john banks

Nigel wrote in

chiark

You could "bi-amp" a system: this would involve using a pair of passives. At its simplest this would be one inductor and one capacitor. One channel would run the tweeter with a cap in line, and the other would run the woofer with an inductor in line.

Phase problems would plague you doing this, so you'd need a bit of a better crossover design, but in theory it's possible. You would not be able to use your existing crossovers unless they were designed for bi-amping. Some x-overs from MB Quart were designed with this feature in mind...

Unless I'm greatly missing the point and you're talking of using a passive x-over upstream of an amp??? In which case, I want some of whatever you're on as the impedances are so vastly different it would not even almost work.

john banks

I don't know what I'm on these days, so you can't have any

"using a passive x-over upstream of an amp" is what I thought Nigel might be on about. Your reply seems to suggest what I thought - ie the much higher impedance of line level compared with speaker level would completely change the crossover points. But might it work with speaker level inputs to the amp???

I was thinking L/C PAIR for EACH arm of the crossover hence 12dB/oct with no phase problems - ie the same design as most passives in outboard boxes.

Could a 12dB/octave active crossover be designed with L/C pairs just like the passive boxes with suitable adjustments in inductance and capacitance of the components to work at line level? Can this be done without using op-amps? Any idea what sort of losses this would draw from the signal?

I suppose it would be so easy to biamp if 4 channel amps just had a wider range for their HPF/LPF - mine seem to be 40 to 240ish Hz. Obviously need around 3.5KHz ballpark for most tweeters. Surely wouldn't be too difficult for manufacturers to do - and put a secondary fixed 80Hz HPF/LPF whilst they're at it so we can bandpass the midwoofer, highpass the tweet and lowpass the sub. And make it 6 channels to do the whole lot in one box. Now I am dreaming and really have had too much of that stuff!

But knocking up a few L/C networks may be an interesting experiment meantime. I did notice Maplin have very little selection of bits for audio quality inductors - no make your own kits even. Can't seem to easily find suppliers. Any ideas? (prob should just go buy an active x-over box and try that!)

[This message has been edited by john banks (edited 18 June 2001).]

rockin'Ru

Yes John,please share with the rest of the class.
John,there are many companies that make 2,3 and 4 way active xovers.They are truly marvelous tuning tools.A friend of mine bought a Radio Shack 4 way that is awesome in ability(but sucks as far as quality ).He runs the same 3way speaker setup as I do,but I run a passive xover board.He has two 4 channel amps.
The sub section
50,80,100,130hz xover points with a variable bass boost from 30 to 100hz.
Mid bass section
50,80,100,130 high pass points and 200,300,500 and flat low pass points.
Midrange section
1,2,4000hz and flat high pass and 2,3,5 and 700hz low pass
Tweeter section
1,2,4 and 6000hz
All 4 channels have their own level controls.He has tone and volume control over each of the six speakers up front and two 18" subs,all with 2 four channel amps.
Only problem with it is it's cheap.Has a signal to noise ratio of 70.Not too dynamic sounding,but he only paid $70 for it just to get started.
Ready for the next level John???

chiark

Living in Leeds, I have the joys of Farnell just down the road (http://www.farnell.com) and also RS Electronics, who have trade counters throughout the country -

Adam M

Nick, cost is the reason in the home hifi world.

In the realms of real hifi, my active crossover alone cost me £650, on top of that there is the power supply to drive it @£500 then the extra cable at £140. My speakers were designed to specifically prevent bi amping, my guess is cause the manufacturer was forcing its clients to buy the complete active set up, and like a mug, I did.

I dont think what you are suggesting possible. Active crossovers have been around for some time because they are necessary.

As was mentioned, the only way to amplify the line level signal without distorting it in the process, is to use an op amp. To explain to others who may not understand, these are electronic switches which when set up in feedback loops with their own power supply can produce a signal which is identical to that going in (eg. from the headunit). They have enormous input resistance (mega ohms) so that effectively no current from the music signal can get through.

I cant see how you could take such a tiny signal such as a line level and pass it through even the correct L and C circuits and expect it to generate an oscillating field without degrading the signal beyond comprehension.

Look at it this way, if it could be done effectively (and we arent talking complex electronics here) it would have been done already.

chiark

And if it hadn't been done already, do you know a good patent lawyer?

Cheers Adam, thanks for clarifying those points.

Nigel Bowles

Thanks guys for expanding on the comment.
It WAS 9am on Sunday and I was only half awake.

Reading through the thread there are several interesting comments and thoughts. Several are backed up by mathematical theory ...
BUT it doesn't always work like that (don't ask why, it just IS)

I was running a Rockford Fosgate setup, it started as a 250w 2 channel running bridged into a passive xover to RF mid/tweeter for each side of the car. The tweeter was very harsh as standard or with the 3db cut sounded muffled. Moving the tweeter was not an option.
I then went for a Rockford EPX2 (glorified EXPENSIVE 28 band Xover). I split the left channel to one 250x2, mids through one channel and tweeters through the other (same for the right).
The result was the ability to adjust the xover point and the gain to each channel whilst leaving the amp gain low.
The subs were run by a pair of amps bridged together. These filled in the bottom end nicely.
The clarity of the new setup allowed the system to be driven harder and overcame the reduced power.
As I said, the theory said the performance should have been halved, but it had a gain.

Other option, (not tried myself but seen working).
* Use different resistance speakers for the mid and tweeter
* Add a rear pair of mids to change resistance and increase volume

For active cross overs, as well as RS, try Maplins

john banks

Thanks all. Some ideas to work with here. Didn't realise that biamping uses passive components on amp outputs - thought it was the same as putting an active crossover before the amp input. Biamping doesn't sound like the way to go then.

Nigel Bowles

Edited message to read in ENGLISH !!!

Adam M

Nigel,

what you did makes complete sense, but power claims from manufactures should always be taken in context. They are claimed at specific values of distortion and over a range. The efficiency of a peaker is certainly not linear over a frequency range, so it is possible to reduce the claimed out put of a speaker to solve a problem, then accentuate the frequencies you like the sound of. You may then find you have hit an amp and speaker coincidental sweet spot, which makes the output much louder. On top of this, you may have pushed the amplitude of an unwanted frequency that makes a specific panel distort right down, so that you can no longer hear how that interfers with the sound field.

when installation and environment are brought into the equations, then car audio is certainly not an exact science.

But if we are talking pure circuit design on paper, then the numbers do hold.