me.

If you are stuck with them you may need a notch circuit. Looking quickly the next generation seems to have a zobel network of some kind in olay that the schematic in green does not.

A bit tired to really check it out but maybe tomorrow.

Mark

Hi Mark

You are dead right - the config is as follows:
HF: 1" Metal dome
MF: 5" Kevlar cone
LF: TWO x 8" Cobex cones

The Matrix 802 (either Series-2 or 3) didn't have the Zobel network but the Matrix 801 Series-2 did & they skipped that in the Series-3.

If you compare the HF circuit, between the Series-2 and Series-3 in green, they are kinda similar but there's a haywire diff in the Mid. Again, the LF is kinda similar.

I am looking to either
(1) run the LF 1-2dB hotter OR
(2) cut Mid slightly by say 1dB or maybe 2dB but then, in the design, the Mid & HF are slapped-on together unless I start tweaking ONLY in the Mid's circuit.

[Inductor]

Originally Posted by arupg

I've a pair of B&W Matrix 802 Series-2 (1988-1993) ...

I would also check for older caps.
In a speaker, not to mention alterations in a zobels or other, it's possible to win a few dB's (1/2) in the woofer by just choosing a low/no resistance inductor (quality/gauge/core) when comparing with the one that comes original. I'm not saying in this case I'm doing this. I don't know what the problem is. Without measuring the speaker, I would do this:
Check the caps (measure) for old ones. They are not very old, to compare, look at the new xover I posted in the link above, with better components.
Check the Mid. If not the tweeter, how do you know? (is not the tweeter). You can connect the attenuator at the ends of the speaker drivers for the tweeter and for the mid (between the xover and driver) but it's not changing the in-balances in frequency in the same driver, if they exist. You can make 3.4R bigger (mid) also... whatever.
Try getting the curves for similar situations, if not measuring yours, like soongsc said.

[mwmkravchenko]

Can you do any kind of measurement?

It will be very telling of what exactly is going on.

There are all kinds of quick and dirty ways of pulling this off.

This is free and works well.

ARTA Home

Many inexpensive microphone capsules are available that will measure the speaker with adequate fidelity. You don't need a measurement mic. You need what is called relative measurement. Even an onboard voice only mic on your computer can be helpfull. THe point being to measure in the same location at the same distance every time you measure and make tweaks.

I really think without seeing the response that padding down the driver will not remove the offending noises. But again I'm guessing from a couple thousand kilometers away!

Mark