I'd like to design a first order three way line level passive crossover.
I know how to get the calcs and make such a thing at the speaker level, but at the line level I don't know where to start.
A note - it doesn't actually have to be passive, but I want to avoid op amps. It can incorporated into the build of a Chinese knock-off NAC42.5 if that helps.
I know how to get the calcs and make such a thing at the speaker level, but at the line level I don't know where to start.
A note - it doesn't actually have to be passive, but I want to avoid op amps. It can incorporated into the build of a Chinese knock-off NAC42.5 if that helps.
First order won't give much separation between tweeter and bass unit - fourth order is commonly used for line-level crossovers as its cheap to do (and avoids the losses of passive crossovers at high power - which are often 2nd order as a compromise). First order is just RC stages of course, f = 1/(2πRC) - all you need to know.
A 3-way crossover is just a cascade of two 2-way crossovers - ideally with buffer interposed to give consistent source impedance
to the second crossover.
Well you can run standard Sallen Key filter topology with emitter followers, but honestly avoiding opamps is illogical - they do the job and they do it well.
A 3-way crossover is just a cascade of two 2-way crossovers - ideally with buffer interposed to give consistent source impedance
to the second crossover.
If you really want a passive XO then you need to know the output impedance of the pre-amp and the input impedance of the power amp. Then you know what value of RC filter you can hang on the pre-amp and what value the power amp input will not swamp. If you were looking for more than 1st order, each stage should be 10x the impedance of the previous one to avoid loading. Inductors at high impedance is probably not practical.
But like Mark saz, op-amps work well, and you can run them from a wall-wart PS. As long as you don't want an adjustable frequency, it's easy. If you want an adjustable frequency then maybe you want a digital system.
But like Mark saz, op-amps work well, and you can run them from a wall-wart PS. As long as you don't want an adjustable frequency, it's easy. If you want an adjustable frequency then maybe you want a digital system.
I do know the input and output impedances of the pre (600 ohms) and power (10K).
Spice - as soon as I finish posting this I'll be working through a KiCad / Spice tutorial (now I have a design to play with - keeping me up late at nights for a week). I really like the idea of Spice, if it had existed in 1979 I'd have carried in with electronics at school!
First or higher order - I loved the Eposes but I'm scared of the phase shift. I'll go second.
Active v Passive - I come from a Linn / Naim philosophy where you keep anything you can out of the signal path. But Active is a lot easier! So I'm building a modular system with Active plug in cards that can be removed and replaced with Passive ones. I'll start with Active. It's a lot easier to get technically correct. Once I'm happy, I'll see if I can get rid of the op amps.
Finally, I've ordered a three way fourth order active crossover from China. I understand what I'm doing just about well enough to switch the components around. I'll try it against my second order Passive and Actives.
Wall wart PS - er, next job is to build a regulated +/- 15v supply. We do things properly around here!
Thanks all - having lots of fun but not much sleep. On the other hand, thanks to modern simulation software, my spend is 0.00 so far.
Spice - as soon as I finish posting this I'll be working through a KiCad / Spice tutorial (now I have a design to play with - keeping me up late at nights for a week). I really like the idea of Spice, if it had existed in 1979 I'd have carried in with electronics at school!
First or higher order - I loved the Eposes but I'm scared of the phase shift. I'll go second.
Active v Passive - I come from a Linn / Naim philosophy where you keep anything you can out of the signal path. But Active is a lot easier! So I'm building a modular system with Active plug in cards that can be removed and replaced with Passive ones. I'll start with Active. It's a lot easier to get technically correct. Once I'm happy, I'll see if I can get rid of the op amps.
Finally, I've ordered a three way fourth order active crossover from China. I understand what I'm doing just about well enough to switch the components around. I'll try it against my second order Passive and Actives.
Wall wart PS - er, next job is to build a regulated +/- 15v supply. We do things properly around here!
Thanks all - having lots of fun but not much sleep. On the other hand, thanks to modern simulation software, my spend is 0.00 so far.
It did exist, it was first released under that name in May 1972, written in Fortran and needing a mainframe computer to run 🙂
Isn't such a kind of philosophy tantamount to wishful thinking (it sounds great, little to do with reality though) - minimizing error is not about minimizing complexity or BoM, its about minimizing error.
So use techniques and components that reduce error (and hopefully cost too). A good design gets worse if any component is removed, showing it was doing something useful. Just because a simple design performs well doesn't mean a more complex design can't perform way better.