Thanks Paul !
As posted by paulb:
"Originally posted by Fred Dieckmann
<no useful content>"
Precisely...... Thanks for summerizing my views so eloquently. It is the lack of useful content that I would think most would find so alarming. Grey says this all so straight forward and easy that can't imagine anyone needing any more help since the basic topology and references have been set down already. How much hand holding do you need. what were you expecting? A full schematic with component values?
Or maybe as Mr. Pass states on page 22 of the excellent manual for the XVR1 as good advice to all.
"Much of what you think you know about active crossovers is probably wrong. "
As posted by paulb:
"Originally posted by Fred Dieckmann
<no useful content>"
Precisely...... Thanks for summerizing my views so eloquently. It is the lack of useful content that I would think most would find so alarming. Grey says this all so straight forward and easy that can't imagine anyone needing any more help since the basic topology and references have been set down already. How much hand holding do you need. what were you expecting? A full schematic with component values?
Or maybe as Mr. Pass states on page 22 of the excellent manual for the XVR1 as good advice to all.
"Much of what you think you know about active crossovers is probably wrong. "
Re: Thanks Paul !
But why would expect more from Grey? He isn't obliged to provide more that what he feels like.
I'm clearly disappointed with the reaction of some of the members. Did you forget how to do your homework (or maybe you have a dog)?
Fred Dieckmann said:
But why would expect more from Grey? He isn't obliged to provide more that what he feels like.
I'm clearly disappointed with the reaction of some of the members. Did you forget how to do your homework (or maybe you have a dog)?
Re: Thanks Paul !
You've chosen to misinterpret my post. Do this simple test: How much useful content does this thread lose if your posts were removed? I would say zero.
Grey has a talent for explaining things in simple terms, and while you may choose to pick it apart from a subject matter expert perspective, it is still very accessible to us mere mortals. If he has made an error, a polite clarification would be appreciated, but the inflammatory tone of your posts makes it sound like you are motivated by jealousy or vindictiveness. I therefore tend to ignore your posts but read Grey's with interest.
I've wasted far too much time on this already. That's it from me.
You've chosen to misinterpret my post. Do this simple test: How much useful content does this thread lose if your posts were removed? I would say zero.
Grey has a talent for explaining things in simple terms, and while you may choose to pick it apart from a subject matter expert perspective, it is still very accessible to us mere mortals. If he has made an error, a polite clarification would be appreciated, but the inflammatory tone of your posts makes it sound like you are motivated by jealousy or vindictiveness. I therefore tend to ignore your posts but read Grey's with interest.
I've wasted far too much time on this already. That's it from me.
Yes we did expect more and still do
"But why would expect more from Grey? He isn't obliged to provide more that what he feels like."
No you are right he is under no obligation. I understand he has other more pressing obligations and his time is limited but...........
He posted that he had a design and even a PCB layout. What I have seen seems to be a rehash of references that others have posted, speculation, and now an unstable discrete op amp design that appears to never have be built and tested This latest circuit from him, the uses 32 volts suply rails allowing the circuit, used as a follower
to swing about 60 volts using transistors that are only rated for 50 volts. Simulation so far shows it be unstable (oscillates) and it has no provision adjusting the offset voltage by allowing adjustment of one of the current sources.
He is also not obligated to ride on Mr. Pass's coat tails by pressuming to know what Nelson is doing in his designs and associating his project with Mr. Pass's projects by tacking "xen" in the name, something even Nelson kidded him about by suggesting the Rollover as a name. This would all be pretty trival and funny except quite a few people expect him to deliver a working and useful circuit. As Joe Friday used to say "Just the facts Ma'am." Maybe just an admission that it is not done and needs some more work would clear the air and take the pressure off. Maybe a new name for it while it is a work in progress would be a good idea. Maybe the "do-over" or "lay-over" would be appropo. Most people actually wait until the baby is actually delivered to christen it. As for doing the homework, I actually did mine on the circuit for a discrete op amp that I posted and am working on the resistor and capacitor matrix, without which the whole concept of an easily tunable crossover is pretty well out the window. The "Double Cross" schematic posted by Jam seems a fairly decent and simple game plan and could be easily fleshed out for the inclusion of resistor and capacitor networks with Q adjustment. I don't own a dog but one would be real useful in sniffing out where this project is hiding.
"But why would expect more from Grey? He isn't obliged to provide more that what he feels like."
