Designing Crossovers and Measurements / Listening

I started with a rough 2nd order crossover at 3500Hz and refined it from there. My first measurement showed a bit of peaking in the midwoofer response around 3800Hz, so I changed the filter on that driver to a 3rd order and lowered it to around 3000Hz. I then adjusted the 2nd order filter on the tweeter to that lower frequency and arrived at a pretty good response. I may do some more fine adjustments, but this is a very good start.
I also tried a first order crossover on both just to see how that looks. Again, it let in a bit too much of the peaking in the midwoofer response, so I ruled it out.
As for how it sounds, it's early yet, but I'd say it's very good. The midrange sounds clear and smooth and blends seamlessly with the tweeter. Lower midrange sounds full and neutal. I played a couple of songs (in mono) that I'm very familiar with to judge. Unfortunately, I can't record those in a published video, since that would be copyright infringement.
Next step is to make the solid wood baffles for both speakers and that will be the subject of the next video.
This is part 2 in my series on designing making a pair of open baffle 2-way speakers to use as "surround" speakers in my listening room.

You can help support the work I do in making these videos:
Support this channel on Patreon:

#diyspeakers
#johnheisz
#audio

My "Scrap bin" channel:

My main channel:

[IBuildIt]

Lost? Don't know what I'm talking about in this video? Watch the first video on this project:
I started with a rough 2nd order crossover at 3500Hz and refined it from there. My first measurement showed a bit of peaking in the midwoofer response around 3800Hz, so I changed the filter on that driver to a 3rd order and lowered it to around 3000Hz. I then adjusted the 2nd order filter on the tweeter to that lower frequency and arrived at a pretty good response. I may do some more fine adjustments, but this is a very good start.
I also tried a first order crossover on both just to see how that looks. Again, it let in a bit too much of the peaking in the midwoofer response, so I ruled it out.
As for how it sounds, it's early yet, but I'd say it's very good. The midrange sounds clear and smooth and blends seamlessly with the tweeter. Lower midrange sounds full and neutal. I played a couple of songs (in mono) that I'm very familiar with to judge. Unfortunately, I can't record those in a published video, since that would be copyright infringement.
Next step is to make the solid wood baffles for both speakers and that will be the subject of the next video.

[michaeltablet]

Another great build! Thanks for sharing your gifts with us! Love all your channels.

[mndlessdrwer]

Glad to see more people discovering the lovely dual ring radiator tweeter design. They're one of my absolute favorite tweeter designs out there, though there are some challenges to using them, namely finding a midrange driver that suits their extraordinary sound diffusion.

[AmmoDude]

On two ways, I start with a poly cap on the tweeter and an air core inductor on the woofer and go from there. I find 2.5K or less works well with a two way, usually settling around 2K. Sometimes the woofer needs a zobel filter, sometimes the tweeter needs a resistor or two to take the harshness out. All done by ear. People that hear/buy my "creations" cannot believe the sound I can achieve from 6 1/2", 8" or 10" woofer and dome tweeter. My position is less is more, I never have to go past a 2nd order xover on a 2-way system. Neither does JBL, Klipsch, AR, Polk, and many others....AR was famous for using only one cap on the tweeter, woofer full range. My JBL 4312 Control Monitors run the woofer full range, no components on the woofer at all.

[NeilBlanchard]

I like dual ring tweeters - and they typically have a nice low Fs. So the crossover could be much lower - like around 2kHz. This would avoid the breakup mode on the woofer.
And something I have done to design crossovers that is quick and very flexible - use a miniDSP 2x4 HD and separate amp channels for each driver. You can quickly change the crossover points on each driver and the slopes, as well as the type of filter. No soldering and you are not limited by the values of parts that you have on hand.

[strongchallenger]

Wish I have the knowledge like you do. I love your video. So educational. ♥️👍🙏

[isaeagle]

Cone breakup. Another way to mitigate that is with a tank filter. A .33uf cap with a 3 ohm resistor in series and the combo in parallel with the inductor. The resistor is necessary to keep the impedance safe above 20khz.

[mrpres]

I'm surprised you didn't recess the tweeter. There was a video I saw once where tests were done on recessed vs non recessed tweeters and the data showed that the tweeters measured much better when flush with the baffle

[weeverob]

John, can you recommend any books on designing and building crossovers for people who know nothing about them. do you think active crossovers are the direction they will be going in speaker design in the future? tx!

[danielesbordone]

Very nice speakers, maybe a crossover point of 2500 Hz. could have been enough with that tweeter because of it's low Fs value of 500 Hz.

[amdenis]

It’s a dip at 5K that you needed to solve for, more than the peak at 3.5K, since it is at 2.5K as well, and is twice as wide there.

[edverbeek]

Which filter sounded the most natural and smooth?

[martinenstrom]

You need to think about summation and phase. You don’t want amplified frequencies thrown in a different direction.

[johncnorris]

Thanks for the video.

Once the woofer is in an enclosure won't that effect the results of your measurements? Were you trying to get a rough idea of how the speaker would sound or was there another objective for testing it early?

[impuls]

I have a trained ear after playing around with REW too much and that peak would have driven me nuts ;) Bought some SB acoustic 6" which has that brakup pushed all the way up to 9kHz and I still noticed it light up with a 2. order at 3kHz. At 3. order 2500kHz it was still noticable but I calculated the quarter wavelength and put a foam piece that diameter in front of the 6". That made it go completely away :). Using foam for absorbtion and diffraction took removed it away. Btw the brakeup did not show up in REW with the 3. order but I could still hear it!