Well strictly speaking, every speaker-mounted driver crossed over its input signal bandwidth, varying behavior and response in significant ways -- whether or not electrical, mechanical, and/or positional means were utilized to control and integrate their sound at the listening position (in stereo, both ears). Any BR crossed over from direct output to Helmholtz. In this forum however, XO by convention referred to electrical filter, passive/active/digital, no? So (electrical) XO-less is not necessarily single-driver. Run LXmini straight-through -- do they not belong in "multi-way"?
Maybe Dave can say what he meant? He promotes successful full range designs, though - frugal horns etc.
There is a lot of cross-over between forums. Comes up a lot when people ask for more niche topics.
No (explicit) crossover is an answer to “optimal”. And often people just dismiss 1-way speakers, nit realizing how good they can be these days.
Optimal XO depends on many factors. In this case, textbook is very, very rarely optimal. they should work on drivers with almost nop reactance like the Heil tweeter.
dave
No (explicit) crossover is an answer to “optimal”. And often people just dismiss 1-way speakers, nit realizing how good they can be these days.
Optimal XO depends on many factors. In this case, textbook is very, very rarely optimal. they should work on drivers with almost nop reactance like the Heil tweeter.
dave
reference 16 and 17 here is the ticket to filter wisdom: https://www.linkwitzlab.com/publications.htm
And google
RaneNote 160
for a easier to understand version
And google
RaneNote 160
for a easier to understand version
Last edited:
🙂 Ya know, do believe optimum crossovers are totally available.
And I do mean optimum...as in acoustically complementary for speaker drivers, not just "complementary electrical crossovers".
Optimum as in no group delay, no phase rotations to surf.
The price? always a price in audio heh?
Multi-way active, with FIR processing.
and FIR software that generates FIR filters that make acoustic outputs precisely match whatever desired complementarity "textbook" crossover order.
True price for sure.
But if it's within project reach you really should try it ...makes at least 95% of all crossover and filter agonizing seem plain primitive.
And hey, the kicker is, you won't believe how much easier it is to get killer measurements and sound than the old passive, ...or even IIR active way.
I began to read the thread and hoped to hear about some interesting topology curing a multi-way from its bad time response with non linear phase in the so important middle frequencies between 200 hz and 5khz.
I am a fan of fullrange drivers because they bring time coherency naturally with them.
And yes. You can hear that. Long time good time response was neglected but in tests of coax speakers in professional audio magazines they admitted that they have better imaging....
Since the availability of digital FIR filters you can construct a loudspeaker with some pretty standard filters as long as you take some overall FIR EQ and make it all flat in phase and frequency response.
However distortion and polar response should not be ignored.
I am a fan of fullrange drivers because they bring time coherency naturally with them.
And yes. You can hear that. Long time good time response was neglected but in tests of coax speakers in professional audio magazines they admitted that they have better imaging....
Since the availability of digital FIR filters you can construct a loudspeaker with some pretty standard filters as long as you take some overall FIR EQ and make it all flat in phase and frequency response.
However distortion and polar response should not be ignored.
Yes.
Since phase is tied to response in a good design, individual FIR shouldn't be necessary if you can get the response right.. and if you can't then you have acoustic issues and FIR won't fix those.
IMO one of the main issues of this thread is the incorrect use of optimal and ideal.
I presume that the OP meant an ideal crossover (which this one isn't of course and not only because such a thing will never exist) while he called it optimal.
An optimum crossover is the one that is the best overall compromise for a given situation and target behaviour. So it wouldn't even look the same for a given speaker topology but different design philosphy. One developer is optimising for amplitude response linearity while the other one is optimising the temporal behaviour for instance. Both their designs are optimised for coming as close to their intended target as possible and both crossovers will definitley look different and will also have different downsides. And I haven't even started about taking driver properties into consideration....
Regards
Charles
I presume that the OP meant an ideal crossover (which this one isn't of course and not only because such a thing will never exist) while he called it optimal.
An optimum crossover is the one that is the best overall compromise for a given situation and target behaviour. So it wouldn't even look the same for a given speaker topology but different design philosphy. One developer is optimising for amplitude response linearity while the other one is optimising the temporal behaviour for instance. Both their designs are optimised for coming as close to their intended target as possible and both crossovers will definitley look different and will also have different downsides. And I haven't even started about taking driver properties into consideration....
Regards
Charles
Last edited by a moderator:
Nice thread.
I hereby copyright the following crossover frequencies:
500Hz
4000Hz
If you plan to use any of them my PayPal can be asked for on PM
Thanks.
I hereby copyright the following crossover frequencies:
500Hz
4000Hz
If you plan to use any of them my PayPal can be asked for on PM
Thanks.
I will protect my rights on 501 hz and 4001 herz!
Any offence will be prosecuted!
"The first to surround a piece of land with a fence and to think of saying, 'This is mine,' and who found people foolish enough to believe him, was the real founder of bourgeois society. How many crimes, wars, murders, how much misery and terror the human race would have been spared if someone had uprooted the stakes and cried out to his fellows: "Beware of believing the deceiver; you are lost if you forget that the fruits belong to everyone, but the earth belongs to none."
―Jean-Jacques Rousseau -
Read the original:
https://archive.org/details/discourseoninequ0000rous/page/n1/mode/2up
Sorry archive org did put a fence around it, here for free
https://www.files.ethz.ch/isn/125494/5019_Rousseau_Discourse_on_the_Origin_of_Inequality.pdf
Any offence will be prosecuted!
