It's not just about drivers being physically close togther, in fact that is merely second most important thing. The most important this is the driver behavious in the region where the roll off is taking place. Because the roll-off is so feeble you need very capable drivers.
Hi,
low slope crossovers (single pole) will better suit low numbers of drivers. i.e. 2way and just maybe 2way plus sub.
If one plans going multiway, then to obtain the benefits of narrow banding each driver and minimising driver interference (phase, lobing, etc.) then one is almost compelled to go to multipole filters.
A 4way plus sub will be putting about 2octaves into each driver. A bandpass filter using single slope filters will never reach a flat pass band when it is only 2octaves wide.
The out of band signal will extend many octaves beyond the passband and interfere with not just the adjacent driver but the one beyond.
Another advantage of narrow banding the drivers is the ability to select a driver that has exemplary sound characteristics in the narrow passband but may have compromised sound qualities just outside the passband to achieve that exemplary performance. That type of driver cannot be used with a simple low slope crossover.
low slope crossovers (single pole) will better suit low numbers of drivers. i.e. 2way and just maybe 2way plus sub.
If one plans going multiway, then to obtain the benefits of narrow banding each driver and minimising driver interference (phase, lobing, etc.) then one is almost compelled to go to multipole filters.
A 4way plus sub will be putting about 2octaves into each driver. A bandpass filter using single slope filters will never reach a flat pass band when it is only 2octaves wide.
The out of band signal will extend many octaves beyond the passband and interfere with not just the adjacent driver but the one beyond.
Another advantage of narrow banding the drivers is the ability to select a driver that has exemplary sound characteristics in the narrow passband but may have compromised sound qualities just outside the passband to achieve that exemplary performance. That type of driver cannot be used with a simple low slope crossover.
These certainly are complex and interrelated issues.
What I have not ever been able to find is a text which really analyzes all the issues for active crossovers.
The audiophile community has been slow to adopt such active crossovers.
What I have not ever been able to find is a text which really analyzes all the issues for active crossovers.
The audiophile community has been slow to adopt such active crossovers.
http://www.linkwitzlab.com/
These speakers are multiamped and have active crossovers.
I don't know the specifics of how they work, but I would love to.
These speakers are multiamped and have active crossovers.
I don't know the specifics of how they work, but I would love to.
There does not seem to be much literature on active crossovers, especially subtractive.
There are going to be component tollerance issues, and phase issues.
A subtractive filter will not have any of the standard transfere functions used by passive filters.
Then there is the issue of how it really sounds. It is all done assuming spatial summation.
The current passive crossover practices are far from ideal, and they evolved over time.
I know that commerical PA, and a very few audiophile builders are using active crossovers, even active subtractive.
I would like to see some literature that analyzes it all, transfer functions, component value sensitivities, realization schemes, and current practices.
There are going to be component tollerance issues, and phase issues.
A subtractive filter will not have any of the standard transfere functions used by passive filters.
Then there is the issue of how it really sounds. It is all done assuming spatial summation.
The current passive crossover practices are far from ideal, and they evolved over time.
I know that commerical PA, and a very few audiophile builders are using active crossovers, even active subtractive.
I would like to see some literature that analyzes it all, transfer functions, component value sensitivities, realization schemes, and current practices.
zenmasterbrian, as said the Linkwitz site contains some very good information. Also check out the active filter cookbook by Don Lancaster.
It sounds like you are interested in the various ways of implementing filters, so have a look up state-variable filters, biquads, multiple-feedback filters.
It sounds like you are interested in the various ways of implementing filters, so have a look up state-variable filters, biquads, multiple-feedback filters.
Richie00boy,
There are people who post on these threads all sorts of stuff which contributes nothing to the topical discussion. Best just to ignore it as noise. Eventually they will have to learn that if they don't like a thread, or don't like a person, they should just keep quiet. They do not have the right to try to disrupt or shut down threads. Moderators should be deleting posts that are negative meta. That would solve the problem completely.
As far as Don Lancaster, I am intimately familiar with his Active Filter Cookbook.
One thing missing from that book is treatment of an inverting form, to complement the noninverting form, of the Salen Keys circuit.
If you use Salen Keys circuits, you can come up with something which works similar to passive crossovers.
