Crossover design

I will paypal somebody to design a crossover for my 3 way system as I do not have the knowledge to correctly design one myself. Im going to use 2 HI-VI M8N woofers, 2 HI-VI M4N as midrange and a Vifa XT25BG60-04 double magnet version for the highs. Power will be from a Sony STR-DN1030 which as stated is 140watts per channel at 8ohm. I don't know how much more information is needed but I will post specs down below this. Thank you in advance!

Specs for HI-VI M8N(I know this is a bit confusing but couldn't find any other site with specs on it other than the sellers):
Hivi M8N Alloy Cone Woofer 16 Ohm | eBay

General Data Nominal Power Handling (Pnom)(W) 80 Max Power Handling (Pmax)(W) 160 Sensitivity (2.83v/1m)(dB) 88 Weight (M)(Kg) 2.3 Electrical Data Nominal Impedance (Z)(Ω) 8 DC (Re)(Ω) 13, 16 ohms nominal Voice Coil and Magnet Parameters VC Diameter (mm) 35 VC Length (H)(mm) 17.5 VC Former SV VC Frame Kapton Magnet System Outside Magnet Former Ferrite Gap Height (He)(mm) 6 Linear Excursion (Xmax)(mm) 5.8 T-S Parameters Moving Mass (Mms)(g) 36.5 Effective Piston Area (Sd)(m2) 0.0214 Equivalent Air Volume (Vas)(L) 53.5 Resonance Frequency (Fs)(Hz) 33

Specs for HI-Vi M4N:
HiVi M4N 4" Aluminum/Magnesium Midbass 297-434

Power handling: 15 watts RMS/30 watts max • VCdia: 7/8" • Impedance: 8 ohms • Re: 6.5 ohms • Frequency response: 69-8,000 Hz • Fs: 69 Hz • SPL: 82 dB 2.83V/1m • Vas: 0.16 cu. ft. • Qms: 5.90 • Qes: 1.35 • Qts: 1.08 • Xmax: 3 mm • Net weight: 1.32 lbs. • Dimensions: Overall diameter: 4-5/8", Cutout diameter: 3-15/16", Depth: 2-5/8", Magnet diameter: 2-5/8".

Specs for Vifa XT25BG60-04 1" Shielded Dual Ring Radiator Tweeter:
Vifa XT25BG60-04 1" Shielded Dual Ring Radiator Tweeter 264-1012

Power handling: 90 watts RMS/140 watts max • VCdia: 1" • Impedance: 4 ohms • Re: 2.9 ohms • Frequency response: 1,500-40,000 Hz • Fs: 630 Hz • SPL: 92.4 dB 2.83V/1m • Dimensions: Overall diameter: 4.125", Cutout diameter: 3.125", Depth: 2.20".
 
Fox,
Someone could easily give you a schematic for a crossover at whatever slope you or they may chose. In reality much of what you read here about crossovers here is pure bs. Unless you take the actual drivers you are using, and after break in of at least many hours running at maximum excursion the published values are often meaningless. The other problem is that what you truly want are impedance curves of the drivers in the actual enclosures you are going to use for accurate results. Most of the simulations done here are a joke, they are not accurate anywhere but in the computer programs that they are being simulated in. Once an accurate response and impedance curve is determined in a working enclosure only then can you really design a crossover that will optimize the each device. This also leaves out the conjugate network that should be included to fix the rising impedance curve that will tend to make any amplifier work much harder than it really needs to do. You could as easily as many here plug in numbers to a simple crossover design software, and you will get a semblance of a correct crossover and nothing more.
 

