Hello
I have been searching and simulating for a while on active filters, especailly those Linkwitz-riley and bessel, I was surfing in www.linkwitzlab.com and www.sound.au.com. I think its time to crystallize my thoughts now..
Any way, I have two 12" loudspeakers which are best dedicated for midrange audio spectrum, and two 4 Ohm tweeters, those types which have a coil, means no peizo, cuz I don't like the way it sounds. I will buy th subwoofer later on, while discussing this system on the forum..
If I want to build 4th order LR HP filters based on Rod's formulas, and use a cut-off frequency of 250Hz, or 310Hz. And build a 4th order LR LP filter which sums the L & R channels, and has variable gain circuit based on an opamp like that described in linkwitzlab, also will have a cut-off frequency which is the same as that of LR HP filter. I am aware of the fact that I will have to use Linkwitz transform circuit if I use a closed box for the sub, and may be the sub controller in Rod's pages, so I will not use the 4th order LR LP filter with the later configuration.
Now, the sub will have its own enclosure, but I still don't know which to use, clsed or veted box, and the two 12" speaker + tweeters will be fitted in separate closed enclusures, each closed enclusure will contain a 12"speaker and a tweeter.
I was wondering if I use a 2nd order passive LR HP filter for the tweeters at 4Khz or 6Khz, because I am still afraid that tweeters may be at risk when I have the cutoff frequency of the LR HP active filters at 250Hz or 310Hz . Would that bring problems to my new Bi-amped system if I use passive filters with the LR active crossovers.
I already have 10s of amplifiers ready to pump, ranging from 20 watt to 200 watt, but I only want to apply Bi-amping to my new audio system, although I could use three-way active filters, but this will increase the wiring work, don't care about the cost though, I don't know, may be there is somebody who can bring a better idea justifying that a Tri-amped system is much better, and has more details in the sound.
Also I need a calculator software to make the 2nd order LR passive filter.
Another thing, I saw that Linkwitz uses delays in his crossover circuits, especially at the HP LR filters, shall I worry about that too, in case I go to Tri-amped system, and shall I worry about it if I am still going for Bi-amped system . Also can you guys suggest something practical about power distribution in both Bi-amped, and Tri-amped systems. I am still doubted about the sut-off frequency in deed, though I feel its ok, but deeply, I don't feel its ok.
Thanks in advance
I have been searching and simulating for a while on active filters, especailly those Linkwitz-riley and bessel, I was surfing in www.linkwitzlab.com and www.sound.au.com. I think its time to crystallize my thoughts now..
Any way, I have two 12" loudspeakers which are best dedicated for midrange audio spectrum, and two 4 Ohm tweeters, those types which have a coil, means no peizo, cuz I don't like the way it sounds. I will buy th subwoofer later on, while discussing this system on the forum..
If I want to build 4th order LR HP filters based on Rod's formulas, and use a cut-off frequency of 250Hz, or 310Hz. And build a 4th order LR LP filter which sums the L & R channels, and has variable gain circuit based on an opamp like that described in linkwitzlab, also will have a cut-off frequency which is the same as that of LR HP filter. I am aware of the fact that I will have to use Linkwitz transform circuit if I use a closed box for the sub, and may be the sub controller in Rod's pages, so I will not use the 4th order LR LP filter with the later configuration.
Now, the sub will have its own enclosure, but I still don't know which to use, clsed or veted box, and the two 12" speaker + tweeters will be fitted in separate closed enclusures, each closed enclusure will contain a 12"speaker and a tweeter.
I was wondering if I use a 2nd order passive LR HP filter for the tweeters at 4Khz or 6Khz, because I am still afraid that tweeters may be at risk when I have the cutoff frequency of the LR HP active filters at 250Hz or 310Hz . Would that bring problems to my new Bi-amped system if I use passive filters with the LR active crossovers.
I already have 10s of amplifiers ready to pump, ranging from 20 watt to 200 watt, but I only want to apply Bi-amping to my new audio system, although I could use three-way active filters, but this will increase the wiring work, don't care about the cost though, I don't know, may be there is somebody who can bring a better idea justifying that a Tri-amped system is much better, and has more details in the sound.
