3way XO help greatly appreciated!

[Andy G]

Andy , imho, does things with almost 'clinical precision' with well behaved paper or poly cones and series crossovers. The results I heard were outstanding

Thank you, kind sir !

Ok, what I'm now considering is this...I already have 4 nice 8" poly drivers, which should go ok to maybe 600Hz 2nd order
acoustic or maybe even 220Hz 1st order series electrical? I have 2 nice, but old, D25TG-55-06 tweeters which will definately be ok crossed at 8kHz first order.

IMO, you don't want to take the Jaycar 8" Poly's much higher than about 300Hz. They do bass very nicely, but can get a bit muddy in the low mids. And crossing to the D25 at 8k is a waste, aim more at 3-4k with a good 4" or 5" mid. Peerless and Vifa make (made) some very good mid drivers suitable for the purpose .

As you know, I rather like the M11WH, which is currently on sale at Madisound. So is the P13WG, which is also a very nice driver, 4 ohm version so good for and MTM (wired in series).
A nice WW+MTM. not that difficult with the right drivers y'know

Waiting for tax refund, probably do some spending myself

[grantnsw]

Omni,
I just re-read my last post to you on impedance issues etc..lol, hey, it seems kinda confused! But I know what I meant! How is your tweaking going? Cheers mate! ( many things happening here , so sorry for the delay)

Andy,
Many thanks for your kind advice on driver selection. I've been wondering about this for a VERY long time now!
The yellow-glass Vifas seem really good. Indeed, your Juke mids sounded great. Thanks for the info on the M11WH - it looks good too. For me the MG10SD's might even be better? ..see the attached, x-axis not aligned
So, after all this time, lol, what I'm thinking is this ..TMWW , ie, D25TG, Vifa MG10 and 2X Jaycar8" '2136'
in a tower. Series 1st order, Thanks! grant

[grantnsw]

Serendipity! The Vifa 'MG10' at about 1700-1900Hz has a small peak to counteract the 'box losses' due to diffraction, etc in my enclosures.

[Volvo_Victim]

Does anyone know where i can find a manufacturer or schematic which fully implements Steen Duelunds 3-way phase-coherent design for filter co-effiecient a = 2sqrt2 ?:

http://www.meta-gizmo.com/Tri/speak/STEEN.html

I'm attempting a 3-way active system with a filter co-effiecient of a = 2sqrt2 = 1.41 as described and graphed in Steen's Theory, yet haven't found any schematic or guide for the implemtation of this passive filter theory to construct this as an active filter from start to finish.


If anyone knows where i can commission or buy such a filter please contact me!

Cheers
Ben

[sreten]

Originally posted by Volvo_Victim
Does anyone know where i can find a manufacturer or schematic which fully implements Steen Duelunds 3-way phase-coherent design for filter co-effiecient a = 2sqrt2 ?:

http://www.meta-gizmo.com/Tri/speak/STEEN.html

I'm attempting a 3-way active system with a filter co-effiecient of a = 2sqrt2 = 1.41 as described and graphed in Steen's Theory, yet haven't found any schematic or guide for the implemtation of this passive filter theory to construct this as an active filter from start to finish.


If anyone knows where i can commission or buy such a filter please contact me!

Cheers
Ben

Hi,

You may need to commission it. You cannot buy it.

For said 3-way the dueland principle gives you the acoustic targets.

For an active speaker the filter function is the (target - raw) response
of each driver in its acoustic environment. The raw response cannot be
assumed to be flat / perfect - a common active c/o error.

The lower order bandpass is the most problematic.

You do not understand the theory, a = 1.414 gives a two way.

http://www.linkwitzlab.com/crossovers.htm

You could go for 10) in the above or a = 2*1.414 in the article.

(edit : 2sqrt2 = 2.828 so some confusion above)

/sreten.

also see : http://www.linkwitzlab.com/filters.htm

[Volvo_Victim]

Thanks Sreten- quite correct- i forgot to multiply the sqrt2 * 2 = 2.828.

I certainly understand what you are saying about actual acoustic response of drivers in a system.


Why is the lower order bandpass the most problematic?

[Volvo_Victim]

I noticed #10 from Linkwitzlabs shows the function for a = 4, where crossover points are -6db.

I am looking for a way to express in an active filter the function of a=2.828427 (2*square root of 2)

To quote Steen directly:

"With that value of "a" you get a rather straight lined phase turn What that means exactly has been subject to discussions, where we couldn't agree. By experiments it seems as this steady curvature of phase also means a steady sound picture.

Common point for bass and treble at -18 dB and mid is dampened -2.5 dB."

[sreten]

Hi,

For 10) SL has squared the function again (H3^4) and used a=3.
(note mid is 0dB db down, bass / treble -36dB)

for 7) H3^2 and a=4. (note mid is 1.2db down, bass / treble -24dB)

So you are after H3^2 and a=2.828:

http://www.meta-gizmo.com/Tri/speak/Image9.gif

As suggested by SL this could be approximated with 2nd order
functions for the low and high pass, thse are easy to fiddle with
in a circuit simulator - you only have Q and F to worry about.

/sreten.