Hi,
This assumes a perfectly flat woofer passband to several octaves above the crossover point which is rare, if it ever exists and it assumes a specific Qt for MF/Satellite speaker to form the other halve of a 4th order LR, which essentially again rarely if ever exists.
So using a textbook crossover in this situation is unlikely to result in a reasonably integrated, flat response. Of course this may not what you want. One also would enquire if you verified, post implementation, by measurements, that the textbook crossover produced the desired results?
Ciao T
This assumes a perfectly flat woofer passband to several octaves above the crossover point which is rare, if it ever exists and it assumes a specific Qt for MF/Satellite speaker to form the other halve of a 4th order LR, which essentially again rarely if ever exists.
So using a textbook crossover in this situation is unlikely to result in a reasonably integrated, flat response. Of course this may not what you want. One also would enquire if you verified, post implementation, by measurements, that the textbook crossover produced the desired results?
Ciao T
Hi,
Actually, I do not offer this as a solution, I warn against it and point out that is a most stupid way to do things, regardless if done active analogue, passive analogue or DSP.
I agree. Even before exposure to Calsod and more advanced modelling systems I did not attempt to ever work with textbook crossovers, they really do not work.
Ciao T
Actually, I do not offer this as a solution, I warn against it and point out that is a most stupid way to do things, regardless if done active analogue, passive analogue or DSP.
I agree. Even before exposure to Calsod and more advanced modelling systems I did not attempt to ever work with textbook crossovers, they really do not work.
Ciao T
Whoa !
Using the 5532 today *anywhere* in an audio circuit that is after some sonic quality (not mere measurable specs !) is even more of a museal thinking and acting than I can stand...
Also *if* you gonna propose such mad circuits, you will have to add a ton of layout tricks and other related stuff to make it finally work for the usual DIYer
No good attempt IMO
Besides that - as already perfectly stated - "simple" optimizing measurable performance is way outdated - for good reason...
Michael
Last edited:
Hi,
I cannot see what the argument is about. Information is information,
if you don't like it, don't read it. If it isn't done done the way you'd
prefer, then you right the book, you don't shoot the messenger.
Differences in philosophy do not change the validity or not of information.
Just as vinyl has never died, analogue amplifier design will never die.
I've found all D.Selfs books informative, you don't have to use the
information presented in the manner described, that is up to you.
rgds, sreten.
I cannot see what the argument is about. Information is information,
if you don't like it, don't read it. If it isn't done done the way you'd
prefer, then you right the book, you don't shoot the messenger.
Differences in philosophy do not change the validity or not of information.
Just as vinyl has never died, analogue amplifier design will never die.
I've found all D.Selfs books informative, you don't have to use the
information presented in the manner described, that is up to you.
rgds, sreten.
I see that I need not doubt your firm grasp of the obvious. But be careful with this statement. Although the lowest TV band is around 50MHz, the fundamental difference between TV and (non-RF) audio is downconversion. After downconversion follows a band filter (in zero-IF a LPF with fc of ~5MHz) and a low-IF DVB-T channel filter has even smaller bandwidth. I'll agree with you that 5MHz is still 10^3 larger than 5kHz, but with complicated audio filters, capacitor values can get very close between the two.
Good for you. 2n2 means nice little MKP. I'm guessing this was not a high-order elliptic filter like the ones Herr Doktor Joachim Gerhard might perhaps have dreamed of during his time at Audio Physic.
The short answer is yes.
Although, from your question, I suspect you might either think this not a serious topic or even call it hokum, I will still expand on my answer, since the matter is of course vastly more complicated than simply "Volt in => SPL out".
Let's suppose we have a tweeter and midwoofer sitting next to each other. Like a double-slit experiment with lasers, this setup tends to create radial peaks and troughs in the radiation pattern, along the axis on which the drivers are aligned. With time-delay or a deliberate phase relationship between the drivers (or just tilting the whole box...), the main lobe (peak) can be directed. There are rules-of thumb for how much the amplitude increases in the crossover region due to this effect (Linkwitz, IIRC).
There is, however, a second effect, similar to two other things I understand in terms of tapered transmission lines. The first (this is fun, maybe a bit less for you as a professional than for me as an amateur Hi-fi fan) is that when you hold your car key against your chin, you extend its useful range. The human body has an impedance in between the antenna impedance and the impedance of open air, acting as tapered transmission line. The second is that I conjecture that a coaxial audio interconnect should be wider near the amplifier and narrower at the source, if it at all matters, because the source has low impedance and a good amplifier a high input impedance (and a wider coax has a higher impedance).