Also I need a calculator software to make the 2nd order LR passive filter.
Another thing, I saw that Linkwitz uses delays in his crossover circuits, especially at the HP LR filters, shall I worry about that too, in case I go to Tri-amped system, and shall I worry about it if I am still going for Bi-amped system . Also can you guys suggest something practical about power distribution in both Bi-amped, and Tri-amped systems. I am still doubted about the sut-off frequency in deed, though I feel its ok, but deeply, I don't feel its ok.
Thanks in advance
Hi
Have you considered buying a digital xo like the DCX2496 ?, then you can try out various orders, gain, delay and xo frequencies.
after trying out the best configration you can build a "supercircuit".
and if you dont want the DCX after that you can allways sell it (cheap)to me. I want one more
DVC Keld
Have you considered buying a digital xo like the DCX2496 ?, then you can try out various orders, gain, delay and xo frequencies.
after trying out the best configration you can build a "supercircuit".
and if you dont want the DCX after that you can allways sell it (cheap)to me. I want one more
DVC Keld
Hi
Simply NO, cuz I don't know where to get that digi thingy from, if you want more filters, I can sell them to you readily assembled on vero boards, but without opamps
I still have the option to experiment using a breadboaord...
Any way, I have tinned and green masked vero boards which are more useful and trustable than the ordinary non-tinned. They cost me 2.75$ each
I think I like this one , as it seems alot impressive to me in deed, but won't build them as I don't know how to make it two-way, also linkwitz seems to have optimized this three-way crossover for a specific speaker system.
I will post my schematics later
Simply NO, cuz I don't know where to get that digi thingy from, if you want more filters, I can sell them to you readily assembled on vero boards, but without opamps
I still have the option to experiment using a breadboaord...
Any way, I have tinned and green masked vero boards which are more useful and trustable than the ordinary non-tinned. They cost me 2.75$ each
I think I like this one , as it seems alot impressive to me in deed, but won't build them as I don't know how to make it two-way, also linkwitz seems to have optimized this three-way crossover for a specific speaker system.
I will post my schematics later
metal said:
They most certainly would be at risk without the passive crossover,
which should not affect the operation of the active crossover.
metal said:
http://www.mhsoft.nl/spk_calc.asp
Though reality is a little more complicated ........
Its highly unlikely a standard LR c/o will suit your drivers.
metal said:
More advanced c/o design that the above link takes into account
driver amplitude response, phase response, and mounting effects.
The power distribution depends somewhat on the relative efficiencies
of the sections you are driving, but for c/o points of 300 and 3kHz i'd
say ~50% bass, ~30% midrange, ~20% treble.
/Sreten.
Re: Re: Filter Questions/Suggestions for a Bi/Tri-amped System
Thats good news but...
Could you explain more please...I don't get it, is it better to go Tri-amped instead of Bi-amped and using 2nd or 4th order HP passive LR filter..also did you mean I have to make advanced measurements for the drivers to find out which c/o frequency to use...
I found that to be the case while I was reading the explainations about that crossover network, the guy did really hard work, then I will have to measure alot here and there...to find my speakers Fs, Qts, and Vas.......also may have to use TrueRta to measure responses, unless there is a better software
That seems reasonable and logical, thanks
sreten said:
Thats good news
but... sreten said:
Could you explain more please...I don't get it, is it better to go Tri-amped instead of Bi-amped and using 2nd or 4th order HP passive LR filter..also did you mean I have to make advanced measurements for the drivers to find out which c/o frequency to use...
I found that to be the case while I was reading the explainations about that crossover network, the guy did really hard work, then I will have to measure alot here and there...to find my speakers Fs, Qts, and Vas.......also may have to use TrueRta to measure responses, unless there is a better software
Originally posted by sreten
The power distribution depends somewhat on the relative efficiencies
of the sections you are driving, but for c/o points of 300 and 3kHz i'd
say ~50% bass, ~30% midrange, ~20% treble.
