2 nd order L/R cause spike?

[mcmahon48]

I am using Xover 3 Pro for a project
When I use a 2 nd order LR filter I get a sharp spike pass the the crossover point than the 12 db
I am using Silver Flute W14RC25-04 ohm 5-1/2" Wool Cone from Madisound in parallel Xovered at 2750 or 3000

[Cal Weldon]

The response graph for that woofer shows a spike about 5K

[mcmahon48]

There should be no sharp spike apparently one of the speaker specs having to due with the capacitance of the coil is an issue. A LR 2nd order xover should be -12 at the xover point and it should be about -20 db at at 5 khz and if I use a 1st order Xover the spike goes away which has a 3 db drop at the crossover point.
And the spike is a very sharp spike no way is it similar to the rise characteristics of the manufacturer's specs

[bjorno]

Originally Posted by mcmahon48

I am using Xover 3 Pro for a project
When I use a 2 nd order LR filter I get a sharp spike pass the the crossover point than the 12 db
I am using Silver Flute W14RC25-04 ohm 5-1/2" Wool Cone from Madisound in parallel Xovered at 2750 or 3000

Hi mcmahon,

My 2 cents of XO hints: See the picture.

b

[mcmahon48]

when I Expanded your picture it was to small to make it out , Thanks for the effort and I look forward to see it larger sized

[tinitus]

try reverse tweeter polarity

[wintermute]

Is the spike you are seeing present in real measurements of the speakers using the built crossover or only in the simulation?

I've noticed in speaker workshop large peaks in response (also with second order electrical crossovers) at places where it shouldn't happen, Note that this has been when modeling the crossover for a single driver, nothing to do with summing with the other driver. I've put it down to the simulator showing resonances between the cap and the coil which shows up as a peak in the response. I haven't actually built one of the circuits to see what happens in real life, but my suspicion is that it won't happen... I'm basing this purely on the fact that they are passive components, and shouldn't be causing peaks in response of a driver that are not there without the crossover.

Tony.

[Cal Weldon]

Originally Posted by mcmahon48

to small to make it out

click on the four arrow cluster in the bottom left corner. It will expand further.

[mcmahon48]

Since the electrical energy is being passed to ground before reaching the speaker, I will build and see what happens though, but it make it harder to get a good starting point in Xover pro. And the part of the resonance is do to being in a box, whereas I will be using a hybrid design to fight the box effect.

[wintermute]

Try modelling with a small resistance in series with the cap in the low pass section (try values between 0.1 and 1). This will probably get rid of the resonant peak if that is what it is. when it comes to the real implementation you can try with and without the resistor.

edit: I've added a graph, blue is the actual woofer response, red is 4th order crossover without a resistor in series with the first cap. black is response with a one ohm resistor in series with the first cap, yellow is the 4th order LR target response. Xover with resistor is the second attachment. I'm assuming that this is the type of behaviour you are seeing?

Tony.