Variable frequency active crossover

[lorylaci]

Hi!

I have already built an active crossover with a fixed cross-frequency

I'm looking for an active crossover with a variable crossover frequency too. I would only need a 1:4 range (250 Hz - 1kHz). I was thinking about rotating switches (switch between premade crossover circuits). Using dual pots would not precision enough i think.
If some could send some schematic I can start with that would be great.

[Bigun]

I think it's going to be difficult to control the values of the components and have good tracking between filters unless you use some kind of digital potentiometer as suggested. But it seems possible. There are a number of commercial variable-XO's and perhaps some internet searching will identify how some of them have been designed.


Another choice is to have a subtractive filter. In this approach you tune only one filter. The other filter subtracts the output from the tunable filter from the full-range signal to generate the other output.

I'm not much of an expert on such things, but I did simulate how this would work using two discrete amplifiers:

Bi-amplification to Dx Amplifiers, a passive input filter

[picowallspeaker]

There are also discharge-capacitor filters , sort of SMPS (one type , at least) applied to audio frequencies , IF I remember , the type LMF 100 was an 8 pin device , with a few components around , i did a sub processor which worked fairly well.

[Bigun]

Interesting device. These boys don't leave much for the discrete audio builder..

http://www.swarthmore.edu/NatSci/ech...eet/LMF100.pdf

[picowallspeaker]

Not very interesting for an audio enthusiast , Bigun. Anybody can hear the grainy sound produced by it . I did put sort of compressor after it to override the problem .
It was an LMF 60 , may be ,or 40 . Are you going to project a sub amp that incorporates it in the FB chain of a T2024 ??? OR for a subtractive filter design to drive a supertweeter ??

[AndrewT]

Hi,
does the adjustment have to be continuous or would stepped be acceptable?

Does it require to be externally adjustable, or would internal switches and/or plug-ins be acceptable?

I bought a ready made subtractive crossover filter about 20years ago.
It was terrible. Good for the case and PSU, all else deserves stripping out and binning.

[Bigun]

Originally posted by picowallspeaker
Not very interesting for an audio enthusiast , Bigun. Anybody can hear the grainy sound produced by it . I did put sort of compressor after it to override the problem .
It was an LMF 60 , may be ,or 40 . Are you going to project a sub amp that incorporates it in the FB chain of a T2024 ??? OR for a subtractive filter design to drive a supertweeter ??

I recently built a sub-woofer with two drivers in push-pull configuration for better linearity. It uses an off-the-shelf 500W ClassD plate amp. That was before I started this amp building hobby. I have noticed that the exact XO frequency is not that critical, especially with a sub. I don't know what method the plate amp uses to adjust XO frequency, but a stepped adjustment would have been perfectly fine. When you cross-over a sub at 80Hz you are already at the frequency where humans can't hear very well, can't even tell which direction the sound is coming from.

I am thinking about a bi-amp project when I start my next speaker project, which will be a 2-way (my first 2-way). I don't like what I have read about the passive XO's, that they are hard to design, use exotic passive components and have to be tested and tuned by equipment I don't have or by my ears and that I can spend a lot of money on different component values to optimize with fancy parts. And I hear that off-the-shelf passive XO can not be as good. That all seems like it would be interesting if I wanted to learn about passive XO, but I have no interest right now. Active XO sounds a better approach, but I would design it with fixed XO not with adjustable.


Andrew T - what was bad about the subtractive XO, was it poorly implemented ? I can't see why it should in principle be a bad approach ?

[infinia]

IMO most sensible approach is DIP replaceable RC filter packs for Sallen–Key topology. But as someone pointed out higher order requires a bit of precision.

State variable topology shows the most promise for continuously variable solution. But requires precision ganged pots which is big bucks.

Subtractive is only good for 1st order filter, so limits it possibilies for an all in one solution.

Switched filter is more for instrumentation.

[AndrewT]

The cross over slopes did not match.
The volume coming from a driver that was supposed to be in the stop band was ridiculously high.
The stereo effect was lost.
The drivers trying to operate outside their passband stressed them too much, maybe?
It just sounded terrible.

[Bigun]

Interesting, Elliot also has bad things to say about subtractive filters. Particularly that the slopes are limited by phase issues and you can't get a flat response. But I didn't see this when I used a power amp as the active filter, where the subtraction was done as part of the gnf loop. Perhaps there is something further to explore there ?