Piezo crossover

[Pbassred]

I've BOUGHT (no offence) A pair of Carlsbro cabs for my P.A. (public abuse) system. They consist of an 8ohm 10" plus a Piezo horn. Carlsbro claim a crossover frequency of 3.5kHz.

Naturally, I pulled them apart They are throwably solid, lined . The lack of performance below 80Hz is of no consiquence since I bough them to run as tops above 300 - 800Hz (haven't decided.) at 150W

I am confused by the Xover. The 10" has no filter at all. THe Piezo has a 16ohm resistor in series ( so far so good.) It also has apparently a butterworth high pass filter with a 2.2uF cap in series. and a parralell inductor with no markings.

Now, Reading this forum.....
a/ Don't piezos need a different kind of filtering?

b/ No filter at all on the LF speaker means that it is trying to produce HF sound - poorly.

c/ Butterworth filters result in a 180deg phase shift. So the piezo must be out of phase with the 10" driver.

So its trying to produce HF sound, poorly, out of phase. Am I wrong? What would you suggest?

[djQUAN]

most low cost systems don't use lowpass filters for the woofers. the low pass is achieved by the upper limit of the woofer response. the 16ohm resistor is used to isolate the capacitive nature of the piezo from the amplifier (prevent it from oscillating)

just some ideas, although I may be wrong......

[Pbassred]

low pass is achieved by the upper limit of the woofer response

Or as WE call it "when the woofers sound nasty"

Some 10" drivers are spec'ed up to 5kHz, but that doesn't mean its a good idea! The components to add a low pass filter might be $3 but I reckon it was made in China for $30. That makes the extra components a big deal to the factory.

Anyway, just because its cheap, we don't have to live with it. What simple things can we do to make it not distort, and be in phase?
Suplimentary question . I don't know how to spec the power ratings for the components.

[BillFitzmaurice]

I suspect that your horn isn't piezo, as one should never use an LC highpass on a piezo. If it indeed has no magnet then somebody goofed.

[Pbassred]

Dear Respected Bill,
Yeah, I figured that for $30 you don't much intellience - and not from China. Am I correct in my assumptions?
If I removed the coil but left the Resistor and the 2.2uF cap would that still change the phase?

How do I get from goof to better?

[djQUAN]

I also bought a three way speaker system (very cheap) with 8 ohm drivers, 8", 4" and piezo horn. although it sounded quite decent, I wasn't satisfied. I removed the passive crossover, added a few pairs of binding post and went tri amping. (quad amp if you include the subs) it sounded quite good!

[BillFitzmaurice]

How do you know what the phase should be? Perhaps that part was properly accomplished; the only way to know is to view the system output with an RTA to see if there is a dip where the driver outputs overlap. Without a resistive load the cap isn't doing anything and it wouldn't hurt to pull it and the coil, if you are absolutely sure it is not a dynamic tweeter. I'm still not sure anyone would add to the cost of the cabinet by installing unnecessary components.

[Swedish Chef]

Phase confusion:

c/ Butterworth filters result in a 180deg phase shift. So the piezo must be out of phase with the 10" driver.

Well, since you don't know the value of the inductor you have no way of telling whether the filter has a Butterworth response or not. Anyway, it is a 2nd order (12 dB/oct) filter and if it is of Butterworth response it will have a phase shift of 90 degrees at the crossover point (-3 dB). If there was a 2nd order crossover for the woofer too the two drivers would be out of phase by 180 degrees relative to each other @ the XO point, hence one usually reverse the polarity of the tweeter. But in this case you only have a benign shift of 90 degrees so it isn't much to worry about. Focus on the fq response instead (which will probably be anything but flat for these babies). Either by crossover design or with an EQ, the latter not wasting amp power.

But like Bill I have my doubts whether this tweeter really is a piezo. If you measure it with an ohmmeter what do you come up with?

/M

[Pbassred]

Having unscrewed the driver - Definatly piezo. Its a nasty one. I've tried pluging it in reversed and I can't hear the difference Replacing the piezo isnt a problem. I'll throw a CTS (Motorola) in there soon.

Compared to my old Warfdale XP2s - lifeless and distant. Cheap does as cheap is. Hey.... I'm not griping! great boxes and I can buy "real" drivers later.

My question was relly about whether The Peizo should have that kind of filter at all! A series resistor and cap I could understand.

Also what to do about providing a filter for the LH driver? I can't properly spec the coil type ( the values I guess at 0.5mH and 8uF).

[Swedish Chef]

My question was relly about whether The Peizo should have that kind of filter at all! A series resistor and cap I could understand.

I would not use any crossover at all for the piezo (apart from a resistor which is good for some of them). For some exceptionally nasty piezos like the KSN-1005 it might be beneficial though.

For the woofer, a 2nd order LP filter may clean up the breakup distortion up high. But it pretty much requires a spectrum analyzer or very good modeling to get it right. Take a look at this document written by Wayne Parham at Pi Speakers, http://www.pispeakers.com/Speaker_Crossover.doc
IMO the best introduction to passive crossover design I have ever read.

/Magnus