"The first to surround a piece of land with a fence and to think of saying, 'This is mine,' and who found people foolish enough to believe him, was the real founder of bourgeois society. How many crimes, wars, murders, how much misery and terror the human race would have been spared if someone had uprooted the stakes and cried out to his fellows: "Beware of believing the deceiver; you are lost if you forget that the fruits belong to everyone, but the earth belongs to none."
―Jean-Jacques Rousseau -
Read the original:
https://archive.org/details/discourseoninequ0000rous/page/n1/mode/2up
Sorry archive org did put a fence around it, here for free
https://www.files.ethz.ch/isn/125494/5019_Rousseau_Discourse_on_the_Origin_of_Inequality.pdf
Last edited:
I was under the impression the OP already called dibs on 3500 (and 350) but 250 should be fine.
Edit: on a second thought maybe it’s the combinations of low-mid mid-high we can copyright not the individual xo-points, you may be within acceptable deviation from the OPs copyright, especially if you use 1st order and/or 3 order. Personally I call dibs on reversed polarity for mids as of now. Sorry I’m a little bored on vacation I’ll stop this now.
Lol
Last edited:
And probably more than one set of compromises.Charles gives a good example.
dave
Hi Charles,
I can't see any reason we shouldn't have both amplitude response linearity and optimal temporal behavior together.
Hey good ole flat mag and phase.
(Both on-axis and off-axis smoothly extending, per acoustic design.)
In fact, I'd say that's my definition of optimum...🙂
And I think if have a lot of crossover compromises to consider, after having built a physical acoustic design, it usually means we have a compromised acoustic design we are trying to juggle with.
Last edited:
Dear Scottmoose
I would like you to show me a 3-way 2nd order Butterworth crossover that has a constant power output at all frequencies 20 to 20kHz. Have the crossover points 360 Hz, and 3600 Hz, Use only standard value parts (that could be purchased from places like Parts Express). Publish it so we can run it through our little circuit analysis programs.
I would like you to show me a 3-way 2nd order Butterworth crossover that has a constant power output at all frequencies 20 to 20kHz. Have the crossover points 360 Hz, and 3600 Hz, Use only standard value parts (that could be purchased from places like Parts Express). Publish it so we can run it through our little circuit analysis programs.
Allenb
You can calculate power. If you know the input impedance. Let us suppose for simplicity the crossover is lossless. My crossover is in fact lossless. It has no resistors so it is lossless. If a given crossover has an impedance, 15.01 -j1.451.
At 400 Hz this crossover’s input impedance is: 15.01 - j1.451.
Let us make our reference input voltage 2.83 volts.
To calculate the power input to the crossover, when a voltage of 2.83 volts is applied to the input with a complex impedance of 15.01 - j 1.451, we will follow this process:
First, we need to find the resistance (Z) of the load, which is the magnitude of the complex impedance:
Z = √(15.01^2 + (-1.451)^2)
Z = √(225.3001 + 2.103401)
Z = √227.403501
Z ≈ 15.083
Now, we can calculate the power into the load at 400 Hz.
P = E^2 / R
P = (2.83)^2 / 15.083
P = 8.0089 / 15.083
P ≈ 0.5314 watts.
Therefore, when a voltage of 2.83 volts is applied to the load with a complex impedance of 15.01 - j1.45 (approximately 15.083 ohms), the power consumed by the load will be approximately 0.5314 watts. That is
2.8 dB below the reference level of 1.0 Watt into a 8.0 Ohm load.
Ezpz right 🙂
*****
You can calculate power. If you know the input impedance. Let us suppose for simplicity the crossover is lossless. My crossover is in fact lossless. It has no resistors so it is lossless. If a given crossover has an impedance, 15.01 -j1.451.
At 400 Hz this crossover’s input impedance is: 15.01 - j1.451.
Let us make our reference input voltage 2.83 volts.
To calculate the power input to the crossover, when a voltage of 2.83 volts is applied to the input with a complex impedance of 15.01 - j 1.451, we will follow this process:
First, we need to find the resistance (Z) of the load, which is the magnitude of the complex impedance:
Z = √(15.01^2 + (-1.451)^2)
Z = √(225.3001 + 2.103401)
Z = √227.403501
Z ≈ 15.083
Now, we can calculate the power into the load at 400 Hz.
P = E^2 / R
P = (2.83)^2 / 15.083
P = 8.0089 / 15.083
P ≈ 0.5314 watts.
Therefore, when a voltage of 2.83 volts is applied to the load with a complex impedance of 15.01 - j1.45 (approximately 15.083 ohms), the power consumed by the load will be approximately 0.5314 watts. That is
2.8 dB below the reference level of 1.0 Watt into a 8.0 Ohm load.
Ezpz right 🙂
*****
Last edited:
When power is mentioned in the context of speaker design, it normally refers to the acoustic power effected by the speakers into the room. This depends on certain acoustic matters as well as transducer properties so from the point of view of input power, these properties need to be specified or we don't know anything useful. Besides, dealing with speaker impedance is the job of the amplifier. I'm not saying that's trivial, what I'm saying is that the amplifier is assumed to be ideal from the point of view of speaker design.. (unless said design requires specific amplification.)
Easy. As you know, the crossover you picture in your first post is in fact nothing more than the standard 2nd order electical Butterworth values (rounded to nearest commonly available components) for the stated frequencies, which you have plagiarised by attempting to claim copyright for. So if you are saying that provides what you are defining as power to your satisfaction, you have answered your own request. QED.
As a brief side-note, you claim 'your' crossover is lossless because it has no resistors. So you seem to be claiming that capacitors have no ESR and inductors no DCR (or any other losses, for that matter, however small). That has some fairly formidable obstacles in its way, since it contravenes some of the more basic laws of physics & engineering. 😉
Last edited: