jewilson said:
and what about 1st order filters?
not bessel (they are butterwotrh) and have a good phase and impulse response
jewilson said:😱
Use them all the time.
You use them all the time?
Where can I see your loudspeakers? 😉
yeah but its still the same: one disadvantage brings up another advantage.
If you want 2-way 1st order without vertical dispersion irregularities you'll almost need to make use of coaxial designs (with the small choice in drivertypes and doppler/modulation distortion etc. all of that). It would probably work nice with very long vert. ribbons placed very close to each other. (bass speaker array next to long ribbon etc)
It is certailny nice the advantage of the 1st order but the overlap-band is huge, and drivers need to be much more broadband. Very little phase-rotation but also not a lot of filtering. And the distance between the drivers play a bigger role in the vertical behaviour, compared with higher orders.
Steeper filtering (with the disadvantage of phase steps) has also a lot of pro's, because the problem-area can be made smaller and maybe less audible.... Btw is there anyone who has heard such a phasestep in a proper configured design (not in a misaligned config orso) I don't know. If I configure filters I always end up with something inbetween the both, giving best results.
I quite much agree with jewilson.
If you want 2-way 1st order without vertical dispersion irregularities you'll almost need to make use of coaxial designs (with the small choice in drivertypes and doppler/modulation distortion etc. all of that). It would probably work nice with very long vert. ribbons placed very close to each other. (bass speaker array next to long ribbon etc)
It is certailny nice the advantage of the 1st order but the overlap-band is huge, and drivers need to be much more broadband. Very little phase-rotation but also not a lot of filtering. And the distance between the drivers play a bigger role in the vertical behaviour, compared with higher orders.
Steeper filtering (with the disadvantage of phase steps) has also a lot of pro's, because the problem-area can be made smaller and maybe less audible.... Btw is there anyone who has heard such a phasestep in a proper configured design (not in a misaligned config orso) I don't know. If I configure filters I always end up with something inbetween the both, giving best results.
I quite much agree with jewilson.
jewilson said:
The 2 way TM will be my 1st project 😉
I've been gathering informations for nearly one year...
The Sonus TM are very nice sounding speakers. Of course there are a lot of great 2 way speakers.
jewislon,
I know about basic filter theory but since there are a number of solutions that gives perfect or almost perfect phase response (perfect time domain behaviour) I was surprised to read what you wrote.
I guess it would be correct to say someting like "of the most commopnly used high order filters, the bessel has best phase response.." or something like that.
1st butterworth pass squarewaves as far as I know. Then there are other alternatives as filler driver (originally B&O I believe) and various subtractive delay networks and so on..
Maybe you are aware of John K´s work on transient perfect x-over networks?
/Peter
I know about basic filter theory but since there are a number of solutions that gives perfect or almost perfect phase response (perfect time domain behaviour) I was surprised to read what you wrote.
I guess it would be correct to say someting like "of the most commopnly used high order filters, the bessel has best phase response.." or something like that.
1st butterworth pass squarewaves as far as I know. Then there are other alternatives as filler driver (originally B&O I believe) and various subtractive delay networks and so on..
Maybe you are aware of John K´s work on transient perfect x-over networks?
/Peter
SF speakers is nice but those I´ve heard can be bettered on many parameters... it´s a matter of taste in the end of course.
Due to the design they are very costly for what you get in sound though.
I know of no SF speaker that is transient perfect and I have yet to se a SF or Dynaudio speaker that is "TP".
Using a cap (or a sinecap filter) to the tweeter and a coil to the midbass does not normally result in TP even if it has a flat Fr. response. Possibly Dynaudio has true 6dB/oct slopes, but seem to revers the polarity on some of the drivers making them fail the squarewave test.
/Peter
Due to the design they are very costly for what you get in sound though.
I know of no SF speaker that is transient perfect and I have yet to se a SF or Dynaudio speaker that is "TP".
Using a cap (or a sinecap filter) to the tweeter and a coil to the midbass does not normally result in TP even if it has a flat Fr. response. Possibly Dynaudio has true 6dB/oct slopes, but seem to revers the polarity on some of the drivers making them fail the squarewave test.
/Peter
On the "Cookbook"- and some philosophy
I believe the original call was for topologies, and questions for books about loudspeakers...
Vance Dickasons Cookbok is exactly what it claims to be;- a cookbook. It contains a lot of information for people to get started with loudspeaker construction and building,- as pr. "cookbook " principles. It does help to know some electronics , and even better some physics as well. In Addition it gives several sets of "ready-to -use" formulaes for getting started with a very interesting pastime occupation.. for those so inclined 😉
The cookbook must not be looked upon as "The Holy Book" of loudspeakers, and I don't think Mr. Dickason has implied so either. Getting deeper and more seriously into loudspeaker theory, is rather like digging for pyramids,- the more you find, the more you realize the size of what has not yet been unveiled......
The physical implication and understanding of the loudspeaker as a whole, with each operating element being what it really is,- an electro-mechanical resonance system, with all parts electrically equivalently presenting themselves as a complex circuit of inductances, capacitances and resitances, gives just the first clues as to what we are up against.......
The problem with any loudspeaker, unless we manage to produce full short curcuits somewhere, is that no matter how false we thread, - there will always be sound, - and invariably we will perceive it a product of our strong efforts, each of us with our personal bias, be it in one direction or the other😀
The "Cookbook " is truly a piece of great work, but I am sorry to say,- it's just the beginning!
If you really want to get to grips of what's going on- get your math and physics hat on, and start reading- it will truly be a long journey! This said, it is absolutely not to frighten anyone from having great fun as we go along!
Basic litterature; Leo Beranek- Acoustics The mother-of-all book in the Acoustics field
Harry F. Olsson - Musical Engineering. Both books from the 50-s but available second hand ( Amazon ?)
Martin Colloms- High performance loudspeakers ( still in print, - new edition out). don't be afraid of the long s's.....( integral math 😉 )
Good luck and have fun......
( I've had for 35 years, and there are still times I wonder if I realy understand a sh*t!??)🙂
I believe the original call was for topologies, and questions for books about loudspeakers...
Vance Dickasons Cookbok is exactly what it claims to be;- a cookbook. It contains a lot of information for people to get started with loudspeaker construction and building,- as pr. "cookbook " principles. It does help to know some electronics , and even better some physics as well. In Addition it gives several sets of "ready-to -use" formulaes for getting started with a very interesting pastime occupation.. for those so inclined 😉
The cookbook must not be looked upon as "The Holy Book" of loudspeakers, and I don't think Mr. Dickason has implied so either. Getting deeper and more seriously into loudspeaker theory, is rather like digging for pyramids,- the more you find, the more you realize the size of what has not yet been unveiled......
The physical implication and understanding of the loudspeaker as a whole, with each operating element being what it really is,- an electro-mechanical resonance system, with all parts electrically equivalently presenting themselves as a complex circuit of inductances, capacitances and resitances, gives just the first clues as to what we are up against.......
The problem with any loudspeaker, unless we manage to produce full short curcuits somewhere, is that no matter how false we thread, - there will always be sound, - and invariably we will perceive it a product of our strong efforts, each of us with our personal bias, be it in one direction or the other😀
The "Cookbook " is truly a piece of great work, but I am sorry to say,- it's just the beginning!
If you really want to get to grips of what's going on- get your math and physics hat on, and start reading- it will truly be a long journey! This said, it is absolutely not to frighten anyone from having great fun as we go along!
Basic litterature; Leo Beranek- Acoustics The mother-of-all book in the Acoustics field
Harry F. Olsson - Musical Engineering. Both books from the 50-s but available second hand ( Amazon ?)
Martin Colloms- High performance loudspeakers ( still in print, - new edition out). don't be afraid of the long s's.....( integral math 😉 )
Good luck and have fun......