Maybe this is the best way, but maybe not.
Limitations are first, it is not flat summing. Second, the accuracy of the sum it does make, depends on component tollerances. A crossover requires two filters, one HP and one LP. You are effectively asking for a matched pair of filters.
Other ways need to be explored, like using one filter and subtracting.
Now, this will not yield a standard filter transfer function for the subtracted signal. But it will be some kind of a transfer function.
So how, it could be worked out and analyzed. It may even be possible to have symetrical slopes. There also needs to be consideration of phase linearity.
Much earlier in this thread someone posted a very complex schematic with 8 op amps, for an active crossover.
It would seem to be something like a symmetrical filtering and subtraction scheme.
But what I have never been able to find is an analytic treatment where the various approaches are considered and evaluated.
Nor have I ever been able to find a text which reviews the current practices. The audiophile world has been slow to go to multi-amping and active crossovers. Linkwitz is one exception.
But commercial sound has been that way for a while. Also, most new systems now have a powered subwoofer which uses an active crossover.
I have never seen a text which reviews the current circuit practices.
See, if one does go to subtractive filtering, there are still component tollerance issues in the summing. They have to do with the accuracy of the subtraction.
There are still questions about how it is going to sound. The present passive crossover conventions have been developed over decades, and quite hueristically.
Going to active crossovers, ones that use different transfer functions, is a very open ended subject.
I believe there is literature out there. I have been looking and watching for years. Some useful sources have been mentioned on this thread. But I also believe there are people who will come along that know of something more.
As far as active filters, a more analytic book that I like is:
Active filters: lumped, distributed, integrated, digital, and parametric. Lawrence P. Huelsman, editor. With contributions by William J. Kerwin [and others] 1970.
Check out the links below. There will not be any negative meta there. The environment will be more conducive to open ended inquirey.
There are people who post on these threads all sorts of stuff which contributes nothing to the topical discussion. Best just to ignore it as noise. Eventually they will have to learn that if they don't like a thread, or don't like a person, they should just keep quiet. They do not have the right to try to disrupt or shut down threads. Moderators should be deleting posts that are negative meta. That would solve the problem completely.
As far as Don Lancaster, I am intimately familiar with his Active Filter Cookbook.
One thing missing from that book is treatment of an inverting form, to complement the noninverting form, of the Salen Keys circuit.
If you use Salen Keys circuits, you can come up with something which works similar to passive crossovers.
Maybe this is the best way, but maybe not.
Limitations are first, it is not flat summing. Second, the accuracy of the sum it does make, depends on component tollerances. A crossover requires two filters, one HP and one LP. You are effectively asking for a matched pair of filters.
Other ways need to be explored, like using one filter and subtracting.
Now, this will not yield a standard filter transfer function for the subtracted signal. But it will be some kind of a transfer function.
So how, it could be worked out and analyzed. It may even be possible to have symetrical slopes. There also needs to be consideration of phase linearity.
Much earlier in this thread someone posted a very complex schematic with 8 op amps, for an active crossover.
It would seem to be something like a symmetrical filtering and subtraction scheme.
But what I have never been able to find is an analytic treatment where the various approaches are considered and evaluated.
Nor have I ever been able to find a text which reviews the current practices. The audiophile world has been slow to go to multi-amping and active crossovers. Linkwitz is one exception.
But commercial sound has been that way for a while. Also, most new systems now have a powered subwoofer which uses an active crossover.
I have never seen a text which reviews the current circuit practices.
See, if one does go to subtractive filtering, there are still component tollerance issues in the summing. They have to do with the accuracy of the subtraction.
There are still questions about how it is going to sound. The present passive crossover conventions have been developed over decades, and quite hueristically.
Going to active crossovers, ones that use different transfer functions, is a very open ended subject.
I believe there is literature out there. I have been looking and watching for years. Some useful sources have been mentioned on this thread. But I also believe there are people who will come along that know of something more.
As far as active filters, a more analytic book that I like is:
Active filters: lumped, distributed, integrated, digital, and parametric. Lawrence P. Huelsman, editor. With contributions by William J. Kerwin [and others] 1970.
