[Joules]

What Is the best  way to split the frequencies for a Bi amped two way that requires at least a 3rd order slope.
iIC OP amp based x-overs don't make it.
Any Thoughts ???

[Gordy]

Hello Joules!  And Welcome!

The Marchands are well thought of and offer a lot of options... http://www.marchandelec.com/xovers.html

[andyr]

What Is the best  way to split the frequencies for a Bi amped two way that requires at least a 3rd order slope.
iIC OP amp based x-overs don't make it.
Any Thoughts ???

Strange, as I actively tri-amp my Maggie IIIAs with op-amp based crossovers from Rod Elliott (www.sound.au.com).

My slopes are 18/12 dB for base LP/mid HP and 12/12dB for mid LP/tweeter HP.  This is my second active crossover iteration ... the first was 24dB all the way through.

Regards,

Andy

[woodsyi]

joules,

I think all that would really depend on the speakers and the individual drivers.  You have to know the specs for each driver and their physical orientation to determine the slope and frequency of the poles.  Curious about 3rd order.  Why 3rd?

[Joules]

Gentlemen - Thanks for your help.
My original post was meant to start a discussion to explore various possible line level x-overs. If you have had good success with IC based circuits, I still would be very interested in hearing about it.

In an MTM speaker, to achieve the intended vertical polar pattern, requires odd order slopes and 1 is too little, unfortunatley, so 3rd order is it. The speakers will be crossed over before they begin to roll off on their own. (Not set in stone yet.)

My feeling is that the least amount of circuitry in line the better, but we still need to cross over some how.

Three line level RC filters in a row would incur too much signal loss. One or two RC filters might be ok (for signal loss) but that's not 3rd order.

I have been thinking that if a suitable 1-2 Henery coil could be made, then one might be able to use a LC in line with an RC filter to get the 3rd order I'm looking for, and with minimal components.
I don't know - just a thought.
What do you think?
Thanks a lot for your input.

[mac]

Are you planning on addressing baffle step & delay (allpass)?  If so, more components.  If not, your results will be far from ideal.  Cheers, mac.

[andyr]

.... Three line level RC filters in a row would incur too much signal loss. One or two RC filters might be ok (for signal loss) but that's not 3rd order.

I have been thinking that if a suitable 1-2 Henery coil could be made, then one might be able to use a LC in line with an RC filter to get the 3rd order I'm looking for, and with minimal components.
I don't know - just a thought....

Joules,

You've identified the problem with line-level passive crossovers - too much loss.  If you're sticking to 1st or second order then you can generally deal with it.

That's why the lord invented active crossovers!  If you're a tube maniac, go for Marchand.  If you can accept that op-amps don't destroy the enjoyment of music, I thoroughly recommend Rod Elliott's.  He only sells you the PCB and you can then choose the components you want to use; Marchand sell you the complete kit.  With his, the easiest configuration is 24dB L-R.

Regards,

Andy

[Joules]

"Are you planning on addressing baffle step & delay (allpass)? If so, more components. If not, your results will be far from ideal. Cheers, mac."

One problem at a time  

"That's why the lord invented active crossovers! If you're a tube maniac, go for Marchand. If you can accept that op-amps don't destroy the enjoyment of music, I thoroughly recommend Rod Elliott's. He only sells you the PCB and you can then choose the components you want to use; Marchand sell you the complete kit. With his, the easiest configuration is 24dB L-R. "

This is on the list of options. - thanks

[Gordy]

Hi Andy,

Just a small clarification   Marchand offers bare boards, partial kits, full kits and fully constucted x-overs.

Another option would be Delta Audio.... http://www.delta-audio.com/Active%20filter%20two.htm

[Joules]

I am wondering ... for a satalite speaker such as this if a single amp may be the way to go, (and using a speaker xover). It would be simpler.

I'm about to buy one or two ASKA  N+ amps and from what I've read, Hugh Dean has gone to great lengths to create a fantastic sounding amp. I want to compliment his amps with the best posible xover, passive, active or speaker.

