3 Way Active Crossover Design project

[powagrip]

Hi,

I'm a student currently working on an active crossover for a 3 way speaker system. Since entering into the world of audio I've noticed that there are many ways of achieving the same goal with some people prefering different methods to others (more so than other industries).
the design I'm going for is a 4th order state variable linkwitz riley design as it seems to be the best solution in my eyes.
I would appreciate learning from any audio professionals why they choose the designs they do as it would make an interesting subtopic for my report. Also hearing what's coming up in the future of active crossovers would be very interesting too and learning any new directions that haven't really been looked into.
thanks for any feedback.

Simon

[lufbramatt]

hi- im also a student at loughborough- are you building the xo for a set of speakers that you have built? we're also into speaker construction, doing an industrial design degree.
Currently im constructing a pair of standmounts, but would at some point in the future like to build a sub module ala wilson watt/puppy, utilising an active xo between the sub and the monitors.
I would be interested to see how u are going about your design, and pass on any tips?
cheers, matt.

[AndrewT]

Hi,
I have rarely(never) heard a xover using state variable filters.
Get some recommendations first before you commit.
I recommend unity gain S-K or ECV S-K or MFB in increasing order of complexity but improving flexibility.
Design for interchangability between Butterworth/Linkwitz-Reilly/Bessel so you can assess best for your system.

[keyser]

Hi!
I'm a student too , and I'm currently working on bi-amping my Canton RC-L's. The Cantons are 3-way speakers too. I've already bought a crossover, the Behringer cx2310, http://www.behringer.com/CX2310/index.cfm?lang=ENG
and an extra poweramp. I too believe a 4th order linkwitz-riley gives best performance for the price.
All I have to do now is rewiring and partly disconnecting the active crossover. Perhaps through this thread we could exchange our experiences of going from a passive x-over to active. My speakers have passive Xover points @ 300 & 3500 hertz. The active crossover will cross @ 300 hertz, leaving the 3500 hertz cross to for the passive.
Keyser

[Alex Kotov]

My project: (RUSSIA) http://akotov.narod.ru/8.html
recommended

[JensRasmussen]

I know it's a 4 way filter,

But you can use it as an idea

http://www.delta-audio.com/Active%20filter%20one.htm

There is a manual for the filter, where you can find formulas for the different sections.

Jens

[Bricolo]

Originally posted by AndrewT
Hi,
I have rarely(never) heard a xover using state variable filters.
Get some recommendations first before you commit.
I recommend unity gain S-K or ECV S-K or MFB in increasing order of complexity but improving flexibility.
Design for interchangability between Butterworth/Linkwitz-Reilly/Bessel so you can assess best for your system.

What does ECV mean?
In what point is a MFB more flexible than a SK?

[5th element]

I would appreciate learning from any audio professionals why they choose the designs they do as it would make an interesting subtopic for my report.

With a passive xover the slopes and frequency decided are usually picked to achieve a good phase response, atleast they are when I make a passive. With an active xover this is less important because you can use delay circuits.

Ofcourse this is all within the sensible operating ranges of the drive units involved.

[RHosch]

State variable is a good option if you want to have high, low, and bandpass outputs with changable center frequency. If you're using it for just high or low pass, the component count is much higher than S-K, but if you need all three 2nd order outputs the component count is actually reduced. It is also more tolerant of component value inaccuracy than S-K designs. Not sure about the multiple feedback designs though... and to be honest I'm not sure why MFB might be desired when S-K seems to work just fine for practically everyone in the world.

You do have a problem with 4th order SV designs in that you either need cascaded sections or a true 4th order section as described by Rane. IMO, cascading seems to be the better option as it is more flexible, but be prepared to look at digital pots or switched resistor arrays if you want to control everything at once with a single knob.


Jens... haven't heard from you guys in a while. I've completed the circuit schematic in functional blocks and am ready to assemble into a completed schematic to prepare for PCB design. If you guys are still interested in helping, shoot me an email - Richard Hosch.

[JonMarsh]

For an example of a 4th order state variable design, download the kit manuals for an XM9 crossover from Marchand.


Most "pro" analog crossovers use this topology, for the ease of tuning the crossover frequency over a wide range; then it requires ganged pots that track well, or electronic potentiometers.

Marchand Electronics