• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Help me make sense of this passive crossover

arvidb

Member
Joined
Dec 1, 2022
Messages
96
Likes
114
Location
Sweden
Hi!

I have these old (and probably rather cheap) speakers. I thought it might be a fun exercise to bi-amp them. Not that I expect to be able to make them sound great, but rather as a learning experience. If I mess them up it's not the end of the world.

I'm thinking a single board computer running Linux, CamillaDSP, maybe an ICEpower 200AS2 amp. (I'd love to tri-amp but synchronized output of more than 2 channels doesn't seem to come cheap: €500+ for a multichannel DAC seems unreasonable to me, when a 2-channel one costs €30 or so. Anyone know of, say, a 4-channel DAC module for the Beaglebone McASP interface?)

Here are the guinea pigs (one with four oscilloscope probe leads coming out of its tweeter hole):
SAM_3549_pigs.jpg


The name plate says:
Pioneer S-X450
Impedance: 8 Ω
Music Power (DIN): 140 W
(Whatever that means. Is/was there really a DIN standard for "music power"?)

The drivers' active diameters measure approximately 164 mm (6.5 in or 0.21 french royal feet ;) - probably marketed as 8 in?), 69 mm (2.7 in), and 25 mm (1 in), respectively.

Construction is rather thin particle board and old-school point-to-point crossover. Hmm, these actually seem even worse than I expected! Oh well. Maybe I should have expected this given the asymmetric, surface mounted drivers.


Anyway, here's the schematic of the crossover:
SAM_3547_schematic.jpg


I don't have any experience of crossover design. Woofer section looks like a 2nd order filter, right? Are the midrange and tweeter third order? Or what's with those "extra" "3.3/63V-" capacitors?

(The midrange damping resistor is a bit funny, I think they've used what's actually a thermal fuse using a low temperature solder bead to spring the wire open at a certain temperature. I've heard about these but I've not seen one in real life before! Anyway, I guess it's only used as a resistor here.)


Now the fun part: I hooked the crossover up to a Siglent SDS1104X-E oscilloscope and used its Bode plot functionality to characterise it:
SAM_3545_internals.jpg


SAM_3542_osc.jpg

How this works is that the 'scope tells a signal generator to output a series of frequencies - here the signal generator is my PC, hooked up to the 'scope over Ethernet; the computer's audio output is then connected to an audio amplifier and then on to the speaker - and then compares the input (channel 1 here, hooked up at the input of the crossover inside the speaker) to the other channels (ch2 being the bass here, ch3 midrange, and ch4 tweeter) and records the amplitude and phase differences. It does this for a configurable number of frequencies per octave to generate the Bode plot. The scope's small screen gets pretty cluttered so I saved the data to a USB stick and plotted it with gnuplot instead:
Pioneer_S-X450-2022-11-30-ampl.png

Pioneer_S-X450-2022-11-30-phase.png



I'm surprised at the high crossover frequencies (about 1.9 kHz and 6.1 kHz). I'm wondering if I've made some kind of mistake here or if this is reasonable.
 

Eetu

Addicted to Fun and Learning
Forum Donor
Joined
Mar 11, 2020
Messages
763
Likes
1,180
Location
Helsinki
Strange, the XO's are almost what you would expect from a tweeter + super tweeter but the driver sure looks like a midrange unit.

Can't help you with the schematics but there's some inexpensive 8ch sound cards out there that could work with Linux, this for example.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
2,018
Likes
1,241
Location
Australia
Hi!



I'm surprised at the high crossover frequencies (about 1.9 kHz and 6.1 kHz). I'm wondering if I've made some kind of mistake here or if this is reasonable.

Nice work so far, and welcome.

I am not overly surprised by the XO freqs.
(It is much safer to cross them a bit high.)
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
I think this is typical of older pioneer design. They seemed to run woofers pretty high.
 
OP
arvidb

arvidb

Member
Joined
Dec 1, 2022
Messages
96
Likes
114
Location
Sweden
Thank you, all, for confirming that my measurements may not be completely off. :) And thanks for your kind words and welcome!