That seems reasonable and logical, thanks
Re: Re: Re: Filter Questions/Suggestions for a Bi/Tri-amped System
Hi,
hard to say which is better - depends on circumstances.
Regarding c/o design - say your target is 4th order L/R
acoustic response for the mid/treble crossover.
This cannot be implemented by an off the shelf c/o, to achieve
this design goal the c/o must be exactly matched to the drivers
when they are mounted in the final enclosure - not easy.
/Sreten.
metal said:
Hi,
hard to say which is better - depends on circumstances.
Regarding c/o design - say your target is 4th order L/R
acoustic response for the mid/treble crossover.
This cannot be implemented by an off the shelf c/o, to achieve
this design goal the c/o must be exactly matched to the drivers
when they are mounted in the final enclosure - not easy.
/Sreten.
I'd also add that a summed sub channel going as high as 300Hz is going to not be good for the stereo image. You want a summed channel to be 100Hz or lower really. If you must cross that high then use stereo subs.
I also question why you are trying to use 12 inch speakers as mids, but also crossing them far too high for high fidelity. Are you building a PA system?
I also question why you are trying to use 12 inch speakers as mids, but also crossing them far too high for high fidelity. Are you building a PA system?
Hello All
In deed, I was amazed this morning, I used STK4191V @-/+ 36 Volts, and wanted to test the two 12" speakers to see where to choose the cut-off frequency. I used TrueRTA, I started from the lower end @ 20Hz, and put a 1K resistor in series with one of the speakers, and started measuring the voltage drop across that resistor to see where I get the highest voltage drop at the first lowest frequecy while sweeping, I got that at 565Hz, but I still can find that there are dips and peaks in the speaker response at many different frequencies, any way, that will not work.
Thanks to David B. Weems book "Great Sound Stero Speakers", I saw that he explained how to measure Fs, Qtc, and Vas of a speaker, I will apply these tests to the two 12" speakers, and start designing a closed box for them, may be not !!
I also read in another book that its best to have two bass woofers, and be able to sum their inputs to have a mono bass output which means you can have stereo or mono bass output, thats because if you are watching a movie, then you switch to mono, saying that the results are much more thundering rather than a stereo when listening to horses or gun shots, while for music, you beter switch it to stereo. And as richie said, that filter was far below 100Hz, it was tuned at 70Hz, thats correct richie, thank you
Any way, returning to the 12" speakers, I removed the 1K resistor, and started listening, what the hell I was trying to do man, it seems If I use these speakers, I will end up with a PA system, the sound was too loud, and I had to close my ears, becasue I could not stand both the High-end bass, neither the mids
I concluded that I have high SPL speakers, any way, that STK was barely hot, even whan I drove the speakers very loud, till I really could not stand it, and got out of the room to listen.
In deed I am almost lost now, I think I will have to go for 8/10" speakers, put them in a closed box, and use 12" subwoofer @ cut-off frequency @ 100Hz as richie said, as I don't need a PA system in my room, so at last, I will be kicked out, as was gonna happen today, when police came asking who is the hell was playing too loud, neighbours were just sick of it
Also I will use STK4191V for the sub, bring two other STk4161V/4141V for the speakers and tweeters respectivelty. I don't know, may be you guys have some thing else to say
That seems to demand lots work and measurement, but does it really worth the try Streten, I don't know, it seems I am too lazy to do that
In deed, I was amazed this morning, I used STK4191V @-/+ 36 Volts, and wanted to test the two 12" speakers to see where to choose the cut-off frequency. I used TrueRTA, I started from the lower end @ 20Hz, and put a 1K resistor in series with one of the speakers, and started measuring the voltage drop across that resistor to see where I get the highest voltage drop at the first lowest frequecy while sweeping, I got that at 565Hz, but I still can find that there are dips and peaks in the speaker response at many different frequencies, any way, that will not work.
Thanks to David B. Weems book "Great Sound Stero Speakers", I saw that he explained how to measure Fs, Qtc, and Vas of a speaker, I will apply these tests to the two 12" speakers, and start designing a closed box for them, may be not !!