No you are right he is under no obligation. I understand he has other more pressing obligations and his time is limited but...........
He posted that he had a design and even a PCB layout. What I have seen seems to be a rehash of references that others have posted, speculation, and now an unstable discrete op amp design that appears to never have be built and tested This latest circuit from him, the uses 32 volts suply rails allowing the circuit, used as a follower
to swing about 60 volts using transistors that are only rated for 50 volts. Simulation so far shows it be unstable (oscillates) and it has no provision adjusting the offset voltage by allowing adjustment of one of the current sources.
He is also not obligated to ride on Mr. Pass's coat tails by pressuming to know what Nelson is doing in his designs and associating his project with Mr. Pass's projects by tacking "xen" in the name, something even Nelson kidded him about by suggesting the Rollover as a name. This would all be pretty trival and funny except quite a few people expect him to deliver a working and useful circuit. As Joe Friday used to say "Just the facts Ma'am." Maybe just an admission that it is not done and needs some more work would clear the air and take the pressure off. Maybe a new name for it while it is a work in progress would be a good idea. Maybe the "do-over" or "lay-over" would be appropo. Most people actually wait until the baby is actually delivered to christen it. As for doing the homework, I actually did mine on the circuit for a discrete op amp that I posted and am working on the resistor and capacitor matrix, without which the whole concept of an easily tunable crossover is pretty well out the window. The "Double Cross" schematic posted by Jam seems a fairly decent and simple game plan and could be easily fleshed out for the inclusion of resistor and capacitor networks with Q adjustment. I don't own a dog but one would be real useful in sniffing out where this project is hiding.
It looks to me like Grey tried to steal some limelight by announcing that he had a crossover with board layouts but when he was called to the mat, he had to play catch up because he did not have the goods.............well it seems that way.
He made a statement in another thread
Which I asked him to explain but still have no response.
Maybe he has made an amazing scientific discovery or maybe this is pure.................
He made a statement in another thread
Which I asked him to explain but still have no response.
Maybe he has made an amazing scientific discovery or maybe this is pure.................
Attachments
Tis the gift to be simple this the gift to be free
"Grey has a talent for explaining things in simple terms, and while you may choose to pick it apart from a subject matter expert perspective, it is still very accessible to us mere mortals."
Until you try try to build something based on generalizations, wild guesses, and untested circuits. I guess style is more important than content. Why simulate, build, measure, and listen to something when it's so much easier to just say " This is probablly what Nelson did in his design."
You know there really must not be anything to audio design if you can look at a couple of simplified schematics and pull a few parts values and transistor types out some other designs and throw it together and say it works. I guess if you never really plan on building the stuff and having it work, not blow up, and sound decent it's all no sweat and simple, that is if you just want to play audio designer on the forum and not actually deliver the goods and state the specifics of a circuit. If I pulled one tenth of this nonsense I would be tarred and feathered. Yes I'm jealous, of the ability to dupe so many people with so little effort and have them come back for more. I will go backing actually doing some design work now instead of pontificating "this is what Nelson probablly did" based mostly on idle speculation.
"Grey has a talent for explaining things in simple terms, and while you may choose to pick it apart from a subject matter expert perspective, it is still very accessible to us mere mortals."
Until you try try to build something based on generalizations, wild guesses, and untested circuits. I guess style is more important than content. Why simulate, build, measure, and listen to something when it's so much easier to just say " This is probablly what Nelson did in his design."
You know there really must not be anything to audio design if you can look at a couple of simplified schematics and pull a few parts values and transistor types out some other designs and throw it together and say it works. I guess if you never really plan on building the stuff and having it work, not blow up, and sound decent it's all no sweat and simple, that is if you just want to play audio designer on the forum and not actually deliver the goods and state the specifics of a circuit. If I pulled one tenth of this nonsense I would be tarred and feathered. Yes I'm jealous, of the ability to dupe so many people with so little effort and have them come back for more. I will go backing actually doing some design work now instead of pontificating "this is what Nelson probablly did" based mostly on idle speculation.
and in this corner we have Fred and Jam! the Intercontenental Tag Team Champions!!
love those tights guys!!
love those tights guys!!
<<message deleted due to the fact that my knowledge would contribute only slightly more than other people's criticism>>
Happy Thanksgiving.
Sandy.
Happy Thanksgiving.
Sandy.
Happy Thanksgiving.