Check out the links below. There will not be any negative meta there. The environment will be more conducive to open ended inquirey.
There is a thread here somewhere that deals quite extensively with subractive XOs. Nelson also has an article on his passdiy site.
dave
That thread is referenced on earlier on this thread. It is long.
I'm sorry but I don't know who Nelson is.
I'm collecting all the good links at the sites listed below. I will also post the link summary here.
I'm sorry but I don't know who Nelson is.
I'm collecting all the good links at the sites listed below. I will also post the link summary here.
http://www.diyaudio.com/forums/showthread.php?postid=961621#post961621
The above reference on this thread would sound quite promising, if it were locatable.
The above reference on this thread would sound quite promising, if it were locatable.
I believe this is the thead in question:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=5435&perpage=10&pagenumber=1
When I started this thread, I actually did not know about the term "subtractive crossover". Other wise I would have found that thread and just continued it.
http://www.national.com/an/AN/AN-346.pdf#page=6
Above is also an informative link.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=5435&perpage=10&pagenumber=1
When I started this thread, I actually did not know about the term "subtractive crossover". Other wise I would have found that thread and just continued it.
http://www.national.com/an/AN/AN-346.pdf#page=6
Above is also an informative link.
http://www.diyaudio.com/forums/showthread.php?postid=963889#post963889
This post, earlier on this thread addresses some of the analytic issues at stake.
This post, earlier on this thread addresses some of the analytic issues at stake.
Summarizing the key links on this thread. This is the complex schematic mentioned:
http://www.diyaudio.com/forums/showthread.php?postid=963935#post963935
Here is where it should be possible to get the AES anthologies, which are claimed to have relevant materials:
http://www.audioxpress.com/bksprods/bkindexltheory.htm
Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
http://groups.google.com/group/audioex_amps_dght
http://groups.google.com/group/audioex_power_supplies
http://groups.google.com/group/audioex_thermal_mechanical
http://www.diyaudio.com/forums/showthread.php?postid=963935#post963935
Here is where it should be possible to get the AES anthologies, which are claimed to have relevant materials:
http://www.audioxpress.com/bksprods/bkindexltheory.htm
Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
http://groups.google.com/group/audioex_amps_dght
http://groups.google.com/group/audioex_power_supplies
http://groups.google.com/group/audioex_thermal_mechanical
The problem is that there is no general recipe available (apart from the fact that there is not much info around) for the use of subtractive crossovers. You'd have to be able to define your overall target response and model the responses of the drivers you want to use. Then you'd have to decide on a case-by-case basis what subtractive topology you want to use and how you want to implement it.
There are some speakers out there that use such topologies but info on those is scarce. Some use a combination of subtractive crossovers and ordinary crossovers with allpass-based group delay compensation.
Some examples:
Active studio monitors by Studer (professional division of Revox).
Active studio monitors by PSI.
Some mid-eighties active speaker by Mitsubishi.
The speakers by Celtic audio seem to be implemented purely subtractive.
Most speakers by Meyersound seem to use classic X-overs that are phase-compensated with delays.
Regards
Charles
There are some speakers out there that use such topologies but info on those is scarce. Some use a combination of subtractive crossovers and ordinary crossovers with allpass-based group delay compensation.
Some examples:
Active studio monitors by Studer (professional division of Revox).
Active studio monitors by PSI.
Some mid-eighties active speaker by Mitsubishi.
The speakers by Celtic audio seem to be implemented purely subtractive.
Most speakers by Meyersound seem to use classic X-overs that are phase-compensated with delays.
Regards
Charles
The scarcity of literature is a problem. Active crossover systems could replicate the transfer functions of passive crossover systems, but they could also use other transfere functions, or even be sutractive. There is need for an ongoing and exhaustive survey, as well as rigorous analysis.
Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
http://groups.google.com/group/audioex_amps_dght
http://groups.google.com/group/audioex_power_supplies
http://groups.google.com/group/audi...rmal_mechanical
Audio Explorations:
http://groups.google.com/group/audioex_amps_atob
http://groups.google.com/group/audioex_amps_dght
http://groups.google.com/group/audioex_power_supplies
http://groups.google.com/group/audi...rmal_mechanical
- Status
- Not open for further replies.