Can you use a digital active crossover to design a passive analog crossover ?

[GoatGuy]

Beware "stronger slopes" - the balance between "paper perfect" and "sounds like sh*t" is exemplified by filter-roll-off design. Theoretically you can get -100 dB/octave ... but the group-phase-shift of frequencies passing thru the crossover is so ridiculous that the sound is like ... well, its terrible.

Adopt the "soft, gentle, 'just enough'" approach. If 12 dB/octave does the job (including the natural 6-12 db/octave of the cones), well, let it be. Musicality is on many levels about showing evidence that you worked really hard (i.e. "building a story"), and used many subtle sophisticated things (i.e. embellishing the story) ... but in practice barely deviating from natural resonance mitigation, natural system response, and barely fed-back amplification.

That's where the gold is.

And... just again: PLEASE consider making "rolling your own inductors" a hobby. Its easy, its cheap (in the end), and you can make an endless succession of inductors that quite simply you wouldn't ever be able to buy. Its the last frontier of true DIY audio.

GoatGuy

[RickDangerous]

Originally Posted by GoatGuy

Beware "stronger slopes" - the balance between "paper perfect" and "sounds like sh*t" is exemplified by filter-roll-off design. Theoretically you can get -100 dB/octave ... but the group-phase-shift of frequencies passing thru the crossover is so ridiculous that the sound is like ... well, its terrible.

Adopt the "soft, gentle, 'just enough'" approach. If 12 dB/octave does the job (including the natural 6-12 db/octave of the cones), well, let it be. Musicality is on many levels about showing evidence that you worked really hard (i.e. "building a story"), and used many subtle sophisticated things (i.e. embellishing the story) ... but in practice barely deviating from natural resonance mitigation, natural system response, and barely fed-back amplification.

That's where the gold is.

And... just again: PLEASE consider making "rolling your own inductors" a hobby. Its easy, its cheap (in the end), and you can make an endless succession of inductors that quite simply you wouldn't ever be able to buy. Its the last frontier of true DIY audio.

GoatGuy

Ok, noted. I guess i have to dip those rolls of wire in varnish to keep them vibrating, right?

My speakers (the ones i actually use, anyway) have 4th order 24dB slope filters and sound pretty darn good to me.
JBL :: Product
Given that they cost 400$ a pair with 2 Samson stands & 2 XLR cables, i'd say the components inside were probably no more than 50$ per speaker, if that.
Are those filters really so expensive to buy in the DIY world ?

[picowallspeaker]

There's a thread going on from 250 pages about Behringer b2031 Vs.Orion ...
And many others which talk about these cheap studio monitors
...biamplifed, similar to yours.
If you could only put your hands on the Dared ...

Also, there are numerous talks about passive Vs. Active, semplicity & minimalism & tubes & ICs Vs. Monitor & broad/wide & nearfield Vs Stage & pics & SS

[GoatGuy]

Originally Posted by RickDangerous

Ok, noted. I guess i have to dip those rolls of wire in varnish to keep them vibrating, right?

Could... but if you are using separate-bobbin winding, you'll be able to achieve a quite-tight wrap of the wire itself. Using the forever-stick tape (usually a kind of cloth-impreg-with-plastic stuff), you "finish it off" with the final wrap. In the way-old days, they used to take the transformers and coils, stick them in a vacuum-bucket, pull a strong vacuum, then after an hour or so, dump in varnish, then let the air back in. Atmospheric pressure would ensure varnish would get into every nook and cranny. This still is done for some magnetics such as guitar-pickups. But really, its not necessary nominally. A good tape wrap with the right tape, does it. Not too surprisingly, even bog-standard (but high grade) duct tape works well, since if it really is "duct" and not "duck" tape, it is design to hold for decades.

My speakers ... have 4th order 24dB slope filters and sound pretty darn good to me.

Glad to hear it. 4th order isn't beyond reasonable. Just don't start thinking that if "4th order is good", then 8th order will be better.

