Hey all,

Recently a guy gave me a couple old home-built speakers that I scrapped out for parts.

The passive crossovers were made by Peavey (and mis-matched), and I can't seem to find the specs on them.

I'm not sure how to describe them, as they seem to be some sort of 1st order-Low / 2nd order-High (??).

PV# 70501558, C1=2.0uF, L1=0.363mH, L2=0.60mH
PV# 30501592, C1=1.5uF, L1=0.342mH, L2=0.738mH

Can anybody decipher the xover freq's on these?




Regards,
JohnR

 

Without knowing the inductance of the coils getting exact responses is not likely. If you have an oscillator and an AC voltmeter and 2 resistors of around 100 ohms you can connect the oscillator to the input and the resistors to the outputs and then measure the response at different frequencies with the voltmeter. Do it every 50 Hz up until about 500 Hz, then every hundred Hz up to 2000 Hz that should cover the crossover range for a two-way crossover.

 

Between 6k and 7k is what I calculate for the high pass as a 2nd order.
But if I figure the low pass as a 1st order, I come up with 1800-2000-ish?
which means there will be a big "hole" in the middle?
Again, that is why I pose this question, I'm not familiar with this topology, and I'm quite certainly confused.
Perhaps there is some other interaction between the HP and LP that I don't understand?

Forget about speaker response and impedance curves, and what we "guess it might be"...
I want to understand this circuit in a theoretical 8 ohm load on each output.
I already tried plugging it into XSIM with FRD/ZMA files for a known tweeter and woofer, and it gave me a bizarre response.
there is something about this design that doesn't make sense.

JohnR

 

The bass loudspeaker inductance will have an effect on the bass part of the crossover. The inductance is generally chosen based on the frequency response of the bass unit which will naturally roll off around the 1kHz to 5kHz region depending on the design of the cone.

What does the tweeter look like? You might be able to get a spec for its recommended crossover frequency.

 

The explanation for how the bass portion works is not as a filter but as a series impedance, The reactance of the 0.60 or 0.73 L2 inductor in series with the resistance of the 8 ohm speaker. So the response half power point is where Xl = L2 x F x 2π = 8Ω, because at that point half the voltage is across each element is the same.
For the tweeter it is a bit more complex because there are two inductances in parallel, unless it is not an inductive tweeter. So at the half power point, the series reactance of the capacitor, Xc=1/C x F x 2π = 8Ω (in parallel with L1 x F x 2π)

So filter equations are not applicable, it is more functionally two resistances in series with one of them dependent on frequency.

 

The explanation for how the bass portion works is not as a filter but as a series impedance, The reactance of the 0.60 or 0.73 L2 inductor in series with the resistance of the 8 ohm speaker. So the response half power point is where Xl = L2 x F x 2π = 8Ω, because at that point half the voltage is across each element is the same.
For the tweeter it is a bit more complex because there are two inductances in parallel, unless it is not an inductive tweeter. So at the half power point, the series reactance of the capacitor, Xc=1/C x F x 2π = 8Ω (in parallel with L1 x F x 2π)

So filter equations are not applicable, it is more functionally two resistances in series with one of them dependent on frequency.

Not so:
Firstly, what you described IS a first-order LR filter.
Secondly, you are assuming that the bass speaker is a pure 8Ω resistance at the -3dB frequency, which it isn’t. It will be about 6Ω in series with an inductance somewhere between 300uH and 1mH.
The treble filter does not form two parallel inductances, because one of them (the tweeter itself) is an inductor in series with an 8Ω resistance.

 

Not so:
Firstly, what you described IS a first-order LR filter.
Secondly, you are assuming that the bass speaker is a pure 8Ω resistance at the -3dB frequency, which it isn’t. It will be about 6Ω in series with an inductance somewhere between 300uH and 1mH.
The treble filter does not form two parallel inductances, because one of them (the tweeter itself) is an inductor in series with an 8Ω resistance.

OK, so calculate the LR filter characteristic curve when the only resistance is the wire and only the inductance of L2 is known. T the then compare those results with the results of my evaluation. The TS already did the math and the results were not reasonable, both thy the TS evaluation and by mine. I will certainly look forward to seeing how your results compare.