Hi Sandy,
Well there's good and bad on boths sides of this good versus evil duel.
However, Fred does make some valid points in his post 27 but the come back verbatim is wrong Man, all Wrong when all thats required is a post saying please clarify or can you state specifics PLEASE.
It pays not to make assumptions or unrealistic expectations about another's technical ability particularly in this sort of informal environment, its all too easy to be intolerant if you judge others by your own technical statue.
These behaviours are very typical of certain technical self obsessed personalities who are socially un balanced and struggle with anyone who does not quite fit in their tiny tiny little word.
And then there are repeat offenders. Putting them back in line is like slapping a young adolescent around the head with an attitude adjuster to make them listen.
macka
Hi Sandy,
Well there's good and bad on boths sides of this good versus evil duel.
However, Fred does make some valid points in his post 27 but the come back verbatim is wrong Man, all Wrong when all thats required is a post saying please clarify or can you state specifics PLEASE.
It pays not to make assumptions or unrealistic expectations about another's technical ability particularly in this sort of informal environment, its all too easy to be intolerant if you judge others by your own technical statue.
These behaviours are very typical of certain technical self obsessed personalities who are socially un balanced and struggle with anyone who does not quite fit in their tiny tiny little word.
And then there are repeat offenders. Putting them back in line is like slapping a young adolescent around the head with an attitude adjuster to make them listen.
macka
You guys are getting way too excited, and it's time to ask
the nurse for your meds.
The central problem in a crossover network is not what to
include, but what to leave out. For example, I have only
encountered one case where a 24 db/oct crossover sounded
good, and then only after a considerable quantity of wine....
The issue is resolved by using electronic crossovers day to day,
and noting what kind of solutions tend to work (objectively and
otherwise) and what don't. By this means we distill the options
down to the useful stuff.
I once built a crossover that could anything (much like the XVR1)
except the adjustments were all on the front panel. It drove me
crazy. I found that putting them on the inside kept the insanity
down to a dull roar due to the extra work making the change.
Those of you expecting a miracle circuit from me in this
category are going to be disappointed, as the ony miracle
will be the best compromise between an XVR1 and all of the
inappropriate "formula" crossovers out there on the market.
I've said it before and I say it again - formula crossovers only
rarely approach optimal performance, and usually they suck, and
by formula I mean "XX dB/oct at XXX Hz" with a classic Q and
both drivers at the same frequency and slope.
My goal here is not to dazzle, but to convince people that
electronic crossovers are better, easy enough, and far more
appropriate to the DIY crowd than the alternatives. They
ultimately will give you more bang-for-the-buck as long as
you are willing to take the time to play, and assuming that
you know what you like when you hear it.
the nurse for your meds.
The central problem in a crossover network is not what to
include, but what to leave out. For example, I have only
encountered one case where a 24 db/oct crossover sounded
good, and then only after a considerable quantity of wine....
The issue is resolved by using electronic crossovers day to day,
and noting what kind of solutions tend to work (objectively and
otherwise) and what don't. By this means we distill the options
down to the useful stuff.
I once built a crossover that could anything (much like the XVR1)
except the adjustments were all on the front panel. It drove me
crazy. I found that putting them on the inside kept the insanity
down to a dull roar due to the extra work making the change.
Those of you expecting a miracle circuit from me in this
category are going to be disappointed, as the ony miracle
will be the best compromise between an XVR1 and all of the
inappropriate "formula" crossovers out there on the market.
I've said it before and I say it again - formula crossovers only
rarely approach optimal performance, and usually they suck, and
by formula I mean "XX dB/oct at XXX Hz" with a classic Q and
both drivers at the same frequency and slope.
My goal here is not to dazzle, but to convince people that
electronic crossovers are better, easy enough, and far more
appropriate to the DIY crowd than the alternatives. They
ultimately will give you more bang-for-the-buck as long as
you are willing to take the time to play, and assuming that
you know what you like when you hear it.
Very nice to hear from you Mr Pass,
The weather is grows very warm here today so this tends stir the devil and it is the custom to sacrific he who has sin'd the most..muhahahaha...
Master, you bring clarity and sanity to the natives. A burning at the stake will wait another day.
Can you offer some enlightenment to your final choice of active buffer so as to bring further relief, alas then the nurse can go home early.
For my own part, I already have the fitted up values for the SK high low filters (biamped speakers) , just some dithering over the buffers.