Jay1111

Member
2013-02-14 6:16 am
AZ
Fox,
Someone could easily give you a schematic for a crossover at whatever slope you or they may chose. In reality much of what you read here about crossovers here is pure bs. Unless you take the actual drivers you are using, and after break in of at least many hours running at maximum excursion the published values are often meaningless. The other problem is that what you truly want are impedance curves of the drivers in the actual enclosures you are going to use for accurate results. Most of the simulations done here are a joke, they are not accurate anywhere but in the computer programs that they are being simulated in. Once an accurate response and impedance curve is determined in a working enclosure only then can you really design a crossover that will optimize the each device. This also leaves out the conjugate network that should be included to fix the rising impedance curve that will tend to make any amplifier work much harder than it really needs to do. You could as easily as many here plug in numbers to a simple crossover design software, and you will get a semblance of a correct crossover and nothing more.

Bah, that's a lot of false generalizations :rolleyes: There are far more simulated designs with posted measurements that closely mimic the simulated result then not. Paul Carmody's write up of the process has a clear example of this, and the PE forum has many such posted examples. The simulation is as good as the entered data.

OP, I would design you a speaker for free, but you would have to pick drivers with posted measurements (preferably from a place like Zaph|Audio). Those 16 ohm woofers do not qualify.
 

Jay1111

Member
2013-02-14 6:16 am
AZ
Getting measurements from some where like Zaph (TS and IB) and then running them through response modeler which simulates baffle effects, will get you extremely close. WAY closer then any text book xover could even attempt. The measurements have proven it repeatedly. You will get a very good speaker.

If your goal is + - .5 db reference type design then yes you need to look elsewhere.
 
re:"Electrical Data Nominal Impedance (Z)(Ω) 8 DC (Re)(Ω) 13, 16 ohms " - so what is the nominal impedance? - I'd be asking for more info from the seller, there's no mention of a 16ohm vesion on the Swan site (doesn't mean that it doesn't exist though...)

The difference in sensitivities between the mid & woofers is going to cause probs in this design too
 
Last edited:
Jay,
Unless you not only have the real values of the speakers in question and the actual box they are used in I beg to differ with you. A simulation is just that, sorry but they are never really what you think you have.


I have to agree. I've tried loads of different crossover configurations. Taken loads of measurements and I feel i'm very close but not quite there yet with my current speakers (One more notch fiter and I think i'm done). For example, pretty much all my music sounds great (I'm over the moon with the sound) but the vocal in beautiful by Christina Aguilera sounds harsh. I'm pretty sure it's caused by a small peak at 4K - I will find out after i've built the notch filter.
There's no way I could have got this crossover through simulations alone.

It took Troels 4/5 months to get the crossover right for his PRELUDE speakers - Read this: PRELUDE
 
Last edited:
Yes, so a 3 way with a nearly 'sub' cut, so again a 6“ mid-wf would be needed.
It's called' filling the User Requirements' before prooceding in making any
order on pieces for the build.
At this point, I would make my own experiments, since the components allow for that; Choose from ePay a classic 12 dB/Oct subwoofer filter ( a coil and a cap) for an 8Ω load ; build a big box or two boxes ( they can be put at the top & bottom of mid-hi unit ).
:p
Now we are at a dead end, because there are two more ( damn ! That mid-woofer needs to be filtered! ) cut-off points to analyze.
Search for a similar design for those components ;)
Cut& paste at the age of Internet :usd:
 
I know this thread is about the drivers in question, and I also think the mid to woofer will have a big sensitivity mismatch, but simulations are only as good as the data procured before the simulation takes place. I've done several simulations where the results clearly matched the simulation. It's not as hard as one thinks if you have all the most accurate data.
Later,
Wolf
 
Jay111,
What most of us are saying is the same thing. You will need to chose three speakers that have well known parameters in a known enclosure size. If you randomly pick speakers that you do not have all the information for you will be doing yourself a disservice. With the help you can get here this can be done and done well. Just don't take the approach to shot in the dark, you will be disappointed. You will have to build three crossover sections. The most critical will be the mid section as you need a bandpass filter to handle that one unit. A second order filter is easy but I don't generally like them and unless the speakers are all very wide band devices the usually is some problem areas with that wide of a cross in frequencies. I always use 4th order Linkwitz-Riley filters, they are just as easy to design and take all of the phase difficulty out of the equation and make the overlapping passband easier to handle.