It would be very interesting to hear Mr. Dean's views and solution to this issue.... I wonder if his GK-1 preamp could be modified in to a active two way xover?

Mr. Dean  ..... you out there?  


Hey woodsyi  - I see you are in the neighborhood. I thought I was the only (home) audio DIY lunatic in the area. glad to meet you!

[AKSA]

Joules,

Thank you for your question;  I appreciate the interest, and the gracious comments.

I am something of a veteran of forums, having participated for about 8 years in public discussions.  However, as a fallible human being with large gaps in my knowledge I am loathe to pass categoric judgement on the heady issues of active v. passive crossovers.  Furthermore, it is not wise to engage dialogues about design philosophy here because there is inevitable disagreement and technical divergence, and incidentally the needs of the DIY audiophile/hobbyist are not always perfectly aligned with the small business environment.

PASSIVES:

1.  Passive crossovers induce phase shift at the crossover point, but partly compensate with impedance correction which can be designed to present a relatively benign load to the amp.
2.  Passive crossovers enable use of just one amplifier per channel.
3.  There are no relative gain issues, unlike their active brethren.
4.  The issue of image depth is much better handled.
5.  Power handling is limited, but this is only an issue in pro-audio.
6.  Cost is MUCH lower.

ACTIVES:

7.  Active crossovers induce phase shift at the crossover point, and unless impedance correction is implemented at the speaker driver the load presented to the amp is quite reactive, chiefly inductive.
8.  Actives require at least two amps, and often three, per channel.
9.  Gain issues are considerable, and normally require specialist calibration.
10. Image depth is a major issue, particularly with 24dB/octave systems.  There is little technical information on this issue, but it is widely realized that 1st and 3rd order are best for passives, so this might also be true for actives.
11.  Power handling is far more optimal;  amps of 20W, 50W and 100W rating can be compatibly coordinated for treble, mid and bass passbands.
12.  IFF using 4th order or beyond, Spl can be far higher because the out-of-band power at each driver is far less.
13.  Cost and complexity is generally three times higher, making these projects almost impractical for most DIYers.

After all this, if you are Clare Brothers, use active.  Go five way.  If you are a HT user, with a 30 seat auditorium attached to your home, consider active.  For all others, bear in mind the difficulties.  It's hard to get right, but a marvellous challenge for the well strapped compulsive-obsessive.

Cheers,

Hugh

[Joules]

Thanks Hugh - Your opinion is greatley appreciated .

[Jens]

Hi Joules,

There's another option that has not been touched upon here: You could actually use a combination of passive and active x-overs.

It can be done like this:

- Use active for the area where the big power is, i.e. bass
- Use passive for the rest

I've done this in my system, where tweeter/mid drivers are crossed passively, using few, high-quality x-over components. Tweeter/mid is powered by an AKSA 55 N+.

Bass is both equalized and crossed using an active x-over (24 dB L-R) and powered by a 500 W class D amp (IcePower).

I have no problems whatsoever with imaging (especially after installing the Nirvana Plus upgrade last week) at all, on the contrary!  

Cheers,

Jens

[Tinker]

A couple of things:
Having built precisely three active systems (and used others in live sound) I can mostly agree with Hugh that active systems are an expensive and tricky business and should not be entered into without a deep breath and adequate measurement equipment.

The main advantage of active is the ability to correct for cabinet/room anomalies and a HUGE increase in power (or huge reduction in size of amp needed - plus lower distortion). A biamp system crossed over at the right point realises four times the power of each individual amp.

Jens is absolutely right. A really good way to build a "3 way" is active crossover for the bass, then passive for midrange. This gets you best of both worlds. Maximum power increase so smaller amps can be used, and the components for a passive crossover between mid and tweeter are much smaller and cheaper. The only place this potentially lets you down is baffle step compensation and schroeder frequency transition equalisation. I made a system like this 5 years ago using the Linkwitz-Riley circuit on Rod Elliot's page for the bass crossover, and Solen caps/Goertz inductors for the passive.

