passiveX-over to allow multiband mix-bus compression . .

Dear all,

I used to use an ATC active crossover to split my mix into three bands, and often used the optical limiting on each band. This was years before TC finalyser, Focusrite, or Tubetech, - but I won't any credit for the concept. That was a suggestion of the marvelous Roger Quested. (to me ,anyway!) (We used to use the ATC when I first worked with him). I would say that the results were a bit crude, but then again, so were my mixes back in those days! I have often wanted to build a crossover enable me to split my signal into 2 or 3 different stereo compressors. I have details of active crossovers, but what I was wondering was would this work with a 12dB/octave x-over, or should I use 24dB/Octave? If 12dB/8ve is ok, could this be achieved passively with LC network? How would I go about calculating the values? I have tried to find some documentation on the web, but to no avail so far. I was also thinking of using transformers to recombine a 2-way split-signal.

- Or should I just buy that £68 Berhing@@@er, 2 way, active, stereo, 24dB/8ve, Linkwitz-Riley box?

Please, tell me that there is another way!


I am forever in your debt, guys! - Andy P

Here's a convenient link http://ccs.exl.info/calc_cr.html with calculators for several different types of passive crossovers.

However, you'll quickly notice that when you plug in reasonable line level impedances the inductor values get quite large. This complicates things dramatically. Better to stick with an active design.

Thomas

dear Barefoot,

Thanks for your reply. I have just found a similar calculator, and found the same thing. When you say that this complicates things, in what way exactly? there are plenty of larger inductors. that can't be the problem . . . are there other criteria coming into play with larger inductors? If so, what exactly?
easy enough to build something active I reckon, but there is something magical in "passive" to me . . . .

Andy P

Good to know, Jaako!

But I want to be able to use ANY stereo compressor in the chain ,not just Forssell Opto or SSL, but you certainly know where I'm coming from!

ALso, Barefoot, If I use Forssell/API/Hardy Op-Amps to drive Passive network(ie potentially as low as a 75ohm load!), are the values of inductor still too high?

Thanks, Guys, You are the best! All my questions answered! (You are all so tolerant . . . . )

Andy P

Hi,

Just done some calculations.

Linkwitz-Riley 12dB/8ve at 250Hz (I guess a good starting point based on my experience with TC Finaliser) using a 75 Ohm load, C=4.2uF/L=95.5mH. Now Sowter have two multi-tapped inductors, 200/160/100/80/50mH + 150/82/68/47/33/27 mH, which would allow for variable crossover frequencies :grin:

Would this be Possible?

Andy P

[quote author="strangeandbouncy"]ALso, Barefoot, If I use Forssell/API/Hardy Op-Amps to drive Passive network(ie potentially as low as a 75ohm load!), are the values of inductor still too high?[/quote]

The series resistance of most large value inductors is always going to be fairly high relative to the load, creating significant insertion loss. It will also require different inductor and capacitor values than modeled in these simple calculators. And if you're going to go to the trouble of building an active buffer, why not just make the whole circuit active? It will be much higher performance. Large inductors also have series capacitances and other nonlinearities similar to transformers. Are you looking for transparency or hoping to get a "sound"?

Dear Barefoot,


I guess I'll be going active then!


Thanks, Andy P

I don't think passive crossovers are at all bad. The inductors used for common 12dB slopes aren't at all large and you can wind them with fat wire for low DCR. You don't need a power supply or a bunch of parts, opamps, transistors, tubes, etc, just a couple of inductors and capacitors.

Places that support speaker building (parst express for one) usually have decent ranges of inductors and capacitors to knock up a few crossovers for fairly cheep. You could easily put together a decent three way passive crossover for less than an active crossover. It may not be as prescise as an active unit because tollerances of these passive components are usually not better than 10% but the sound quality is quite good.

[quote author="thomasholley"]I don't think passive crossovers are at all bad. The inductors used for common 12dB slopes aren't at all large and you can wind them with fat wire for low DCR. You don't need a power supply or a bunch of parts, opamps, transistors, tubes, etc, just a couple of inductors and capacitors. [/quote]
Unless I misunderstood, Andy is talking about using a line level crossover in front of other line level devices, not speakers. In this case active is definitely the way to go, unless maybe you're shooting for a vintage sound or something.

Dear Barefoot,

i was definately refering to a line-level X-over. I am no stranger to passive X-overs in monitors, having worked for Roger Quested in the late '80's, and subsequently built and sold my own passive 8" nearfields comercially in the early '90s. I am well aware of the vagaries of Inductors, especially when attatched to reactive loads(ie Loudspeakers!) I was guessing that if the load was much more constant, ie 75ohm or 150ohm resistor, that design would be simple, and effective, asuming that componant tolerances are reasonably tight. Passive seems sensible to me, but I am very able to take practical from anyone more experienced than I am! It is just as easy for me to build something active. I am looking for something as transparent as possible, and my previous experience of passive eq's led me to believe that a passive cross-over might be transparent too! I May well come back to the passive idea, but right now I need something effective and quickly to help me in da studio . . .


Thank you all for your input!


Andy P

Dear Barefoot,

i was definately refering to a line-level X-over. I am no stranger to passive X-overs in monitors, having worked for Roger Quested in the late '80's, and subsequently built and sold my own passive 8" nearfields comercially in the early '90s. I am well aware of the vagaries of Inductors, especially when attatched to reactive loads(ie Loudspeakers!) I was guessing that if the load was much more constant, ie 75ohm or 150ohm resistor, that design would be simple, and effective, asuming that componant tolerances are reasonably tight. Passive seems sensible to me, but I am very able to take practical from anyone more experienced than I am! It is just as easy for me to build something active. I am looking for something as transparent as possible, and my previous experience of passive eq's led me to believe that a passive cross-over might be transparent too! I May well come back to the passive idea, but right now I need something effective and quickly to help me in da studio . . .


Thank you all for your input!


Andy P