I also read in another book that its best to have two bass woofers, and be able to sum their inputs to have a mono bass output which means you can have stereo or mono bass output, thats because if you are watching a movie, then you switch to mono, saying that the results are much more thundering rather than a stereo when listening to horses or gun shots, while for music, you beter switch it to stereo. And as richie said, that filter was far below 100Hz, it was tuned at 70Hz, thats correct richie, thank you
Any way, returning to the 12" speakers, I removed the 1K resistor, and started listening, what the hell I was trying to do man, it seems If I use these speakers, I will end up with a PA system, the sound was too loud, and I had to close my ears, becasue I could not stand both the High-end bass, neither the mids
I concluded that I have high SPL speakers, any way, that STK was barely hot, even whan I drove the speakers very loud, till I really could not stand it, and got out of the room to listen.In deed I am almost lost now, I think I will have to go for 8/10" speakers, put them in a closed box, and use 12" subwoofer @ cut-off frequency @ 100Hz as richie said, as I don't need a PA system in my room, so at last, I will be kicked out, as was gonna happen today, when police came asking who is the hell was playing too loud, neighbours were just sick of it
Also I will use STK4191V for the sub, bring two other STk4161V/4141V for the speakers and tweeters respectivelty. I don't know, may be you guys have some thing else to say That seems to demand lots work and measurement, but does it really worth the try Streten, I don't know, it seems I am too lazy to do that
Hi,
I agree with the other poster.
That gap between the tweeter and 10inch mid needs to be filled in with a more appropriate mid. Maybe in the range 4inch to 6inch.
This then makes it a 4 way system, allowing you to dedicate the sub to real sub duty say up to about 120Hz.
Can I suggest you build the system as active initially using a DCX2496 to select all the parameters, then sell it on and use the data to generate the active or passive crossovers.
If you go the passive route then also consider using an amplifier dedicated to each driver and its crossover section. (bi-amping or tri-amping in the UK)
I agree with the other poster.
That gap between the tweeter and 10inch mid needs to be filled in with a more appropriate mid. Maybe in the range 4inch to 6inch.
This then makes it a 4 way system, allowing you to dedicate the sub to real sub duty say up to about 120Hz.
Can I suggest you build the system as active initially using a DCX2496 to select all the parameters, then sell it on and use the data to generate the active or passive crossovers.
If you go the passive route then also consider using an amplifier dedicated to each driver and its crossover section. (bi-amping or tri-amping in the UK)
Lots of good advice here already.
I'd like to offer some comments about power distribution across the audio frequency spectrum.
I did some tests on the spectra of popular music, including rock, pop, electronic, dance music. I chose these styles and omitted classical and jazz as I tend to listen to the popular styles at higher volumes and the purpose of the tests were to determine the relative maximum SPL's required across the frequency spectrum.
The results showed an almost exact -1.5dB/octave drop as frequency increased from about 45Hz to 10kHz. In the octave below 60Hz the amount of power dropped 12dB relative to the octave above it. In the octave above 10kHz the power dropped by 10dB relative to the octave below it.
The drop in the top octave may be due to the nature of the source material which was all MP3 encoded.
So by considering your crossover frequencies and driver sensitivities you will have a very good estimate of your power requirements for a given SPL.
The only caveat might be the power required in the subwoofer region for use with movie soundtracks. These sometimes have a lot of low frequency energy. For movie use, a safe estimate might add 12dB to the power required for popular music.
Cheers,
Ralph.
I'd like to offer some comments about power distribution across the audio frequency spectrum.
I did some tests on the spectra of popular music, including rock, pop, electronic, dance music. I chose these styles and omitted classical and jazz as I tend to listen to the popular styles at higher volumes and the purpose of the tests were to determine the relative maximum SPL's required across the frequency spectrum.
The results showed an almost exact -1.5dB/octave drop as frequency increased from about 45Hz to 10kHz. In the octave below 60Hz the amount of power dropped 12dB relative to the octave above it. In the octave above 10kHz the power dropped by 10dB relative to the octave below it.