I have had some sucess with op amp XOs, but PCB layout and circuit complexity are big issues. A good DIY crossover could cost you as much at AU$600 to build (with case and PSU), so for the DIYER/EXPERIMENTER, if you want to make a serious study of active systems it  is worth considering buying a cheap active or even a digital crossover like that made by Behringer. Not ultra-fi, but very good. Plus they can be modded to boot.

Cheers,
   T.

[Joules]

"- Use active for the area where the big power is, i.e. bass
- Use passive for the rest
I've done this in my system, where tweeter/mid drivers are crossed passively, using few, high-quality x-over components."

This is what I have been doing (and not the first time, either).
People keep telling me to Tri-amp. So me being who I am, and having never tried tri - amping before, I set out to learn as much as I can about it. I keep coming back to the same old problem, active 3 way incurrs a lot of stuf in the signel path (particularly for the oh so precious mid range). So I am reminded why I Bi - amped in the first plase (as outline above). I keep hoping I can find something better !!

Presently what I have is very simple and very stunning sounding. I have a CD player (with remote controled variable out put), two dual mono amps,(with active LP for bass and passive HP for the highs, befor the amps) and three way speakers (with two way speaker xover between mid and high). Very simple and very good sounding and now it's time to try and make improvements.

[andyr]

"- I keep coming back to the same old problem, active 3 way incurrs a lot of stuf in the signal path (particularly for the oh so precious mid range).  ...

Joules,

Why do you assume that the mid-range LP filter (which you have implemented passively) is any less a "whole lotta stuff in the signal path" than active components would be? It's just different "stuff"!

I went along your path, first implementing an active crossover between base and mid-range in my speakers (ie. this was base LP/mid HP) and keeping a high-quality passive crossover for mid LP/tweeter HP.

But I liked the results of the active at the bottom crossover so much, I decided I'd go active at the top as well.

Here's a suggestion ... in a 3-way speaker, you have the following crossover sections:
* base low-pass,
* mid band-pass, and
* tweeter high-pass.

Instead of the conventional "active for base LP/mid HP & passive for mid LP/tweeter HP" - which is what you have now, if you want to tri-amp why not implement the following setup:
* active for base LP
* active for mid BP
* passive for tweeter HP?

Regards,

Andy

[Martin]

I know this sounds simple, but my Aksonics are powered by a 100n amp.
My powered sub receives it's signal off the Alps pot on my TLPn.
I suppose this would be considered bi-amped also...sort of.
Passive on the highs & mids, active on the lower end.

Martin
 

[Gordy]

Please excuse my newbie ignorance... what would be the drawbacks of simply using a .1uF cap at line level as a high pass?  Certainly far less dirty sounding than the dbx unit I tried with the tour 100N+...  

Thanks
Gordy

[Jens]

Please excuse my newbie ignorance... what would be the drawbacks of simply using a .1uF cap at line level as a high pass?  Certainly far less dirty sounding than the dbx unit I tried with the tour 100N+...  

Thanks
Gordy

Hi Gordy,

Nothing wrong with your idea. However, the cap neds to be calculated in accordance with the input impedance of the amp. Apart from this, no problem. I've done this on a friend's system, which is essentially identical to mine - works a treat - just haven't got around to implementing it on my system yet  

Cheers,

Jens

[andyr]

Please excuse my newbie ignorance... what would be the drawbacks of simply using a .1uF cap at line level as a high pass?  Certainly far less dirty sounding than the dbx unit I tried with the tour 100N+...  

Thanks
Gordy

Gordy,

A cap in series with the input pin of the RCA socket of the AKSA driving the tweeter is a 1st order Line Level Passive Crossover.  As Jens says, the value for 'R' in the equation is that of the AKSA - which I believe should be taken as 43K (Hugh will be able to confirm the value).

When you feed in the frequency of the tweeter -3dB (crossover) point, you'll get the required capacitance.

Regards,

Andy