Crossover calculator discrepancy

[GaryM44]

Hi all . . . .

I'm trying to teach myself crossover design so that I can build a crossover for a 2-way kit I want to modify. The kit is the SEAS Idunn kit from Madisound. I want to add a second woofer (in parallel) to the design, which should increase overall SPL 3db. It will also change the net impedance of the woofers, so I'll need to modify the crossover. The modified Idunn will be combined in a single enclosure (separate compartments) with a powered sub.

To make sure my method was correct, I tried to derive the values SEAS shows for the unmodified Idunn, using this online calculator:

Passive Crossovers, Capacitor and Coil Calculator

(That is the only one I could find that allows calculation of 1st, 2nd, and 3rd-order filters. The Idunn uses a 2nd order on the woofer, 3rd order on the tweeter).

The calculator gives entirely different values from those SEAS shows (see photo).

I'm assuming the impedance value for the tweeter should take into account the L-pad. That gives a net impedance for that load of 4.75 ohm. I also tried the calculation using the tweeter's nominal 6 ohm value. The values for the tweeter filter I got are:

4.75 ohm
________

C1 10.16 uF
C2 30.46 uF
L1 0.26 mH

6 ohm
_____

C1 8.04 uF
C2 24.12 uF
L1 0.33 mH

Neither of which agrees with SEAS' values.

I also compared the low-pass filter values from the calculator with SEAS' values. They are different also.

The nominal impedance of the tweeter is 6 ohms; 8 ohms for the woofer. Specs are here:

H1571-08 U18RNX/P (woofer)

H1499-06 27TBCD/GB-DXT (tweeter)

Can anyone explain these discrepancies?

Thx!

 

[jjrenman]

In a nutshell the crossover calculator that you are using does not state which type of filter network the calculations are for. As in Bessel, Butterworth etc. The values will be different as each type of filter gives a different "Q" at the x over point.

 

[GaryM44]

In a nutshell the crossover calculator that you are using does not state which type of filter network the calculations are for. As in Bessel, Butterworth etc. The values will be different as each type of filter gives a different "Q" at the x over point.

Isn't that just a matter of the topology? The overall SEAS crossover is a hybrid, with a 3rd order Butterworth on the tweeter, and a 2nd order L-R on the woofer. Isn't a L-R just 2 cascaded Butterworths?

 

[Jim Griffin]

Now I don't know for sure but it is likely that the Seas design was derived from actual in-box measurements and resultant computer optimized crossover network design to a specific desired response curve. The Seas design likely then considers the actual acoustical response of the enclosure and drivers so you are far from an ideal cookbook filter situation.

 

[fastbike1]

Also you'll get 6 dB from a second woofer in parallel. 3dB for reduced impedance, 3dB for doubled cone area.

 

[balerit]

Isn't that just a matter of the topology? The overall SEAS crossover is a hybrid, with a 3rd order Butterworth on the tweeter, and a 2nd order L-R on the woofer. Isn't a L-R just 2 cascaded Butterworths?

Yes but the multipliers are different.

 

[dumptruck]

Yeah, Seas isn't going to be using textbook resistive load filters. If you just want to put a second woofer on the same filter without using two filters, though, you can just halve the inductance and double the capacitance. I can't say I recommend it, though. Two drivers running the same signal is a whole different can of worms, because now you've got to consider the center-to-center spacing between those two drivers, as well as the spacing of each to the tweeter. I guess you could try it as an MTM... Also, you've got to be careful you aren't ending up with 2Ω at some point in the bass range or something.

 

[icsohn94]

As you know well, the theory and practice is different.
There are some factors that makes the discrepancies.

The driver's frequency response is not flat.
The driver's impedence is fluctuated depending on the frequency.
The distortion and directivity is different depending on the frequency and drivers.
When the driver is installed on enclosure, the frequency response is being fluctuated due to baffle step effect.

The aboves are the reasons why the calculator result and Seas' circuit is different.


If you change the woofer from single to double, Dumptruck's comment is basically right.
However, a constructive interference is occured when the woofer is doubled with stacking.
The result is that the decibel of lower frequency is rised upto 6dB, not 3dB.
So, for compensating this effect, the less attenuation of tweeter is needed.

 

[GaryM44]

If you change the woofer from single to double, Dumptruck's comment is basically right.
However, a constructive interference is occured when the woofer is doubled with stacking.
The result is that the decibel of lower frequency is rised upto 6dB, not 3dB.
So, for compensating this effect, the less attenuation of tweeter is needed.

Yes. I figured I'd remove the L-pad in the SEAS tweeter circuit. May also need to add a 1-2 ohm series resistor in the woofer circuit.

Thanks!

 

[GaryM44]

dumptruck ---

Good advice. Have also considered using the 2nd woofer as a BSC driver, with an additional coil between the two.

Thanks!