$2 speaker, $10 crossover (Can somebody critique my design?)

[Jim85IROC]

I'm in the process of designing a crossover for the $.87 Onkyo mids (PE # 269-570). and $.88 tweeters (269-776).

Since I have no testing equipment, I'm forced to rely on published specs. In this case, I relied on the following frequency response & impedance drawings that PE provided.

Tweeter: http://www.partsexpress.com/pdf/269-776.pdf
woofer: http://www.partsexpress.com/pdf/269-570.pdf

Anyway, I attached my schematic. It's a 6dB/octave 6khz filter with a series notch filter for the tweeter, per the reccomendation of the write-up on the tweeter description. I also inserted a parallel notch filter on the woofer to take out the nasty 7khz spike. Even though I'm filtering at 6k, that spike would probably be quite audible if I didn't get rid of it, especially with a 6dB/octave filter

I don't really want to do a 12dB crossover (yet) considering that I'm already over $10 in crossover parts for a $2 pair of speakers. Does anybody care to critique my design or offer suggestions though?

[Volenti]

You could possibly need a zobel network on the 4'' to compensate for impedance rise at higher freq, both the x-over and notch filter won't work correctly if the driver does have a signifigant rise, and the frequency that both are being used at is high enough that would suggest to me that a zobel would be required.

[sreten]

simply put you are over complicating the issue, misaligning
the crossover will take care of the tweeter peak i.e. cross
it over higher than your intended acoustic c/o frequency.

You really need second order c/o for this.

Ignore the bass unit peak but don't listen to the
speakers on axis, set them up straight ahead.

Again 2nd order will be better than 1st.

Finally you need to add another bass mid unit and roll it
in first order for Baffle Step correction, or again use a
misaligned crossover, first order set for some BSC.

for a two way I'd suggest :
first order bass set at ~ 750hz,
2nd order treble set at ~ 4khz.

Sounds wrong, but will sound nearer right.

sreten.

[Jim85IROC]

This discussion already went over my head.

[sreten]

Originally posted by Jim85IROC
This discussion already went over my head.

sorry about that, but you can bet the manafacturers using the
drivers don't spend 5 times as much on the c/o as the drivers,
I was trying to ilustrate how they would go about doing it.

Certainly for the tweeter they'd roll it off electrically early, to
compensate for the peak, a 4KHz Butterworth high pass for
4 ohms (easy to find on the net) will give an acoustic c/o
point of around 2kHz, 2 components C series, L parellel.

The bass units peak could be ameloirated with a cheap coating.
Try the coating on a piece of paper first. It shoud dry flexible
and have a "slow" reponse to returning to shape after being
stretched. Soft varnishes are best for this.

BSC you can read up on. The point here is if the unit has a rising
response in the midrange the BSC inductor if used without a
parallel resistor can also flatten the midrange response.
BSC inductor value depends on Z and baffle width, I estimate
with a small baffle with the design f is ~ 600 to 800 Hz, and if
a parallel resistor is not used, acoustic c/o will again be about
2KHz.

Rethinking again its best to error on the too much side, so say 5k
for the treble and first order 600hz for 8 ohms (a single inductor)
for the bass unit.

With the effect of the BSC inductor midrange level from the bass
unit will be in the mid 80's dB range, which matches the tweeter
level quite well.

So I'm suggesting 3 components, (and coating the bass unit ?)

You'll also need to try connecting the tweeter phase both ways
to see which sounds best.

sreten.

[Jim85IROC]

Thank you for the clarification. I'm going to study your recommendations in order to understand them better, then I'm going to start playing with parts and see what I get.

[sreten]

Just to clarify another point :

Both units 30 degree off axis performance is substantially better
than the on axis performance, so an option is to use the speakers
parallel to walls, but design for the 30 degree responses.


sreten.

[Jim85IROC]

My hope is to use 5 of these for home theater usage if they sound adequate, which means that the center channel will be on axis unless I mount it very high. But... a little extra midrange in the center channel shouldn't hurt.

[SQ Kid]

i just bought about 2 dozen speakers from that onkyo buyout. a few different setups to play with, including some of the 4" carbon fiber speakers, a few 6", and 3 different types of tweeters. gotta love getting a ton of speakers for $20 ($30 shipped)

[Variac]

Jim, I think you could tilt the center channel driver up vertically in a sloped front cabinet to get your off axis response.