I assume the more I can leave out the better.
Thankyou again for your input.
Ian
The weather is grows very warm here today so this tends stir the devil and it is the custom to sacrific he who has sin'd the most..muhahahaha...
Master, you bring clarity and sanity to the natives. A burning at the stake will wait another day.
Can you offer some enlightenment to your final choice of active buffer so as to bring further relief, alas then the nurse can go home early.
For my own part, I already have the fitted up values for the SK high low filters (biamped speakers) , just some dithering over the buffers.
I assume the more I can leave out the better.
Thankyou again for your input.
Ian
Attachments
Warning - lenghty post from a non Xenover-member
That reminds me of those days when I carried my TV (a 25kg monster still black/white screen) downstairs in the basement in order to prevent me from arbitrary viewing BS (and 95% of what is shown in TV is just that) and to bring it back only when I want to see something "valuable". It did not always work as intended though
Me (just for example😉 ), haven`t expected the XVR1 to be something else than what it turns out to be foremost - a universal (to the largest extend) adjustable (order of filters, Q and f) active crossover.
For DIYers basicly this is nothing new as a properly tuned crossover, be it active or passive must be tailored to the specific properties of the drivers (and some other things) anyway and more often than not this turns out to be anything else than something calculated according textbook formulas.
I think for most DIYers a XVR1 (or a similar flexibel circuit) is completely overkill and not really needed (though nice to have).
As I see it it`s most useful for folks building active speakers every day or for customers of commercial High-End speakers who wants to "activate" them and which are NON-DIYers - and fully adjustable active crossovers are very rare on the market as far as I know (I don`t mean DIY-market but commercial Hifi-equipment).
Actually the intro of the description of the XVR1 at the passlabs site says pretty much the same IMO:
"The XVR1 electronic crossover network is the result of several years research into the actual needs of loudspeaker designers"
For a DIYer in order to build an active crossover for HIS pair of speakers a XVR1 (or the like circuit) with its great flexibiltiy I consider as rather an excellent tool for experimenting than an absolute necessity to have soundwise (as it appears to me). This only as I don´t really understand the excitement the XVR1 causes among DIYers, although I understand that for NON-DIYers the XVR1 probably is THE MISSING LINK in order to achieve their dream of a fully active (and to a great extend individually configurable) High-End speaker system.
For comparison - I don`t believe somebody would go into troubles and big (cost) efforts to built an universal adjustable/switchable xx-way passive crossover unit unless people that are experimenting crossovers for different speakers every day.
The usual procedure people using during developing a passive crossover (at least I think so) is an experimental soldered setup (rat`s nest) where part values and filter topologies can be swapped easily. When the final topology and parts values are determined at the end (through extensive listening AND measurements), the crossover will be build up properly.
Now that`s exactly what I do as well when building an active crossover and IMO there is no substitution for this process.
Even having a XVR1 which could "simplify" and shorten this, actually I could imagine that such a tool in hand of folks
not yet really very familiar with how filters (together with speakers!) are working, could prevent them to find the right of the xx million possible settings (and I`m quite convinced that for a given speaker system there are not too many that will fit very well) as they are likely tempted just to zap back and forth the filter-settings, without ever being able finding the right ones.
My point I try to make is just this: the likelihood to build a good sounding speaker/ (active)crossover combo is MUCH higher through
first.) getting a more complete understanding of:
a.) how does an (active) filter work by itself basically(*)
b.) what acoustically effects has (or can have) a filter in INTERACTION with an actual speaker driver
c.) what acoustically effects can have two (or a x -number of) drivers with their respective filters and when they work together (+ under consideration when physically placed here or there and in respect to each other)
d.) what effects can have the (dimensions for instance) room and the speaker enclosure and speaker placement on the performance
+second.) WHENEVER possible doing some ACTUAL measurements (drivers individually without crossover, with lowpass, with highpass, all together) - instead of building a universal adjustable circuit that possibly can "everything" and just "zapping-through" settings until it appears to sound somehow decent.