For loudspeaker drivers something similar might be observed. If the effective total cone/membrane area is larger, the acoustic impedance (this terminology is probably wrong) is more closely matched to the room, giving better coupling and therefore higher output SPL/Volt. This total cone/membrane area is largest at the XO point, since two (or, in MTM, three) drivers are seriously pushing the air in the room around at the same time. The effect, to my understanding, will differ per configuration, hence the question regarding a short list of standard configurations accompanied by dB numbers indicating the dip that must be applied to the summed filter output in the XO region (on top of the corrections that are apparent from driver datasheets, baffle shapes, bass reflex ports, etc.). This list would then give a designer a much better first shot at an XO design.
BTW, tapered transmission lines are another topic requiring extreme analytical mathematics techniques. In a 1989 paper, a particular Fourier transform pair was introduced to allow proper analytical treatment.
--
This should conclude our final point of discussion concerning active crossover filters. No hard feelings if you do not care about/for the acoustic impedance argument. I'll enjoy the rest of the flamewar on DSP-vs.-Analog. Sure companies like Behringer and Genelec are starting the move out of analog XOs into DSPs; that does not make the book obsolete. There will still be plenty of opportunity to put all this knowledge to good use. The argument reminds me of a EUR30000 set of speakers with terrible phase alignment in the XO region. Any book on XO design would have been a great help to its designer.@Moondog: Time delay is precisely not what I'm talking about. Time delay is calculated by inserting a ruler into your midrange/widwoofer (obviously not damaging the cone
) and doing the distance->delay time->filter conversion. No rule-of-thumb needed other than the speed of sound.--
Greetz,
MatchASM
It seems some of the posters here are commencing their own books online and criticising a mere table contents that they suspect will not fit their personal or group interests.
There is hypocrisy here too, over new technology with DSP in the presence of 100 year old glass tech. This is a ridiculous platform to argue from, regardless of one's subjective tastes in audio.
How about being a little constructive in the hope of eventually serving the wider DIYAudio community's interests rather than just the current favourite toys? Accepting language difficulties, it still is also better to keep posts down to a length that can be read and considered as concise and answerable queries or suggestions. This way we can all understand what you are actually saying.
However, if some do have a wider understanding that is not buried under metres of new-tech industry dogma and presumptions about the past, I assume you will have already written a huge book and published. Somehow, I doubt that will be the case.
There is hypocrisy here too, over new technology with DSP in the presence of 100 year old glass tech. This is a ridiculous platform to argue from, regardless of one's subjective tastes in audio.
How about being a little constructive in the hope of eventually serving the wider DIYAudio community's interests rather than just the current favourite toys? Accepting language difficulties, it still is also better to keep posts down to a length that can be read and considered as concise and answerable queries or suggestions. This way we can all understand what you are actually saying.
However, if some do have a wider understanding that is not buried under metres of new-tech industry dogma and presumptions about the past, I assume you will have already written a huge book and published. Somehow, I doubt that will be the case.
Hi,
Actually, in our PA (Sound Reinforcement System) we replaced the analogue crossover with a digital one (Yamaha IIRC) in 92 and the EQ went digital in 93. The main reasons where actually adjustability, repeatability and the avialbility of memory slots and the way you could save more settings to Midi Computers.
What Behringer, BSS and dBX among others started maybe a decade or more ago was to push this technology into the mainstream, comoditised Pro-Audio Market.
Ciao T
Actually, in our PA (Sound Reinforcement System) we replaced the analogue crossover with a digital one (Yamaha IIRC) in 92 and the EQ went digital in 93. The main reasons where actually adjustability, repeatability and the avialbility of memory slots and the way you could save more settings to Midi Computers.
What Behringer, BSS and dBX among others started maybe a decade or more ago was to push this technology into the mainstream, comoditised Pro-Audio Market.
Ciao T
Sure, wasn't that kind of input asked for ?
The question below was answered - by some - that present tense and outlook to future is "missing" and also general focus may not exactly be DIY friendly in the context of this board...
But nothing wrong with it as long as many would like to buy "as is" (lukewarm to cold coffee spread over a 500 pages that is)
Michael
Last edited:
Although crossovers are discussed in only a small part of "Sound System Design and Optimization" by Bob McCarthy the approach taken is so novel that it cannot be ignored. By extending the crossover discussion to include, in fact to emphasize, summation *in the room*, and expanding the concept to spatial as well as spectral "crossover", McCarthy forces a rethink of a number of "settled" concepts. No circuit theory or detail at all, though, just a different way of looking at what's important . . .