( I've had for 35 years, and there are still times I wonder if I realy understand a sh*t!??)🙂
So, is there no solution to have at the same time aligned accoustic centers AND a horizontal main lobe?
Bricolo said:
Yeah there is.... coaxial, and even with the use of nonlineair phase filtering, there still is a horizontal main lobe.
oops could have thought of that myself..... 🙂
Well active there is, passive not really, but almost.
And I think if you use a not too big ribbon that still goes quite low , (so more of a mid/high unit from 800hz/1khz and up, like the SA ribbon) it is probably possible to get close with a first order passive filter.
Well active there is, passive not really, but almost.
And I think if you use a not too big ribbon that still goes quite low , (so more of a mid/high unit from 800hz/1khz and up, like the SA ribbon) it is probably possible to get close with a first order passive filter.
dokter dB said:
Can you explain the "almost" solution? 😉
I was more about to use a 1" dome tweeter, but a ribbon isn't excluded
Bricolo said:
Well it is impossible to cross passive, without phase-rotation.
There are passive solutions to compensate a bit with phase-circuits (drivers in a H-network etc).
Maybe someone else has good experiences with that, I never went that way, because I felt I'd never get it really right in terms of efficiency and complexity.
But the closer the drivers are together physically, and the lower you can cross, the broader the vertical lobe will be (directivity of the drivers itself excluded off course). Keep in mind that a too broad lobe can be unwanted because of floor/ceiling reflections, but in MT you can't really do a lot about that, other than with the baffle shape.
The less steep you cross (the less phase-difference between the two drivers you add), the less irregularities in the vertical dispersion-pattern around the crosspoint (vertical dispersion stays more uniform, even around the crosspoint). Unfortunately the drivers need to be closer in 1st order compared with higher order filters, because above the cosspoint there is a bigger frequency part where both drivers work together.
So if you have this goal then you can go for a ribbon that will perform quite low, and is able to cross 1st order 1.5 kHz orso (can be a problem to find one, excursion limits may still be reached too fast under 1.5 k). One disadvantage of the bigger ribbons is that the high frequency vert. dispersion pattern is getting very small (a function of ribbonheight and frequency). The only way to overcome that is with an acoustic lens/horn.
hmm... anyone else/vision?
(trollish?) ANSWER -- stop being an "audiophile" -- also TM(W)
Fullrange
=========
Advantages - It's easy
Disadvantages - Desperation isn't good to write the least. Bad at all frequency reproduction. Not great at highs, mids or lows.
Coax
====
Advantages - Makes some nice graphs to impress you're friends (if you weren't an "audiophile", you would have these).
Disadvantages - Have to combine drivers into one. It's hard making 1 great driver. Tweeters often horn loaded into cone, shape is not good.
TM - MT
========
Advantages - You choose what you like. People notice VASTLY different things in audio, visual, taste, life. OFF-AXIS response; you can get up, other people (you're non-existent friends) can enjoy the sound also.
Disadvantages - Lobing, okay
FAST is a good choice; I don't like a lot of highs.
TMW is a better choice.
TMW is optimal design for most people.
Due to physics, highs are mostly done right, bass is not.
3-4" driver and 8-10" woofer is also a very good choice. Is that "fullrange"? 😆
Wait, no, actually I'm going to agree with COAX 🤣 but not for me 🤔
+1 for a FAST or WAW (Woofer Assisted Wideband) works for me. Similar xover freq as a three way Woofer to Mid i.e. lower xover freq mitigates some of the problems mentioned previously.
See https://www.diyaudio.com/community/threads/10f-8424-rs225-8-fast-waw-ref-monitor.273524/
for a FAST with 1st order xover.
Also what should be mentioned is a tweeter in a waveguide. Narrows the off axis response in the lower treble to 'match' the woofer. Does some time alignment also as the tweeter is further back and increases output at the lower end therefore reduces distortion and allows a lower xover freq.
See https://www.diyaudio.com/community/threads/the-elsinore-project-thread.97043/ for a waveguided tweeter and 1st order xover.