BOB Urz wrote:

I had a customer bring in an Infinity KPX00 crossover
for a 6.5" car unit.

Surely the customer has stereo in the car, so one avenue would be to ask
for the other cross-over as a source of comparison.

Coil L2 has burnt marks on it, Its a 24db linkquitz riley filter
(or so the spec sheet says).

OK. You know the target response, and implicitly it is the tweeter
cross-over slope this is about. Tweeter response is reasonably easy to
measure, so if you replace the coil with a guesstimate coil it is
possible to determine in what way the replacement coil is incorrect.

CALSOD comes with a lil' util called inductor.exe, it is quite possibly
not the only software on the planet for calculating inductors, and its
existence suggests that the4 coil geometry can get you a close estimate
of the inductance of the coil.

Strange, this seems to be a 3rd order filter rather than
what the tear sheets says.

It would be good design practice to design a cross-over that fits the
units so that the number of components can be reduced.

Only one coil on the high pass with two caps.
Minus the pad, it looks like the 3rd order 18 DB filter he

http://www.apicsllc.com/apics/Misc/filter2.html

and these calculated 3500hz values are (4 ohm nominal)
C=7.5 ufd
L=0.1364 mh
C2=22.7

standard replacement coil values are .1,.13,.15,.20,.27,.33 and etc.

Thoughts?

You know the coil geometry, you know the wire diameter, it should be
possible to get an estimate of the lenght of wire in the coil. That
should make it possible to calculate a probable value for its
resistance. Determining the actual resistance and comparing may provide
correction data that can be used to convert the measured impedance of
the coil to what it should have been.

Bob


Regards

Peter Larsen