(*)and this part is really not that complicated as it might seem at the first glance - for beginners I can only (repeatedly, as others have done this before) recommend Don Lancaster: "The Active Filter Cookbook" - You won`t find a book with simpler descriptions (and even me understood it after reading several times, and at that times I still had NO CLUE about this things)
Once somebody has understood the "Cookbook" filter circuits, together with a better understanding of effects of points b.) to d.) - (which are finally the reasons why real-world crossovers differ from textbook designs), he already gets an idea where, why and which kind of circuit is feasible and in principle can build an active crossover, which even by comparing with a XVR1, does not lack any performance - with regard to possibility of topologies and settings only of course, not refinements of the discrete active circuits, whereas IMO the right choice of topology (order of filter) and right parts values (Q and f) and not to forget the relative volume levels of the individual crossover branches (in more than 2-way systems not so trivial and easy as it might look) matter more than the final execution of the active circuits itself (hope this statement does not cause a flame war).
Decent OP-Amps (huuuhh the devil word) and capacitors for line-level circuits are so cheap (much cheaper than an equivalent passive filter circuit after the power amp) and maybe only need slightly more time to breadboard than a passive crossover, so why not a simple experimental OP-Amp setup first, instead of designing a possibly "overcomplicated" and physically comparatively large (and supposed to be) "all-in-one" unit from the beginning?
You even don`t have to solder much - IC-sockets for fast changes of resistor and capacitor filter values (in order to change Q and f) can make life easy (of course You must power off the circuit before You change parts!).
When done with experimenting one can implement the active stages fixed in whatever flavour floats ones boat - breadboarded or PCB - discrete (again in whatever flavour) or just plain simple with OP-Amps (huhh..shame... I said it again).
For specific DIY active crossover needs I believe this approach is cheaper and easier to build and above all from what I said before likely leads to a more deliberate result (and therefore probably better) as this way people are (or must be) more aware of what they`re doing.
Also this ensures that only this kinds of circuity find implementation which are really needed - in contrary to a universal applicable circuit where half of it may be is left unused because not needed (or worse when it turns out that something desirable is not even included).
BTW: I see kinds of applications (again, for the DIYers) not covered even from a XVR1 - what about subtractive filters or LR-transform circuits (or maybe delays)?
That reminds me of those days when I carried my TV (a 25kg monster still black/white screen) downstairs in the basement in order to prevent me from arbitrary viewing BS (and 95% of what is shown in TV is just that) and to bring it back only when I want to see something "valuable". It did not always work as intended though
Me (just for example😉 ), haven`t expected the XVR1 to be something else than what it turns out to be foremost - a universal (to the largest extend) adjustable (order of filters, Q and f) active crossover.
For DIYers basicly this is nothing new as a properly tuned crossover, be it active or passive must be tailored to the specific properties of the drivers (and some other things) anyway and more often than not this turns out to be anything else than something calculated according textbook formulas.
I think for most DIYers a XVR1 (or a similar flexibel circuit) is completely overkill and not really needed (though nice to have).
As I see it it`s most useful for folks building active speakers every day or for customers of commercial High-End speakers who wants to "activate" them and which are NON-DIYers - and fully adjustable active crossovers are very rare on the market as far as I know (I don`t mean DIY-market but commercial Hifi-equipment).
Actually the intro of the description of the XVR1 at the passlabs site says pretty much the same IMO:
"The XVR1 electronic crossover network is the result of several years research into the actual needs of loudspeaker designers"
For a DIYer in order to build an active crossover for HIS pair of speakers a XVR1 (or the like circuit) with its great flexibiltiy I consider as rather an excellent tool for experimenting than an absolute necessity to have soundwise (as it appears to me). This only as I don´t really understand the excitement the XVR1 causes among DIYers, although I understand that for NON-DIYers the XVR1 probably is THE MISSING LINK in order to achieve their dream of a fully active (and to a great extend individually configurable) High-End speaker system.
For comparison - I don`t believe somebody would go into troubles and big (cost) efforts to built an universal adjustable/switchable xx-way passive crossover unit unless people that are experimenting crossovers for different speakers every day.
The usual procedure people using during developing a passive crossover (at least I think so) is an experimental soldered setup (rat`s nest) where part values and filter topologies can be swapped easily. When the final topology and parts values are determined at the end (through extensive listening AND measurements), the crossover will be build up properly.
Now that`s exactly what I do as well when building an active crossover and IMO there is no substitution for this process.
Even having a XVR1 which could "simplify" and shorten this, actually I could imagine that such a tool in hand of folks
not yet really very familiar with how filters (together with speakers!) are working, could prevent them to find the right of the xx million possible settings (and I`m quite convinced that for a given speaker system there are not too many that will fit very well) as they are likely tempted just to zap back and forth the filter-settings, without ever being able finding the right ones.
