crossover active

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ciao.
Spero capirete che in Italia può capitare che gli articoli di elektor siano stampati con errori dovuti alla stupidità delle persone.
Comunque cerco la tabella 4 che viene nominata nell’articolo, la quale indubbiamente dovrebbe contenere le formule per il calcolo dei filtri, alle varie frequenze di taglio
Allego il file con il progetto di elektor, affinché sia chiaro cosa cerco.
Qualcuno ha la versione completa?
Grazie in anticipo a chi mi aiuterà
Amo LME49830.
Guido
Dimenticavo! FRD ci ha regalato degli ottimi programmi per il calcolo dei crossover, quello in cui si calcolava un quattro vie non riesco più a trovarlo qualcuno lo ha e me lo può dare? Era in formato. xls


Hello.
I hope you will understand that in Italy it can happen that the elektor items are printed with errors due to the stupidity of the people.
However I am seeking the table 4 that is mentioned in the article, which undoubtedly should contain the formulas for the calculation of the filters, the various cut-off frequencies
I attach the file with the project elektor, to make it clear what I'm looking.
Does anyone have the full version?
Thanks in advance to those who will help me
I love LME49830.
Guido’


I forgot! FRD has given us some good programs for the calculation of crossovers, one in which it was estimated a four-way I can not find someone has it and can give it to me? He was formed. xls Visualizza traduzione originale
 
Hi Guys

Personally I find it easiest to use the Sallen-Key filters with Riley-Linkwitz alignment. All this entails is adding the factor of root-2 into the star\dard equation, which means you are calculating a 6dB-down point for the crossover frequency.

The preferred alignments are 12dB/oct and 24dB/oct, with the latter preferred overall.

Low-pass is then two cascaded LP filters.
HP is two HP filters cascaded.
Middle bands are two LPs cascaded with two HPs to make a bandpass.

It is best to keep the crossover simple inasmuch as it just provides the crossover filtering desired. if you have a "problem" driver that needs an extra filter effect, add it to that band only and make this filter switchable so the crossover can be used with other drivers later.

It is of no value whatsoever for a DIY or purpose-built crossover to have variable frequencies. This just entails unnecessary expense and introduces opportunities for miswiring. It is good though to provide an output level for each band if you need to tweak the relative loudness of the drivers.

Bi-amping rules! It reduces system distortion by an order of magnitude.

A sub-sat system in the home provides much better movie sound than the artificial tone from a "home theatre" system AND provides far better music reproduction.

Have fun
 
Last edited:
I can't read that Italian article, but I think it is describing a subtractive filter arrangement.
I suggest you look at the simpler lowpass and high pass filters.
ESP has a couple of articles and a calculator for the LR filters.
Linkwitz Riley as described in the Linkwitz site is more comprehensive.
Sreten is correct that it's the total filter of electrical plus acoustic that needs to be your target. But Struth's very much simpler LR4 suggestion to get you started works quite well in some circumstances.

Use a good driver that performs well in it's passband and filter out the signals that make it misbehave and you will find that the steeper LR4 allows a good sound from good drivers very simply.
It's worth trying as a start. You may be happy with what you get.
 
Ciao Guido. :)

Well, I can read Italian, so will comment on the actual article :)

1) it´s an "Elektor Project", with all that implies :rolleyes:
Not dissing them, quite the contrary, Elektor is THE DIY magazine, so at every issue, no matter what, they MUST offer at least a couple new , buildable projects, not an easy task every Month, every year.

And those must be "serious" projects, (not LED blinkers or LM386 amps), of good quality and fully developed and tested,including a proper PCB.
All that makes me admire them immensely because that´s a Herculean task :D

2) if anything, they tend to overcomplicate things, although in a way it´s understandable, I guess they want to stay ahead of the pack ........ which they usually do.

3) Guido !!!! : it´s a 1989 project !!!!!! :confused:
Doubt many will have it available here, but I know old Elektor editions have been scanned and are available as PDF on CD ROM so you might get the 1989 one, just search around .

4) recalculating everything will be a PITA and probably prone to errors BUT: the schematic and article provides values for 500 and 5000 Hz , other frequencies were provided for in the (lost?) table, BUT you can easily convert/transpose values to other frequencies by increasing/decreasing relevant cap values by a fixed factor.
Say, you want to move low/mid crossover, now 500Hz , to, say, 600 Hz, you decrease low band lowpass caps and midband highpass ones by a factor of 1.2 and you have the new ones.

Same with mid to high crossover.

Enjoy :D
 
Last edited:
I can't read that Italian article, but I think it is describing a subtractive filter arrangement.
I suggest you look at the simpler lowpass and high pass filters.
ESP has a couple of articles and a calculator for the LR filters.
Linkwitz Riley as described in the Linkwitz site is more comprehensive.
Sreten is correct that it's the total filter of electrical plus acoustic that needs to be your target. But Struth's very much simpler LR4 suggestion to get you started works quite well in some circumstances.

Use a good driver that performs well in it's passband and filter out the signals that make it misbehave and you will find that the steeper LR4 allows a good sound from good drivers very simply.
It's worth trying as a start. You may be happy with what you get.

ho seguito i consigli grazie dei suggerimenti.
ciao guido
 
The one and only
Joined 2001
Paid Member
The key problem is that the standard sorts of filters are not tolerant of the
amplitude and phase errors of the loudspeaker drivers, and the summed results
for any driver whose bandwidth doesn't exceed the filter by a couple of octaves
are not usually good.

Siegfried Linkwitz dealt with it by equalizing and phase-shifting the driver as
part of getting a good response, although that is more complicated and tedious.

A simpler approach is to individually adjust the number and frequency of
the poles of the filter until you get a decent sum, something that you don't
find in fixed-formula filter sets. It still can be time consuming, but at least
it doesn't require additional circuitry.
 
More explanation here:
3-Way Active Crossover with Linearity Phase eng[
dale + wima good,good,good,good,good,good,good

Explanations fancy.
Once again, this is not a linear phase crossover, just a LR4, spice it, you wil see.
Linear phase subtractive xo needs a time delay path, hard to get acourate
with analog circuitery, see Douglas Self / the design of active crossovers /
Ch 10.8 : delay lines for subtractive crossovers. The Design of Active Crossovers by Douglas Self
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.