diyAudio (
-   Multi-Way (
-   -   3-way crossover expert needed (

andrew01 24th October 2007 09:29 AM

3-way crossover expert needed
2 Attachment(s)
Hi , I'm looking for some assistance with a crossover I built a while ago . Here's the details :

3-way speaker with SS 8565-01 Bass , Vifa P13-WH Mid and Vifa XT-25 Tweeter in bass reflex enclosure , 1" MDF construction , lots of bracing and dampening etc added. Volume = 120L and tuned to 23hz.

I'm looking to improve the sound - At the moment the bass is excellent but the mid and treble are *average* , something is wrong with the crossover but I'm not sure what. It is a 6dB / Series design taken from basic textbook plan with a few minor adjustments to the tweeter padding.

I don't have any measuring equipment or software but am hoping someone can have a look at my crossover and tell me if there's something obviously wrong or what can be improved / fixed.
see attached drawing.


andrew01 24th October 2007 09:34 AM

a bit more info :
xover points are ~ 300 & 3000

I want obviously to keep the drivers & cabinet and just tweak the crossover.
The sound is overall quite *OK* but something is missing when I listen in comparison to some "highend" commercial speakers I have.
The drivers are recessed not surface mounted and in traditional 3-way positioning....T M W
Inductors are aircored low DCR and quality Caps & resisters.

Any thoughts ?:confused:

sreten 24th October 2007 10:06 AM


What is wrong with your c/o ? pretty much everything.

See :


(for elucidation try wading through : )


andrew01 24th October 2007 11:24 AM


Gee thanks , I've played with them before & got similar results to what I built.
What in particular of "pretty much everything" is wrong ?
example ??

PeteMcK 24th October 2007 11:47 AM

Have look at Andy's site:

He has a downloadable spreadsheet which should help you tweek your xover values.
Note that you need to use the ACTUAL impedance of your drivers at the xover frequencies.
If you don't have actual measurements, you can get in the ballpark by using the published impedance graphs (although that's not close enough for some here....)

Is there much difference in the sensitivities of the woofer & mid drivers?

PeteMcK 24th October 2007 12:08 PM

had a quick look at the data sheets, mid & woofer seem to be a good match for sensitivity,
but the woofer has a huge peak at 2.5KHz which will need a notch filter....

sreten 24th October 2007 01:23 PM


Originally posted by andrew01

Gee thanks , I've played with them before & got similar results to what I built.
What in particular of "pretty much everything" is wrong ?
example ??

Everything ......

The crossovers assume 8R drivers - very wrong.
The crossover is 1st order electrical - wrong because its not right.
(you have no idea of the resulting acoustic crossover alignment)
The target acoustic slopes should be acoustic, not electrical - wrong.
Baffle step and diffraction effects are ignored - very wrong.
Driver responses are ignored - wrong.
Phase integration of drivers and offsets is ignored - wrong.


2nd order high pass L/R acoustic slopes on the XT25
(~ 3khz) is possible but it is not simple see :

Complementary (but assymetric due to driver offset) ~ 2nd
order low pass L/R acoustic slopes on the P13 is possible.

Mid highpass / bass lowpass is complicated by baffle effects.
Mid driver positioning and baffle size need to be included.
No simple answer without simulation.

Take a look a RJB's 3-way :


PeteMcK 24th October 2007 11:41 PM

Andrew, I see youve asked these questions before; series xovers can be difficult to understand eh? For that reason, Id go with a parallel xover, easier to tweek.

Heres how I would proceed using the published graphs & off the shelf components:
Mid unit, a good rep for sound but known to vary from published specs, so use real measurements if poss. Id go 1st order on this, .82mH and 74uf (47+27 in parallel) in series with the driver.

Tweeter: an excellent 4 ohm unit but a rep for complaining when driven too low, Id go 2nd order here, 7.4uf (4.7+2.7) in series, .33mH in parallel, tweeter connected out of phase.
L pad, 1 ohm in series & 8.2 ohms in parallel.

Woofer. 1st order needs 3 mh in series, but this needs to be adjusted for baffle step correction; how wide is your cabinet?
(later you may prefer to go 2nd order to let the mid unit do its own thing, and this will help with the peak as well)
Notch filter: using this calc:
I get 12uF + .3mh +12ohms in series across the driver (Ive never needed a notch filter so someone else may be more help here)

Check these numbers with this calculator:
(this will also give you a feel for how component changes affect the result)

tinitus 25th October 2007 12:18 AM

Actually he has most of the components to do an ordinary and "experimental" simple paralel xo:D

4mH in series with woofer
0.42mH and 68uF in series with mid
6.8uF in series with tweeter

"easy" to ad paralel inductor on tweeter and various RC if needed
And yes, woofer may need a notch, but maybe a bit too advanced at this moment

Components are surely not the perfect values, but should give you some "feel" of the drivers and some idea of what should be done

andrew01 25th October 2007 11:33 AM

Thanks Pete.
Some ideas to mull over.
I was going to do a notch filter but the different online calculators i tried all gave different values - so I left it out :xeye:

I also tried a parallel xover as per tinitus' suggestion and I think I proved to myself there is no difference between series & parallel.

Will the cabinet naturally roll-off some of that spike at 2500hz ?

All times are GMT. The time now is 01:30 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio