Determining filter order - What are the rules?

[Refugee]

When determining the order of the crossovers in a schematic I am a little confused. do I just count the caps and coils or do they have to be in certain configurations?
I have searched for a set of rules to determine the order that would apply to all configurations.

EX: on the schematic below -
*the low pass on the woofer (LS1) has a cap in series and a coil to ground. So it is second order.
*the band pass for LS2 has a cap and a coil but they are both in series - it seems this is a first order low pass and a first order hi pass. SO is it a first order or a second order.
*for LS3 there is a cap and coil in series and a coil to ground. My guess is a second order hi pass and a first order low pass. In this case I cannot call it a first order band pass, or a second order band pass, and I have never seen anyone mention a 1 1/2 bandpass so what is the language to use here?
* also, am I correct to consider this as four different filters (one for each driver) working together to make one crossover? The filters are ordered - not the entire crossover - right?

* another example I can think of would be two caps in series - This one I get ( I think) - in that case they would only be be there to achieve a certain value and act as a single capacitor thereby not adding an order.
Can anyone explain a set of rules that will help me define the order or at least correct the guesses that I marked on the schematic under each driver?

 

[GordonW]

You're mostly correct.

The one slight functional error is in the first stage- the low pass. That resistor in series with a capacitor, paralleled with the woofer, is a Zobel. That acts as impedance compensation, to counteract the inductive impedance-rise of the woofer with increasing frequency. Without that Zobel network, the first order crossover would not act as a "textbook" first order- it would have a different effective rolloff frequency, and it would not be truly 6dB/oct. rolloff near the crossover region- it would be a shallower slope, which would allow excess high frequency energy to pass to the woofer. The Zobel flattens out the impedance of the woofer- letting the low-pass inductor act correctly, as a first-order lowpass.

As for nomenclature- there are several standards, from different divisions of electrical engineering, that don't always agree. In classical electrical theory, a bandpass filter is referred to by the SUM of the orders of the internal filter networks. In this case- in the case of the second crossover section- a first order lowpass, in series with a first order highpass, would be a second order bandpass. In the third crossover section- a second order high-pass in series with a first-order lowpass, would be referred to as a third-order bandpass. HOWEVER- in common crossover design notation- we'd just refer to that second crossover stage, as a first-order bandpass (since it was made up of all first-order elements), and the third crossover stage, as a hybrid bandpass made of a first-order lowpass and a second order highpass.

I personally like the latter nomenclature system- that commonly used in speaker design- as it usually conveys more information about the filter design up front...

Regards,
Gordon.

 

[Refugee]

Thanks @GordonW !
Very clear explanation. I like the latter nomenclature system better too.

So R1 is there t for the Zobel effect but R2 and R6 are there for voicing and mostly effect the amplitude of the drivers, not the filter, right?

 

[GordonW]

Thanks @GordonW !
Very clear explanation. I like the latter nomenclature system better too.

So R1 is there t for the Zobel effect but R2 and R6 are there for voicing and mostly effect the amplitude of the drivers, not the filter, right?

Actually, R2 and R6 are both acting as part of a sort of "L-pad" amplitude reduction device (along with R3 and R5, respectively), AND since they're directly in parallel with the drivers- they also act to limit the inductive impedance rise, sort of like a Zobel would. The driver impedance is somewhat stabilized vs changing frequency, by the constant value of the parallel resistor. One component actually doing two things.

Regards,
Gordon.

 

[Refugee]

That clears things up. Thanks. It also gives me the proper search terms to further understand the circuit.