RS225-4 & RST28A-4 8+1 Two Way issues??

[pick]

Over the years, I have built a few 2-way designs based on 8-inch woofers. I was recently reminded how these sound when I built a quick and dirty 8+1 two way for my son’s college house based on the Dayton DC200 and DC28F (picture attached). This design sounds amazingly good for the low price, ease of build, and simplicity of design, even without making the extra effort to match off-axis response of the two drivers.

I also made a 8+1 two way based on the RS225-4 and the RS28A. I have never experienced any issue with the sound due to the breakup modes or off axis response. That said, these are only used as the rear channels to compliment the front channel MTMs based on RS180 and RS28 drivers. I have attached data for this design. The data is based on my measurements, not the stock data for the drivers.

I have ready many posts over the years about the poor performance of 8+2 two-way overall and about the extreme measures taken to use the RS225 products in these designs. I recall Mark K created a design using an elliptical filter with 8^th order response. All of my designs using the aluminum reference series drivers use elliptical filters resulting in 4^th – 5^th order effective response. I often find the extreme slope filters are more difficult to get integrated properly and with good phase alignment. From the attached plots, you can see the drivers integrate well on axis and response peaks at the breakup frequencies are well attenuated in the 4^th order system. I am interested in understanding why so many designers feel more extreme measures are needed to use these drivers in a two-way design. I am interested in opinions related to the driver’s bad behavior at high frequencies as well as beaming effects that cause mispatches in off axis response of the two drivers.

As always, thanks for any responses. I enjoy exploring the opinions of this group, understanding if/why they differ from what I want to believe, and adjusting my design strategies accordingly.

[fpitas]

Off-axis issues often have a lot to do with the listening room and speaker placement. If the room is very reflective, and there are nearby walls, there's a good chance you'll hear issues. At the opposite extreme, if you listen in an anechoic chamber, off-axis response is of little importance.

For a while, it was claimed that the best listening room was highly reflective. Even JBL fell for that; the Salon2 room at Times Square was designed that way. I hated it. Like listening in a tiled bathroom. Later they added some serious wall absorbers, and it sounded much better.

[pick]

Thanks for your feedback. I also don't like a highly reflective environments. My listening spaces are typically drywall constructions that does have nearby walls but not within a few feet of the speakers (except the back wall is very close and offers some low-frequency reinforcement). Also, most of my time is spent listening on axis.

[fpitas]

If you're a speaker manufacturer, you have no control over what room or placement is used. So, it's a good idea to have good off-axis performance. Of course, that never hurts anything.

[isaeagle4031]

That's a very low crossover point for this tweeter and not one I would recommend beyond very low listening. It just doesn't have the capability to be used that low without Xmax and distortion being issues. A crossover point of 1.8 to 2khz will go a long way in protecting the tweeter and still get a good off axis and power response.

[Geoff Millar]

I'll happily defer to someone who knows infinitely more than I about crossovers and driver behaviour, but the Tritrix speakers also cross the DC28 fairly low and sound good - especially for the price - even when turned up a bit. Is that because the 5" mids integrate better with the DC28 than the single 8", or is it due to the Tritrix' crossover slope for the tweeter, which is fairly steep?

Interested in your thoughts, thank you

Geoff

[pick]

To be clear, the picture and graphs are not from the same design. I should not have shown them that way. The xover used for the DC28 is just over 2kHz. The graph shows the design using the RS28A, which had a resonance frequency below 600Hz. As I recall, this design crossed over the tweeter at 1400-1500 Hz.

[Wolf]

In 8" 2ways, it also has to do with the directivity error at the woof to tweet transition. There are ways to mitigate it, whether waveguided, early rolloffs, or other means.

Steep filters help suppress breakup, as do elliptical or notch filtering. -25dB from ref should be minimum, -40dB is better, and -50dB is basically gone. These benchmarks help a lot.

Wolf