The drop in the top octave may be due to the nature of the source material which was all MP3 encoded.
So by considering your crossover frequencies and driver sensitivities you will have a very good estimate of your power requirements for a given SPL.
The only caveat might be the power required in the subwoofer region for use with movie soundtracks. These sometimes have a lot of low frequency energy. For movie use, a safe estimate might add 12dB to the power required for popular music.
Cheers,
Ralph.
Hi all,
my views on power distribution vary somewhat from the concensus posted here.
Power or energy distribution do not enter the equation to a significant degree.
My first assumption is driver of equal voltage sensitivity (db output for 2.8V input)
All the drivers need a similar peak voltage drive.
the treble amp needs the same voltage rails as the mid amp and as the bass amp.
This has the effect that all the amps should deliver the same power into the same impedance.
However the treble driver draws much less average current than the lower frequency drivers resulting in an amp design that concentrates on quality of output rather that quantity. This in part is due to the relatively short signal peaks that the amp has to amplify. From the other extreme the bass amp will be asked to deliver high currents for sustained periods that will be many times longer than the higher frequency drivers. The PSU and output stage should be designed to deliver this energy without compromise.
The big difference comes when my basic assumption no longer holds true. Drivers of different sensitivity can and should modify the amplifier requirements. The advantage that can be gained by adjusting the relative power outputs to account for sensitivity variations can result in significant economy and also improved performance.
In conclusion, forget energy distribution and instead use sensitivity variation to guide your design power variation to the dedicated amps. eg. using a baseline 92db mid driver. 96db treble needs 63% of voltage = 40% power. 86db bass needs 200% of voltage = 400% power. Using real numbers could give bass = 200W, mid = 50W, treble = 20W ( might suit ClassA) .
my views on power distribution vary somewhat from the concensus posted here.
Power or energy distribution do not enter the equation to a significant degree.
My first assumption is driver of equal voltage sensitivity (db output for 2.8V input)
All the drivers need a similar peak voltage drive.
the treble amp needs the same voltage rails as the mid amp and as the bass amp.
This has the effect that all the amps should deliver the same power into the same impedance.
However the treble driver draws much less average current than the lower frequency drivers resulting in an amp design that concentrates on quality of output rather that quantity. This in part is due to the relatively short signal peaks that the amp has to amplify. From the other extreme the bass amp will be asked to deliver high currents for sustained periods that will be many times longer than the higher frequency drivers. The PSU and output stage should be designed to deliver this energy without compromise.
The big difference comes when my basic assumption no longer holds true. Drivers of different sensitivity can and should modify the amplifier requirements. The advantage that can be gained by adjusting the relative power outputs to account for sensitivity variations can result in significant economy and also improved performance.
In conclusion, forget energy distribution and instead use sensitivity variation to guide your design power variation to the dedicated amps. eg. using a baseline 92db mid driver. 96db treble needs 63% of voltage = 40% power. 86db bass needs 200% of voltage = 400% power. Using real numbers could give bass = 200W, mid = 50W, treble = 20W ( might suit ClassA) .
Hello All
Thanks for all the advices here, now I think I have a good idea about power distribution now, thanks
I was trying to measure TSP & impedance curve for some speakers using audio tester software, man, this seems to yield strangs readings, although I connected the speaker as it should be, but, no results, it just yields monterious graphs, which has nothing to do with mpedance curve, also it doesn't give me any TSP at all. Can you experst explain the procedures, also how to know what each curve yields and how to make each curve yield what I want.
Thanks
Thanks for all the advices here, now I think I have a good idea about power distribution now, thanks
I was trying to measure TSP & impedance curve for some speakers using audio tester software, man, this seems to yield strangs readings, although I connected the speaker as it should be, but, no results, it just yields monterious graphs, which has nothing to do with mpedance curve, also it doesn't give me any TSP at all. Can you experst explain the procedures, also how to know what each curve yields and how to make each curve yield what I want.
Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.