Floorstanding 2-way - crossover design - guidance

[Csh1474]

I'm building a pair of floor standing 2-way speakers for my rear channel speakers. It includes a dayton DC28FS-8 1.125" Silk dome tweeter ($10) and Peerless HDS Nomex 5.25" Woofer 830873 ($20), I spent more on my center speaker and was trying to reduce cost on the rears.

This is the first set of component speakers that i bought that didn't come with an included crossover, so I've been studying the theory of crossovers for a few days now.

I don't have measuring equipment, but the manufacturers supply Impedance vs Frequency and SPL vs Frequency graphs for the speakers. I know there are some variations from the spec, but it should allow me to get closer with my speaker calcs.

I built an excel spreadsheet that calculates the crossover attenuation of the driver at each frequency. It calculates it for a 1st & 2nd order butterworth and a 2nd order Linkwitz-Riley and graphs it in decibels and power gain (0-1).

I also started to look at a Zoebel circuit and a series notch filter as the impedance is not very flat and the woofer has 0.8 mH inductance, but it looks like i can flatten out the summed output by under crossing the speakers. 2150Hz on the low and 2300 on the High.

I just can't decide between the butterworth 2nd or the L-R 2nd order. The L-R gives me a flatter summed output, but the power drops off a little early on the woofer. The butterworth drops the woofer off steeper, but the output isn't as flat.

I uploaded the spreadsheet, so that you can see the results. Any guidance would be helpful. Is this spreadsheet all wrong in its intention. It doesn't take the SPL vs frequency of the speaker into account yet, so I may need to work that in.

Thanks in advance.

[picowallspeaker]

It doesn't take the SPL vs frequency of the speaker into account yet, so I may need to work that in.

Look at all the parameters given with a speaker (and usually the manufacturer says if they are taken on a IEC baffle or closed box or vented...)
that define the mechanical and the electrical part of it .
Many software running on PC require to insert all those numbers and may model a certain box that suits your targets .That was done even before PCs !
Acoustic loading is as important as calculating filters ..it may be part of it-
A woofer going out of excursion limits is doing it probably in a bass reflex box than in a sealed one ,and so on...Man ! You are going to open a rat's nest !
The crossover is the heart of a loudspeaker system ! Some boxes of the '70s
sounded great and they had only a capacitor , or maybe a coil for the woofer
....

[nac134]

I checked out your spreadsheet. First off, the difference in "flatness" is so small that its negligible. Im guessing that you will probably not be running these speakers full range, as they are HT rears. There is NO WAY you can tell by those graphs which crossover will sound better. You would have to try both. But, the butterworth will most likely play louder, because it places less low frequency load on the tweeter. Ist order with those components is not practical for home theater. All things being equal, steeper is better. Opinions are like... well you know. But I think If you put it to a vote, the 2nd order butterworth would win.

[Csh1474]

Thanks for the advice.

I was leaning towards the second order butterworth as well. The 2nd order L-R seemed to be more affected by the inductance of the woofer and starts reducing the power well below 1kHz even though the cutoff was around 2kHz. The butterworth dropped the power on the woofer more steeply and by underlapping the crossover frequency, it still gets an almost flat summed output similar to the L-R crossover design.

[okmat]

Nice spreadsheed, however there's a lot more things to take into account when designing a loudspeaker without actual measurements...

This is probably the best way to go (and everything is free):

FRD Consortium tools guide

[Csh1474]

Great. That is exactly what I was looking for. Everything else I found was either much simpler or more complicated and required imported measurements.