Sure, wasn't that kind of input asked for ?
.........................................
In subject yes, but in the form of emotionally charged jibes or a long diatribe is never good for credibility.
..........................................
The question below was answered - by some - that present tense and outlook to future is "missing" and also general focus may not exactly be DIY friendly in the context of this board...
But nothing wrong with it as long as many would like to buy "as is" (lukewarm to cold coffee spread over a 500 pages that is)
Actually, I would like to agree with your forward determination but the prohibitive development times, costs, skills and materials for such microscopic SMD requirements means no way for 99% of DIYs. Most will be obliged to leave this for button-pumping commercial toys when there are no alternatives. The rising popularity of active crossover projects on some DIY sites gives me a little assurance of this reality.
Hi Douglas If you are interested in seeing something along these lines (not the time delay, but active implementation of passive LC filters using GIC's) , have a look at my blog post http://www.diyaudio.com/forums/blogs/wintermute/245-synergy-active-crossover.html
Note that I still haven't actually built it, so it might not work, but apart from SY's acheron, I haven't seen anything else like it (ie being used for an active crossover) anywhere on the net. If you follow the link in the blog to SY's gyrator based Acheron crossover (which is what inspired me) then you will be looking at something that has been built and works, I'm taking SY on his word here, (just note that I don't think that SY has updated the typo yet
) I like the philosophy of making an active filter that for all intents and purposes behaves like a passive filter. I think that this has merit when considering designing a filter that is matched to the driver, although until I actually try it, I guess that is only a hunch.
I for one have no interest currently in using a dsp approach, sure it might be the current trendy thing to do and will no doubt become the norm as mass production makes it cheaper and cheaper, but as sreten points out, there are still people listening to vinyl! and there are plenty of diyers doing tube stuff too. No matter how far digital technology goes, I think there will still be interest in good old analogue
It's like high tech engines with ECU's v's old stuff with carburettor's and a distributor... the latter is much more accessible to most DIYers.
Tony.
Actually, I would like to agree with your forward determination but the prohibitive development times, costs, skills and materials for such microscopic SMD requirements means no way for 99% of DIYs. Most will be obliged to leave this for button-pumping commercial toys when there are no alternatives. The rising popularity of active crossover projects on some DIY sites gives me a little assurance of this reality.
Probably you misunderstood - I'm not advocating to go into DSP board DIYing nor into coding tricks.
My suggestion to Douglas was meant to move out of the "center of the middle" (of filter analysis) towards the "bleeding edge" of XO philosophy - providing a glance over possibilities *and brick wall limitations* as a counter balance to the immense expectations of never ending horse power available now and in the times to come that don't hold for easy to understand physical reasons.
This also would include to give up on the museal stand point this current book seems to be devoted to and also to re-think the whole stand point of XO being a "frequency divider tool" in favor to looking at XOs (and all related response shaping ) as to be a "spectrum splitting tool" - further meaning - highly emphasizing on the time domain aspects that got lost in "audio minds" over generations.
Quite a turn around - I admit...
Michael
Last edited:
If you follow the link in the blog to SY's gyrator based Acheron crossover (which is what inspired me) then you will be looking at something that has been built and works, I'm taking SY on his word here, (just note that I don't think that SY has updated the typo yet
No, I haven't.
To be fair, although I got several years use out of it, it's now on a shelf in the lab, with one of those shiny new digital units taking its place in my living room.
I think you've missed the point, here. No one is shooting the messenger.
Douglas asked for our opinions and input. We've been giving it. So far I have not seen any strong objections to the information in the book, just that the lack of info on DSP makes the book a look backward, not a look forward.
I have to say, if it were the year 1952 and I was publishing a big book called "The Design of Railroad Locomotives" that was based mostly on steam engines, or a big book called "Passenger Aircraft Design" that talked mostly about pistons and propellers - I think that I might want to know that those technologies, as wonderful as they are, are on the way out. And that someone designing those things and likely to buy my book might not buy it, because the obvious future path of the technology is given so little ink.
The theories explained, the engineering practice and the problem solving presented would all still be valid and useful, but who would buy such books in 1952 - if there is little to no mention of diesel locomotion or turbine and jet engines? And a book on crossovers that is heavy on opamp design while light on DSP may suffer the same fate in the year 2011.
So I don't see a major criticism in this thread of the information that Douglas plans to publish, just a criticism of giving the obvious future technology of active crossovers (DSP) so little coverage. Can't speak for Douglas, but that is something I certainly would want to hear from my potential audience.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.