How are active crossovers so simple/clean? - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 17th August 2003, 04:49 AM   #1
tcpip is offline tcpip  India
diyAudio Member
 
Join Date: Jan 2003
Location: Mumbai, India
Unhappy How are active crossovers so simple/clean?

Hi all,

This is a newbie question. If it's a FAQ, please can you just point me to where I can get my answers?

My question is about the apparent simplicity of active crossovers versus traditional output-level (as against line-level) passive crossovers. (I'll call them POL XO for passive o/p-level...)

When we see POL XOs for "real" speakers, we often see a lot of complexity and asymmetry at the XO point. For instance, in a woofer-mid XO, you may find that the high-pass for the mid is an electrical second order while the low-pass for the woofer is a fourth-order. This is because the electrical 2nd-order high-pass for the mid is working in conjunction with its sealed enclosure, resulting in an accoustic 4th order. And so on.

With some other speakers (very good ones too), the XO has been designed such that the high-pass and the low-pass don't even cross over at the same frequency... the gap is wider than what the pure electronics theory would necessitate. However, accoustically, there's no dip in the frequency response graph of the speaker, because the humps and peaks of the drivers' response curves are suitably compensated by the wider-than-natural gap of the electronics.

In other words, the electrical XO circuits are rarely simple and symmetrical, yet, in finely tuned designs, the accoustic results are smooth, uniform, and give you a beautiful flat response curve.

However, when I see (analog) active XO based designs, I find people plonking in a 2nd or 4th order LR XO, same slope on both sides of the XO point, no strange wide gaps between the two curves, etc. And they claim it sounds lovely.

My question is: how come we don't see similarly complex, asymmetrical active XO designs? Is it that all the active XO designs I've read about have less-than-mature XOs designed? Is it that no one has designed finely tuned active XO based speakers of the level of complexity and fine tuning as, say, the Ariels? Is it that active XOs make many problems just disappear, simply because there's no POL XO in the path? I suspect it's a bit of all of the above, and some more...

Please can you help? I was thinking of embarking on the design of a new speaker (open baffle to boot)...

Tarun
  Reply With Quote
Old 17th August 2003, 04:54 AM   #2
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Quote:
Is it that all the active XO designs I've read about have less-than-mature XOs designed?
Not all, but many. Maybe most. I've seen some published designs with proper eq, signal delays, and such in Speaker Builder. And of course there's the Linkwitz designs.

Part of it is that one doesn't have to worry as much about the reactive charactristics of drivers. That factor as much as driver acoustic response is responsible for the asymmetry of many passive designs.
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote
Old 17th August 2003, 06:38 AM   #3
diyAudio Member
 
Join Date: Jul 2002
Location: denmark
.....and, without all those passiv parts in the filter - the speaker(s) is a much easier load for the amp(s).
  Reply With Quote
Old 17th August 2003, 01:29 PM   #4
tcpip is offline tcpip  India
diyAudio Member
 
Join Date: Jan 2003
Location: Mumbai, India
Default Oops! Sorry... wrong forum

Dear all,

I just realised now that I have created this thread in the wrong forum. It should have been in the Loudspeaker forum. Is there any way to move a thread from one forum to another?

Really sorry for creating this bit of a mess... Feeling very stupid.

Tarun
  Reply With Quote
Old 17th August 2003, 01:32 PM   #5
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Thanks for pointing that out. It's now in its proper place.
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote
Old 18th August 2003, 01:41 PM   #6
tcpip is offline tcpip  India
diyAudio Member
 
Join Date: Jan 2003
Location: Mumbai, India
Quote:
Originally posted by SY
Thanks for pointing that out. It's now in its proper place.
Thanks, SY.

Going by your answers, does this mean that if I design, say, a three-way speaker (without any XO, but with separate leads for each driver at first), and then test the frequency response curve for each driver, and verify that there's there's a 2-octave overlap between the flat portions of every pair of drivers (i.e. the woofer to mid, the mid to tweeter), I can simply put in steep 2nd order or 4th order LR filters and get a flat response curve from the speaker? (I'll do SPL matching later on.)