 

OK, so calculate the LR filter characteristic curve when the only resistance is the wire and only the inductance of L2 is known. T the then compare those results with the results of my evaluation. The TS already did the math and the results were not reasonable, both thy the TS evaluation and by mine. I will certainly look forward to seeing how your results compare.

First I need to know the inductance and resistance of the speaker and its response curve.
Peavey won't have just thrown together a crossover, it will have been designed in an anechoic chamber to optimise the bass inductor to keep the response as flat as possible.

 

"GOOD LUCK" with creating a flat crossover system. My purpose for having one is to have the system operate more effectively by not heating the drivers with pwer that they can not reproduce. No other reason at all. The crossover is very simply to send the different frequency ranges to the correct transducers. Flatness of the reproduced spectrum is the responsibility of the equalizer and tone controls. And it has been at least 30 years since I heard a flat sound system.

 

And it has been at least 30 years since I heard a flat sound system.

Ha, when I started working in pro-sound back in the 70's all our sound systems were "flat".
Back then, you just did the maths, and everything worked as intended.
It wasn't until everybody started jumping on that inductance/impedance/response curve bandwagon that every got wacky.
I guess my point was that I do not have the PV boxes these were intended for.
I found them in my junk pile from something I scrapped out years back.
I'm in the middle of putting together a couple little monitors for my practice room and thought I might put these to use.

And my question was, what is the difference between these two mis-matched x-overs?

 

The purpose of the crossover is to avoid cone breakup in the woofers besides the fact that tweeters are terrible at pushing air around. It is almost impossible to get "flat" response. Added to that there is phase distortion introduced at the cross-over frequency. And that is unavoidable because that this the nature of frequency filters.

If you want to eliminate the crossover you can have a look at electrostatic speakers.

 

Does phase distortion matter at 96dB sound levels? or at 105 dB levels?? and 20% harmonic distortion levels???

 

Alrighty folks,

Ya'll have done the obvious and taken a simple question and turned it into a battle of "expert" opinions.

I've gotten my answer, and now will bow out and leave this discussion to the multitude of idle hands....

I've got work to do...

JohnR

 

I tossed in phase distortion because it is a fact of all frequency filter networks and it is little understood how it impacts on human perception.

Your can read more about it here:

https://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2

 

Many years ago I created an audio phase shifting system to see what it could do with the effect. The effect was very interesting and it really bothered the cat a lot. The system allowed me to continuously rotate the phase of one signal between two channels of an amplifier so that the difference could vary from zero to 180 degrees. with the spin of a knob.

 

The wavelength of a 1kHz signal is about 34cm.
5kHz wavelength is 6cm.

Now consider ½ and ¼ wavelengths. Just rotating your head is going to easily cause a ½ wavelength shift.
Or just moving one loudspeaker by 3cm can cause ½ wavelength shift at 5kHz.

Can anyone discern such phase shifts with music content?
I am not an audiophool.

 

With the exception of concerts in open stadiums, most folks listen to most music in rooms that reflect sounds, to some extent, from the walls and ceiling. The exception being those who use headphones. So it seems that most music listening is done inareas subject to reflections that result in multiple different delays. So it is not likely that most folks have ever listened to anything without a lot of phase shifts very much.
Given that reality, who is able to discern what sound without any phase shift is like??? Then consider that sound without any phase shifts would consist of steady tones, which in my opinion, would be rather boring after a short time. But, to each their own.

 

Sounds arriving within 40ms of each other are perceived as a single sound and the delay between arrival at each ear used to determine the direction. Sounds arriving after 40ms are heard as an echo.
Look up “Haas effect”
https://en.wikipedia.org/wiki/Precedence_effect

 

All good hifi speakers I have heard sound excellent except one local "high end" brand. Their speakers were very expensive and had very nice looking woodwork. I bought crossover coils from them for the speakers I designed and made, but mine sounded bad so I went to their showroom with a pink noise generator, an amplifier, a sweep detector circuit and an oscilloscope. Their speakers also sounded bad and my tests showed "a hole" in the midrange frequencies, because the inductance of a crossover coil was wrong.
I think the person who made their coils got the "6" upside down making a "9".

 

There may have been an error between engineering and production. That does happen. Often accounting is to blame, sometimes it is purchasing that goofs things up.