How to calculate crossover frequency?

770Hardtop

New Member
Is there a tool on the net that can calculate a speaker's crossover frequency if i input its capacitor and inductor values? I have seen a tool that does the reverse - if you put in the crossover frequency you want and driver impedances you have, it will calculate the values and give a schematic also (eg: http://www.diyaudioandvideo.com/Calculator/XOver/)

But i have some speakers that i want to modify, and i want to know the crossover frequency of this existing crossover to see if it is suitable.
 
I have been looking for these sort of reverse calculators for boxes and crossovers and the like all I have done is to start reverse engineering.
 
The thing you'd have to know is the impedance for which the crossover was designed. One can assume 8ohm or 4ohm in some cases, as some designers don't go beyond these nominal figure, but it's usually more complicated than this. You can find the electrical kneepoints and Q for 2nd order networks this way:

(1/Z)/(2*Pi*C) = f1

Z/(2*Pi*L) = f2

where: Z = driver impedance, C = capacitor in Farads, L = inductor in Henrys

then:

sqrt(f1*f2) = F

where: F = crossover frequency

then:

F/(highest of f1 or f2) = "Q factor", typical figures range from 0.55 to 1.0

Remember that these are the electrical figures only. What usually counts at the end is the acoustical kneepoints and slopes, something you really can not know without more involved measurements.

az
 
OK thank you for the formulae! I have no idea what the Q factor means to crossover design although i can see how you get it. And don't know what the knee point refers to. But i will input some values from my current crossover and see what I come up with.

I like the reverse calculation method too, using that calculator i gave in the link above, just keep on inputting a range of frequency values until i get very close to the component value. I'm expecting about 5500hz although would prefer lower to save myself some soldering. Thanks

But if anyone knows of an online calculator, please advise.
 
you can just keep trying different frequencies until the values of the caps/inductors match.

The problem is that while you may have an '8 ohm driver', it's impedance at the crossover frequency may be quite different. Also, speaker designers may tweak values based upon listening tests that don't align strictly according to the math. So it's a bit of an inexact science.
 
The kneepoint is another term for the cutoff, or -3dB frequency, or half-power point, the point after which the response is said to roll-off. The Q, or Quality factor, describes the filter's behavior at the aforementionned kneepoint. A low value of 0.56 (Bessel) will have a shallow roll-off and good impulse response, 0.7 (Butterworth) will have steeper roll-off and be maximally flat within the passband, this is a very common value to design with off the bat, and 1.0 (Chebyschev) will have the steepest roll-off, but will start getting peaky with somewhat worse impulse response. Values outside this range are seldom practical.

az
 
that's great, thank you. I have a lot more to work with now, and have a much better understanding of the type of crossovers. The only thing i'm not sure of is how phase shift is corrected, with the various Xover designs, if it even needs to be? And how does this affect the sound?

The link such as musichal gave is one i have seen, but it would be nice if it worked both ways, ie, leaving any field blank and pressing 'go' will give the required field (namely, frequency, which is what i need to work out)
 
The only thing i'm not sure of is how phase shift is corrected, with the various Xover designs, if it even needs to be? And how does this affect the sound?

It's been a while since my filters class (and I lent the book to somebody who never returned it :( ) but IIRC the order and type of crossover will determine phase shifts at the crossover points. Keep in mind that the drivers themselves will introduce phase shifts at certain frequencies as well. One of the goals is to aim for at least similar phase shifts of the two drivers at the crossover point. Wildly different phase shifts within a narrow frequency range is not good.
 
Thanks guys, that info is VERY helpful. Kirk for the phasing info and toxcrusadr for that link, that will put me streets ahead and is almost virtually what i need. Apprci8d
 
The thing you'd have to know is the impedance for which the crossover was designed. One can assume 8ohm or 4ohm in some cases, as some designers don't go beyond these nominal figure, but it's usually more complicated than this. You can find the electrical kneepoints and Q for 2nd order networks this way:

(1/Z)/(2*Pi*C) = f1

Z/(2*Pi*L) = f2

where: Z = driver impedance, C = capacitor in Farads, L = inductor in Henrys

then:

sqrt(f1*f2) = F

where: F = crossover frequency

then:

F/(highest of f1 or f2) = "Q factor", typical figures range from 0.55 to 1.0

Remember that these are the electrical figures only. What usually counts at the end is the acoustical kneepoints and slopes, something you really can not know without more involved measurements.

az
What exactly is f1 and f2 in this calculation?
 
I've used the Passive Crossover Designer a few times, presented in an .xl/spreadsheet interface

You can tweak component values and "see" what happens to the crossover point

It's a little over the top but then it's trying to cover all variables. Takes a little time to learn.
 
Speakers are seldom 8 or 4 ohms at crossover points. Thats why you have to use the appropriate test equipment and have a space to run repeatable tests.
 
Back
Top Bottom