My point I try to make is just this: the likelihood to build a good sounding speaker/ (active)crossover combo is MUCH higher through
first.) getting a more complete understanding of:
a.) how does an (active) filter work by itself basically(*)
b.) what acoustically effects has (or can have) a filter in INTERACTION with an actual speaker driver
c.) what acoustically effects can have two (or a x -number of) drivers with their respective filters and when they work together (+ under consideration when physically placed here or there and in respect to each other)
d.) what effects can have the (dimensions for instance) room and the speaker enclosure and speaker placement on the performance
+second.) WHENEVER possible doing some ACTUAL measurements (drivers individually without crossover, with lowpass, with highpass, all together) - instead of building a universal adjustable circuit that possibly can "everything" and just "zapping-through" settings until it appears to sound somehow decent.
(*)and this part is really not that complicated as it might seem at the first glance - for beginners I can only (repeatedly, as others have done this before) recommend Don Lancaster: "The Active Filter Cookbook" - You won`t find a book with simpler descriptions (and even me understood it after reading several times, and at that times I still had NO CLUE about this things)
Once somebody has understood the "Cookbook" filter circuits, together with a better understanding of effects of points b.) to d.) - (which are finally the reasons why real-world crossovers differ from textbook designs), he already gets an idea where, why and which kind of circuit is feasible and in principle can build an active crossover, which even by comparing with a XVR1, does not lack any performance - with regard to possibility of topologies and settings only of course, not refinements of the discrete active circuits, whereas IMO the right choice of topology (order of filter) and right parts values (Q and f) and not to forget the relative volume levels of the individual crossover branches (in more than 2-way systems not so trivial and easy as it might look) matter more than the final execution of the active circuits itself (hope this statement does not cause a flame war).
Decent OP-Amps (huuuhh the devil word) and capacitors for line-level circuits are so cheap (much cheaper than an equivalent passive filter circuit after the power amp) and maybe only need slightly more time to breadboard than a passive crossover, so why not a simple experimental OP-Amp setup first, instead of designing a possibly "overcomplicated" and physically comparatively large (and supposed to be) "all-in-one" unit from the beginning?
You even don`t have to solder much - IC-sockets for fast changes of resistor and capacitor filter values (in order to change Q and f) can make life easy (of course You must power off the circuit before You change parts!).
When done with experimenting one can implement the active stages fixed in whatever flavour floats ones boat - breadboarded or PCB - discrete (again in whatever flavour) or just plain simple with OP-Amps (huhh..shame... I said it again).
For specific DIY active crossover needs I believe this approach is cheaper and easier to build and above all from what I said before likely leads to a more deliberate result (and therefore probably better) as this way people are (or must be) more aware of what they`re doing.
Also this ensures that only this kinds of circuity find implementation which are really needed - in contrary to a universal applicable circuit where half of it may be is left unused because not needed (or worse when it turns out that something desirable is not even included).
BTW: I see kinds of applications (again, for the DIYers) not covered even from a XVR1 - what about subtractive filters or LR-transform circuits (or maybe delays)?
The time needed to "play" with a passive x-over until it`s "right" is at least the same as with an active solution (or more, as interaction of passive filter circuits with a real life speaker drivers sometimes can cause unexpected and surprising effects), passive filter parts are more expensive or how Nelson said active finally is more bang-for-the-buck (and not only this but active ultimately is overall superior IMO - of course when applied with the same care) what only underlines going the active route. Not to mention that some features are just not possible passively and others very hard to do (only think about attenuation of a bass driver where series resistors in a passive network could change box alignment noticable whereas going active You simply need to turn a pot without such effects).
As usual.......
Words of wisdom from Cap'n Pass.
Active crossovers are not that hard to make. The hard part is the fiddling part. And, yes, it will drive you nuts if you let it.
Lancaster's book on active filters has enough good info for all here, including all the "fudge factors" needed to change alignment, while retaining the proper crossover frequency.
Fiddle away, gang..........
Jocko
Words of wisdom from Cap'n Pass.
Active crossovers are not that hard to make. The hard part is the fiddling part. And, yes, it will drive you nuts if you let it.
Lancaster's book on active filters has enough good info for all here, including all the "fudge factors" needed to change alignment, while retaining the proper crossover frequency.
Fiddle away, gang..........
