Help deciphering crossover

[percy]

This is from a Polk Monitor 60 speaker. 2-way 4 driver. Sorry for the scribble drawing and horrible picture.
The circuit in the pic attached is basically the low pass side of the 2-way. The tweeter xover is basically a classic 2nd order high pass filter so dont need much help there but how exactly would you describe this low pass crossover. Three drivers in series, starts off as a low pass with an inductor in series and cap in parallel but everything below is going over my head. They look like zobel type networks (140uf + 2ohm) across the lower two drivers and the 2.7ohm in series with the 10uf cap seems like they are trying to pull off a zobel+low pass in one shot ? Is that even possible ? What do you make of this ? Please explain.
Dont have the impedance on those drivers yet but the DCR measures 1.45ohm.

[sofaspud]

I think you have what amounts to a 3-way, not a 2-way. I'm not a crossover filter expert, but my semi-educated guess says that drawing shows 2 crossover points of roughly 500Hz and 5kHz. Those numbers may be way off; I'd need to look up the correct formulas to use. How closely do they jive with the tweeter crossover? ('s OK I'm prepared for the embarrassment)

[Jay]

Originally Posted by sofaspud

I think you have what amounts to a 3-way, not a 2-way.

It is not 2-way, nor 3-way. It is 2-5/8-way Anyway, useless terminology.

There are only 2 "elements" there. First is lowpass filter (made by the C in parallel with driver), second is low frequency response attenuator (the other driver with C in parallel with it).

The lower woofers are crossed "lower" (C=140uF). Effect of attenuation from 8-ohm upper woofer in parallel with 10uF is minimal.

The upper woofer is crossed "higher" (C=10uF). Effect of attenuation from 16-ohm lower woofers in parallel with 140uF is rather significant, but it affects a very low frequencies (because 140uF creates cut-off around 200-400 Hz in my prediction).

What you have here is 3 woofers being in phase with each other, all summed to a suitable response to match with the tweeter. Upper woofer will handle more midrange. Lower woofers will help on the bottom end.

[percy]

forgot to mention - the spec says "2.25Khz second order low and high pass" for crossover.

The diagram is kinda upside down as compared to the actual location of the drivers on the baffle. The top driver in the diagram is the one at the bottom on the baffle, and the other "pair" of drivers is really towards the top of the baffle with the tweeter in the middle - like a "MTM".
Funny thing is that the lower woofer actually is crossed higher (extends more into higher frequencies) and also goes all the way in the low bass frequencies, whereas the top woofers dont go that low nor that high - like bandpass. (This is just based on ear test by running tones through, not actual measurements).

What are the resistors for btw ?

[tinitus]

3 woofers connected in series, sharing the same big series inductor
but having different zobels they probably roll off different

a kind of BSC EQ

mid and tweeter missing

woofers 'sees' the two zobels in series
plus the other driver
phase response is probably pretty weird

[AllenB]

The lower woofers in the circuit will be largely resistive above some 500Hz or so, so this shouldn't be too different to the typical 2.5 way. This may also be an attempt to deal with floor bounce.

[Jay]

Originally Posted by percy

forgot to mention - the spec says "2.25Khz second order low and high pass" for crossover.

The 3 woofers together form a response that with that of the tweeter crosses at 2.25kHz

Originally Posted by percy

The diagram is kinda upside down as compared to the actual location of the drivers on the baffle. The top driver in the diagram is the one at the bottom on the baffle, and the other "pair" of drivers is really towards the top of the baffle with the tweeter in the middle - like a "MTM".

I cannot see anything wrong with the arrangement. It can be MMTM or MTMM or even MMMT. The size/height of the box mostly determine the optimum arrangement.

Originally Posted by percy

Funny thing is that the lower woofer actually is crossed higher (extends more into higher frequencies) and also goes all the way in the low bass frequencies,

Correct. But why is it funny?

Originally Posted by percy

whereas the top woofers dont go that low nor that high - like bandpass. (This is just based on ear test by running tones through, not actual measurements).

No, it is the single woofer that behaves more like a bandpass because the very low end is attenuated while the other higher frequencies are passed through the 140uF. The top woofers are crossed by the 140uF, so if you compare with the other woofer which is only crossed by the smaller 10uF, the top woofers are crossed lower. It is just that it's response is attenuated (lower SPL) so you think that the lows (which is not the lowest) is not there.

May be you are confused, thinking that this is 3-way or something. Like I said, this is 2-5/8-way . The 3 woofers work more or less on the same frequency band. They are tuned to form a summed response that roll off and crossed with the tweeter response at 2.25kHz.

Originally Posted by percy

What are the resistors for btw ?

Resistor will help you tune the slopes. The bigger the resistor, the shallower the slope. All of these are an effort to shape a proper total response.

[Jay]

I used to draw something like this in the old days.

[percy]

Ok thanks Jay, now while I crunch that information can I throw something stupid out there –

Can this be thought of as a hybrid series+parallel crossover ? Think of the lower woofer pair and the top woofer as two separate first order series low pass filters, and then the inductor is added in series to the whole thing like in a parallel low pass filter ?
(lower and uppper as in the circuit diagram)

[AllenB]

I think you should see it as three series components. The inductor as one, the single woofer with it's zobel (as a single lumped impedance) as the second, and the other woofers and zobel as the third.

The roll-off is under the influence of the inductor, but outside that fact the relative signal to each of the two woofer portions of the crossover is directly related to their relative impedances.

One hypothetical guess may be that after the 10uF cap cuts in, the two resistors would be the only significant components in the circuit other than the inductor. With the single woofer section dropping almost 3dB more of the signal, but assuming 3dB array gain for the lower section, each section would be equal in loudness.

It would be interesting to know how much capacitance it takes to actually flatten the impedance of one of these woofers, to see where they were going with this but it is at least clear that the 2 ohms with the woofer pair is lower than required for that purpose and is intended to cut their level.

The 10uF on 2.7 ohms also cuts in fairly high. Maybe it was tuned to deal with some cone breakup, who knows.