Given that they cost 400$ a pair with 2 Samson stands & 2 XLR cables

Hold on there Hoss... XLR cables? these sound like active speakers to me, not passives. [I go looking at the link] Ah, bi-amplified JBL speakers. Well, I'm sure they're pretty nice. Remember though they do the Linkwitz crossover split on the INPUT side, not after it has been amplified. That's why they're "bi-amped". The Linkwitz-Reilly filter can be built out of inexpensive op-amps and very low power (inexpensive) capacitors and resistors. Quite a bit different from building a crossover inside a speaker, which is in charge of splitting and routing many watts of audio power to the cones. Quite a bit.

NO - they don't have to cost a fortune, your cross-overs. You'll only be limited though by your lust for the "perfect components". The more you read here, the more you'll see heated discussions and ginormous opinions regarding how good a particular $50 capacitor (quantity 1!!!) is over some other $75 capacitor.

Anyway... check this out - a very straight forward 3-way, with all the components listed. Note there are a lot of "air core" inductors. They're even easier to build... no core! Just bobbins, wire, and a number-of-turns counter chuck for your electric drill. And formulas. Make 'em yourself. Now go forth, and PRICE OUT all the components - remembering the power dissipation levels for the resistors!

http://electroschematics.blogspot.com/2012/04/3-way-crossover-speaker-circuit.html

GoatGuy

[GoatGuy]

Try...

3-Way Crossover Speaker Circuit | Electro Schematics

Sorry, bad link code.

[RickDangerous]

Originally Posted by GoatGuy

Try...

3-Way Crossover Speaker Circuit | Electro Schematics

Sorry, bad link code.

Thanks for that.
Looking at the schematic, i dont understand why there is an L-pad on all three drivers, i thought that was to compensate for varying sensitivity. In this case why not just have a -4 dB pad on the mids ?

[RickDangerous]

Also, im guessing that, even with used equipment, the cost of a decent digital crossover plus X amps kind of makes up for the savings on the filter, no?
Plus having a crossover in the speaker is far more convenient.

[RickDangerous]

Originally Posted by GoatGuy

Also, one last thought: You might want to settle on just "bi-amped", but go ahead and have 3 cones per speaker.
(...)
GoatGuy

I like the idea but i'm not sure i get it.
How do you do that? Do you sum the signal coming from the tweeter & mid before you amplify it ?

I am currently looking into the Gainclone as this seems like the best value cheap & simple & small amp there is...or that i can find so far. Maybe tri-amping with these Gainclone thingies won't be so expensive after all.
And maybe i can get them to fit in each speaker so that i don't have to bother with speaker cable as well. My cabinets will probably be quite big/deep so hopefully they accomodate these amps.

Thanks again & sorry for the rookie questions, i feel like i've landed on a different planet.

[head_unit]

Originally Posted by RickDangerous

tri-amping

If you can afford the amps-which are perhaps LESS in total time and money-I'd do that.

Because to get back to you original question: " Can you use a digital active crossover to design a passive analog crossover" the answer in my opinion is "kinda sorta not really"
- First off, digital crossovers have different characteristics than are even possible with passives. Unless you get the IIR or is it FIR, I forget. One of those can have phase and magnitude characteristics like theoretical passive crossovers IF you can get the right digital crossover and set it correctly. I italicize "theoretical" because feeding real speaker drivers, the passive will not act the same so that is another problem. Hence the suggestion above about passive filters feeding the amplifiers.

Passive crossover design is a real rabbit hole. Microphone, measurement setup, software to capture magnitude AND phase of not just frequency response but impedance, software to hard-number-crunch the voltage transfer of arbitrary passive circuits (LEAP was great but founder Chris Strahm died and I don't know what is usable in its place now). And MUCH time to spend. It is simpler-I don't say simple ha ha but simplER-to adjust active crossovers.