Jocko
I keep getting the impression that some people expect an active crossover to solve their problems for them. It won't. It is a tool that allows people to solve problems for themselves. The 'you have to work with it' aspect seems to slide right past them. They want pre-digested answers to their particular crossover needs--12 dB/oct at 2500Hz, etc. The closest that can be arranged in the real world is to provide sufficient flexibility that someone can get the job done without being hampered by the hardware.
Someone seemed unhappy that I posted a Sallen-Key filter in the Xenover thread. I'm not clear what they expected--a new filter topology, perhaps? (As though the world really needs a new filter topology--what we really need are people applying the ones we have now intelligently.) Others seemed to want it more complicated, as though throwing enough parts into the circuit would endow the poor thing with sufficient intelligence to set itself. That Behringer dingus will auto-set itself if you punch the right sequence of buttons, but it's not a crossover, it's an eq. I suppose you could contrive some way to use two of them to eq the tweeters and woofers in a biamp system, but you'd still have to deal with the fact that the result is pretty 'digital' in sound quality. Ugh. Maybe that idea appeals to others, but not to me.
If you can adjust the frequency, Q, slope, and gain for each driver, you've got about all the control there is. If you're not willing to sit down and try things, you're not going to get satisfactory results, no matter how elaborate the filter is.
People who get fixated on slopes (6, 12, 18, 24, etc.) or Qs (Bessel, Butterworth, and I suppose you could include Linkwitz-Riley, as it presumes a predermined Q) are frequently happy with their choices, simply as a function of Beranek's Law. But there are often better choices. I've never needed more than 18 dB/oct to get the job done, and find myself falling back to lower and lower order slopes as time goes by. Still, if the crossover is flexible enough, then even L-R fanatics can get their fix without having the crossover get in the way.
Grey
Someone seemed unhappy that I posted a Sallen-Key filter in the Xenover thread. I'm not clear what they expected--a new filter topology, perhaps? (As though the world really needs a new filter topology--what we really need are people applying the ones we have now intelligently.) Others seemed to want it more complicated, as though throwing enough parts into the circuit would endow the poor thing with sufficient intelligence to set itself. That Behringer dingus will auto-set itself if you punch the right sequence of buttons, but it's not a crossover, it's an eq. I suppose you could contrive some way to use two of them to eq the tweeters and woofers in a biamp system, but you'd still have to deal with the fact that the result is pretty 'digital' in sound quality. Ugh. Maybe that idea appeals to others, but not to me.
If you can adjust the frequency, Q, slope, and gain for each driver, you've got about all the control there is. If you're not willing to sit down and try things, you're not going to get satisfactory results, no matter how elaborate the filter is.
People who get fixated on slopes (6, 12, 18, 24, etc.) or Qs (Bessel, Butterworth, and I suppose you could include Linkwitz-Riley, as it presumes a predermined Q) are frequently happy with their choices, simply as a function of Beranek's Law. But there are often better choices. I've never needed more than 18 dB/oct to get the job done, and find myself falling back to lower and lower order slopes as time goes by. Still, if the crossover is flexible enough, then even L-R fanatics can get their fix without having the crossover get in the way.
Grey
Well put. DIYers should revel in the opportunities provided
by simple Sallen-Keys. Crossover theory is great to contemplate
and even occasionally solves a problem or two, but the fact
remains that driver behaviour is approximated at best, not to
mention the rest of the conditions, and not even considering
taste.
I keep repeating: Lots of experimenting and good ears.
by simple Sallen-Keys. Crossover theory is great to contemplate
and even occasionally solves a problem or two, but the fact
remains that driver behaviour is approximated at best, not to
mention the rest of the conditions, and not even considering
taste.
I keep repeating: Lots of experimenting and good ears.
Aw heck. I thought I was the one that came up with this.kelticwizard said:
Key phrase here: "...is convinced that he has made a wise choice of design". Even if you are convinced, posting your supposition here will soon make you retreat in terror, go lick your wounds, and stop liking the sound of your system any more. Like the guy who posted that Bose was his favorite kind of speaker.
It takes a strong will (or unwavering trust in a good pair of ears) to stand against peer pressure.
The question then becomes: Who do you regard as your peers?