My reasoning is that if I operate well within the measured flat areas of the drivers, I'm taking into account all the driver+enclosure interactions in my measurements itself. And crossing over at points where both sides are flat, I don't have to bother about compensating for various types of slopes of the various complex driver+enclosure interactions. Does this sound ... well, sound?

Tarun
  Reply With Quote
Old 18th August 2003, 02:42 PM   #7
diyAudio Member
 
Join Date: Sep 2001
Tarun, correct xover design is a very complex issue, particularly for a passive network.
There is no such thing as a set of drivers in a box that have a substantially flat response, particularly for 2 octaves either side of the xover point. They always need eq as well as filtering. Even if these existed, the phase shift associated with a roll-off 2 octaves away will still affect how well they sum togehter.
A passive network has the extra complexity of working into a highly reactive load (the driver) and so needs extra components to help deal with this, and must also present a reasonable load to the amplifier, without the impedance dropping too low or being too reactive.
Using simple active filters with no active eq simply shows a fundamental lack of knowledge of speaker design. It has been shown numerous times in the literature that the target function for the speaker drivers to sum correctly acoustically is that the combination of driver AND xover should be , for example, 4th order Linkwitz-Riley (if that is your preferred choice). Even this will not sum correctly unless you take into account driver delay due to positioning on the baffle.
The desired acoustic response, and hence the required electrical response, is independant of the requirements of matching the passive xover to the reactive load of the driver.
Hope this helps clarify your confusion
  Reply With Quote
Old 18th August 2003, 04:29 PM   #8
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Yeah, what Andrew said

In addition, you'll need to think about the interactions of the polar response of the drivers and crossovers. There are various schools of thought on what polar patterns ought to look like and how they should vary with frequency; before starting the crossover design, you need to decide that fundamental issue as well.
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote
Old 18th August 2003, 04:38 PM   #9
diyAudio Member
 
Join Date: Jul 2002
Location: denmark
.......my usual answer to selfmade speakerdesigner wannabe's...:

"i'd rather send a man to the moon than to come up with a working speakerdesign"

thats how difficult it really is......but its a lot of fun...
  Reply With Quote
Old 19th August 2003, 02:19 AM   #10
tcpip is offline tcpip  India
diyAudio Member
 
Join Date: Jan 2003
Location: Mumbai, India
Default Whew!

Quote:
Originally posted by AndrewJ
Tarun, correct xover design is a very complex issue, particularly for a passive network.
There is no such thing as a set of drivers in a box that have a substantially flat response, particularly for 2 octaves either side of the xover point.
Wow. Thanks guys. You blew all my optimism and most of my confusion away.

But let me try to pick up the pieces, just for my own sake. First, about the "drivers in boxes with substantially flat responses" bit.

Isn't it possible for me to choose, say, a woofer driver and put it in a box, such that its response will be substantially flat from, say, 60Hz to 400Hz? I was under the impression that it would be possible with some patience but without Linkwitz-like or Olson-like expertise. Similarly, I'll carefully select a midrange and either mount it in an OB or put it in a sealed box such that the midrange + box will give me a largely flat response from, say, 200Hz to 3KHz, when measured. Just by itself. No XO, no integration with other drivers, etc. Is this possible?

I am assuming, of course, that driver selection will be done carefully, and the box will be designed using modelling tools and other software, and may even be fine-tuned iteratively to achieve a driver+box combo which is flat in a range.

This was my first premise. Please guys, can you react to this? I don't see any point in my going ahead with this line of thinking if my most basic premise itself is overly simplistic.

If it works this far, then in my next post I'll ask about XO. And AndrewJ, I'm never thinking POL XO. I'm only thinking of active XO, because that's the only thing I can remotely understand.

Thanks enormously,
Tarun
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Clean crossovers Clipped Car Audio 78 24th March 2008 03:55 PM
Active crossovers GRollins Solid State 42 26th August 2006 09:26 AM
active crossovers, please help herbd2 Multi-Way 2 23rd February 2005 03:21 PM
Active Crossovers. hacknet Solid State 3 17th February 2004 11:15 PM
Active crossovers - is it this simple? annex666 Multi-Way 6 24th September 2002 07:43 PM


New To Site? Need Help?

All times are GMT. The time now is 10:45 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2