Advantages of Series Crossover vs Parallel for Subs & Open Baffle

You can likely still insert the passive bass boost circuit into your series crossover. You just need to model it and work out the kinks. It's going to be very sensitive to DCR of the inductors which needs to be low. I think it will help (4-7dB is good) I just estimate it's not quite enough boost.
Hi Perry, I will play with it.
Until then, I went back to the old trusty EA15A in the series config, with the 1808 (https://www.diyaudio.com/community/threads/my-open-baffle-journey.394806/post-7959804) - the sound is gorgeous.
 
Last edited:
Hi Perry, I will play with it.

I will have to ! Along with ARTA suite, REW, VituixCAD...
All of them are loaded and installed in my PC...
I even have a dedicated external soundcard...
I will have to take the time to return to this kind of softwares and learn their idiosycrasies, after leaving my obsolete LspLAB and SpectraLAB (thank you Windows) that I used for years...
Yes : I'm lazy and miss time to (re)do all this !

T
 
  • Like
Reactions: Plott
The following is my first 3-way build with an entirely series crossover.

Woofer: GRS 8SW-4 HE 8” subwoofer
Midrange: Dayton RS125-4 ohm 5” aluminum cone midrange
Tweeter: SB Acoustics SB19ST-C000-4 with additional 3D printed waveguide (kudos to @wolf_teeth for the mid and tweeter, they were door prizes at his InDIYana event which was fun).

Photo of system:
wolf-speakers-view.jpeg


Crossover simulation in VituixCad:

wolf speakers vituixcad screen shot.png


Discussion:

The impedance peak of a midrange like this is hard to deal with, using a standard parallel crossover. I feel getting this level of performance from a 3 way crossover with only 12 passive components is very attractive.

The impedance curve is smooth, and the drive signal curves from the filter outputs on the right middle panel are very well behaved.

You'll notice a sharp dip at 180Hz. That is very angle and boundary dependent. You see hints of it in the real world measurements but overall it seems to be a non-issue. If you reverse the polarity of one driver you get a huge dip at that frequency.

On the midrange I used a notch filter instead of a standard high pass filter. That reduced phase shift, and got the vertical lobing exactly where I wanted it.

The series crossover also managed to give me “shelf” characteristics for both midrange and tweeter that were exactly what I needed, without extra components.

Measurements:

ON AXIS - The 58Hz peak is a room mode.

wolf_speakers_1.5 feet frequency response.png


Off axis Horizontal performance is outstanding, there is no visible “seam” at the crossover frequency, it exhibits quasi-Constant Directivity behavior above 200Hz.

wolf speakers 0-90 frequency response.png


Polar Heat Map

wolf system polar heat map 0-90degrees.png


I’m very happy with the polar map view as well. I generated this plot using ChatGPT from the data in the previous off axis frequency response graph.

One of the reasons this system achieves Constant Directivity behavior below 1000Hz is the diffraction characteristic of the 8" wide baffle. I discuss this in another thread: https://www.diyaudio.com/community/threads/constant-directivity-without-horns-or-waveguides.422090/

Series crossovers are definitely more tricky to juggle but if you can get the planets to line up you can kill a lot of birds with not many stones.