The question is non-trivial. The more you think about it, the more layers it has; almost a koan. I used to listen to people who pounded a particular drum loudly. Surely someone who's that sure of themselves must know what they're talking about, right? Well...no. Often the exact opposite. These days I try to surround myself with folks who are open to many, many possibilities and try to be that way, myself. I like tubes and solid state. I like low feedback circuits for some things, but higher feedback is useful in other areas. I distrust people who try to shove one narrow viewpoint down your throat, moreso if they're loud and overbearing about it. At best, they're only going to be right under a very limited set of circumstances--and dead wrong the rest of the time.
Like people who grow up thinking Chevy makes the only cars worth owning, if all you hear about in your circle of friends is Linkwitz-Riley...Linkwitz-Riley...it will wear a rut in your mind even before you get around to designing your own speakers. Once you begin making decisions as to crossover points and slopes, the term Linkwitz-Riley is going to pop out, sure as daybreak. Ditto for Butterworth, Bessel, etc. Assuming that my reverse-engineering of Nelson's XVR1 Q selections is correct, none of them are a 'named' Q in the sense of .707 being Butterworth. That's certain to make people who regard themselves as belonging to a certain camp nervous.
Q isn't a choice among two or three discrete possibilities, it's a long, sliding scale. Very versatile. Want true adjustability? Replace the Q-setting resistors in the XVR1 or the Xenover with a pot and dial in whatever Q you want. Who cares if it has a 'name' or not? The only question that matters is: Does it work in your system?
Unfortunately, that requires making decisions, and the decisions will require testing, listening, or both. The vast majority of DIYers have no one to hold their hands as they struggle through the process the first time or two. Add to that the natural human tendency to want instant results, and crossovers (whether active or passive) become instruments of torture rather than the salvation that people expect.
As someone said recently, it's the journey, not the arriving.
Crossovers are definitely a journey.
If you want to arrive instantly at a destination, do not set foot upon the crossover path or, like Frodo and Sam, there's no telling where you'll end up.
Paul,
Beranek's Law is one of the few Laws in electronics that deals directly with the human element. I cherish it.
Being the owner of a perfect pair of speakers, myself, I know whereof I speak.
Now, if only I could remember where I left my sanity, I'd be whole again...
Grey
The question then becomes: Who do you regard as your peers?
The question is non-trivial. The more you think about it, the more layers it has; almost a koan. I used to listen to people who pounded a particular drum loudly. Surely someone who's that sure of themselves must know what they're talking about, right? Well...no. Often the exact opposite. These days I try to surround myself with folks who are open to many, many possibilities and try to be that way, myself. I like tubes and solid state. I like low feedback circuits for some things, but higher feedback is useful in other areas. I distrust people who try to shove one narrow viewpoint down your throat, moreso if they're loud and overbearing about it. At best, they're only going to be right under a very limited set of circumstances--and dead wrong the rest of the time.
Like people who grow up thinking Chevy makes the only cars worth owning, if all you hear about in your circle of friends is Linkwitz-Riley...Linkwitz-Riley...it will wear a rut in your mind even before you get around to designing your own speakers. Once you begin making decisions as to crossover points and slopes, the term Linkwitz-Riley is going to pop out, sure as daybreak. Ditto for Butterworth, Bessel, etc. Assuming that my reverse-engineering of Nelson's XVR1 Q selections is correct, none of them are a 'named' Q in the sense of .707 being Butterworth. That's certain to make people who regard themselves as belonging to a certain camp nervous.
Q isn't a choice among two or three discrete possibilities, it's a long, sliding scale. Very versatile. Want true adjustability? Replace the Q-setting resistors in the XVR1 or the Xenover with a pot and dial in whatever Q you want. Who cares if it has a 'name' or not? The only question that matters is: Does it work in your system?
Unfortunately, that requires making decisions, and the decisions will require testing, listening, or both. The vast majority of DIYers have no one to hold their hands as they struggle through the process the first time or two. Add to that the natural human tendency to want instant results, and crossovers (whether active or passive) become instruments of torture rather than the salvation that people expect.
As someone said recently, it's the journey, not the arriving.
Crossovers are definitely a journey.
If you want to arrive instantly at a destination, do not set foot upon the crossover path or, like Frodo and Sam, there's no telling where you'll end up.
Paul,
Beranek's Law is one of the few Laws in electronics that deals directly with the human element. I cherish it.
Being the owner of a perfect pair of speakers, myself, I know whereof I speak.
Now, if only I could remember where I left my sanity, I'd be whole again...
Grey
- Status
- Not open for further replies.