I have not seen this combination of 2nd order lowpass and 3rd order highpass mentioned/recommended anywhere before. What do you make of it? Could there be some cleverness behind it, or is it more a "capacitors are cheaper than inductors and we can get away with using 2nd orders for the lowpasses" thing?

Looking at the actual rolloffs I get this (from the cursor readback in gnuplot):
  • Woofer LP rolloff: 16 dB/octave (2 kHz - 4 kHz)
  • Midrange LP rolloff: 19 dB/octave (6 kHz - 12 kHz)
  • Midrange HP rolloff: 20 dB/octave (800 Hz - 1.6 kHz)
  • Tweeter HP rolloff: 20 dB/octave (2.5 kHz - 5.0 kHz)
So it's a bit off from the expected 12 dB/octave and 18 dB/octave, respectively. Is it within expected component tolerances and measurement uncertainty? Or is there something else going on here? (Can driver inductance affect the response - and be part of the design - for example?)
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
So it's a bit off from the expected 12 dB/octave and 18 dB/octave, respectively. Is it within expected component tolerances and measurement uncertainty? Or is there something else going on here? (Can driver inductance affect the response - and be part of the design - for example?)
Certainly - actually the driver impedance and its own frequency response must be part of the design.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,191
Location
Northern Virginia, USA
Certainly - actually the driver impedance and its own frequency response must be part of the design.
Yes. What counts in the end is the acoustic crossover slope and magnitude.
 
OP
arvidb

arvidb

Member
Joined
Dec 1, 2022
Messages
96
Likes
114
Location
Sweden
Certainly - actually the driver impedance and its own frequency response must be part of the design.
I see. I'm starting to understand why passive crossovers are considered more difficult to get right. Also I did not realise that it was possible to get filter slopes that are not multiples of 6 dB/octave!

So the question for me now, if I decide to go ahead and bi-amp these speakers, is: do I keep the midrange's passive 3rd order high pass and try to fit the active woofer crossover to it, or do I bypass it?

Except for low frequency protection for the midrange, would I lose anything by bypassing it? What happens to phase matching to the tweeter for example?
 

Waxx

Major Contributor
Joined
Dec 12, 2021
Messages
1,931
Likes
7,688
Location
Wodecq, Hainaut, Belgium
I would first measure the drivers seperate in the cabinet and start again to mke it work. Without raw driver measurements we can't tell much. we need impendance and frequency response measurment done with the mic pointed at the tweeter, preferable on 1m distance and with a test signal of 2.83v (but surely all the same mic position and test signal power).
 
OP
arvidb

arvidb

Member
Joined
Dec 1, 2022
Messages
96
Likes
114
Location
Sweden
Okay, fair enough. :) I already have a UMIK-1, and I might be able to repurpose the bedroom for some quasi-anechoic measurements. How to mount the speaker under test though, that's another matter... I'll have to think about that!
  • Do I aim at the tweeter despite the speakers being asymmetric (and not mirrored pairs, at that)? Or do I aim at the center line at tweeter height? (Does it matter?)
  • Woofer+midrange measurements should be enough, right? (Given midrange/tweeter crossover is fixed anyway.) Or do I need the tweeter data as well? (Since they'll be bi-amped they'll have only two pairs of connection posts and it would be nice to be able to do all tests at once, without having to rewire.)
  • What are reasonable frequency ranges for sweeping the drivers? I guess I don't want to sweep the midrange (or tweeter!) from 20 Hz?
Anyway, I guess a good start would be to install REW and learn how to use that mike!
 

jschwender

Member
Joined
Jan 3, 2022
Messages
30
Likes
11
Location
GERMANY
In 1960 the DIN 45500 was defined, and in 1996 replaced by EN 61305. Among other things it defines how power capacity of a speaker was measured: the signa, the duration…. If you look at other marketing power figures like "PMPO" that all makes sense.
 

muskrat

Member
Joined
Sep 28, 2017
Messages
18
Likes
15
The 2nd cap in a 3rd order high pass is usually larger than these, hence 20db instead of 18. Why? Anybody's guess.
 
Top Bottom