Off axis response and low tweeter x-over points. Never thought about...

[fjhuerta]

...this:

However, some care must be given to low-crossover-point designs to avoid “hot spots” in the off-axis response that can occur as the tweeter begins to pick up in the critical upper midrange region.

From the Dr. K's plans.

What does he mean by "hot spots"? I'm guilty of designing mainly on-axis, and I think I can hear what Darren says happens with low crossover points. I'd call it "beaming"... as if the tweeter, in the midrange region, "beamed" some extra energy to the listener.

I never took any off axis measurements to prove / disprove this theory, but now it sounds a bit logical.

So, what's the correct way to design a speaker off axis with low crossover points? How can those "hot spots" be avoided?

[cjd]

Hot spots?

Yet ignoring poor integration and driver breakup is OK?

I have no idea what he's talking about specifically. Hot spot to me would equate to excess output (i.e. peak in response). I've not experienced this myself.

C

[Mudjock]

Clearly, the dispersion pattern of a 1" dome tweeter is significantly different than that of a 6-7" midwoofer. What you really have is a tradeoff. As you move the crossover point down, the midwoofer beams less and behaves more like the dome tweeter. The downside is that the ear, in general, is more senstive to subtle changes at 1 kHz vs. something higher like 2-3 kHz. The best solution, for most people probably remains to operate your drivers within their ideal range in terms of distortion, frequency response, and off-axis response (which for something like the RS180 tops out around 1.4 kHz - then you need to choose a tweeter that can comfortably handle that.)

For MTM's, the choice is constrained somewhat more by lobing effects caused by interference between the two midwoofers playing over the same frequency range.

[fjhuerta]

Hot spots?

Yet ignoring poor integration and driver breakup is OK?

I have no idea what he's talking about specifically. Hot spot to me would equate to excess output (i.e. peak in response). I've not experienced this myself.

C

I assume he means that off-axis, there would be an excess of energy in the 1 KHz - 3 KHz range when crossing over low.

I haven't taken any measurements in that respect, so I really don't know. Other than LSPCad's predictions, I have never worked with off-axis measurements.

[fjhuerta]

Clearly, the dispersion pattern of a 1" dome tweeter is significantly different than that of a 6-7" midwoofer. What you really have is a tradeoff. As you move the crossover point down, the midwoofer beams less and behaves more like the dome tweeter. The downside is that the ear, in general, is more senstive to subtle changes at 1 kHz vs. something higher like 2-3 kHz. The best solution, for most people probably remains to operate your drivers within their ideal range in terms of distortion, frequency response, and off-axis response (which for something like the RS180 tops out around 1.4 kHz - then you need to choose a tweeter that can comfortably handle that.)

For MTM's, the choice is constrained somewhat more by lobing effects caused by interference between the two midwoofers playing over the same frequency range.

Thanks for your answer.

If I understood correctly, crossing over low means that off-axis dispersion is flatter than when crossing over high... I'd assume a 6" crossed over at, say, 4 KHz would suffer from "suck-outs" from 2 Khz and up.

If this is correct, and the ear is indeed more sensitive from 1KHz ~ 3 KHz, would it make sense to cross over low and use a bit of a BBC dip?

[cjd]

Last time I did measurements, lower crossover to the tweeter provided slightly flatter response over a broader off-axis range...

[fjhuerta]

Last time I did measurements, lower crossover to the tweeter provided slightly flatter response over a broader off-axis range...

So, maybe I'm used to hearing a suck-out in the presence region, and I'm still not used to flat frequency response. It could be...

[cjd]

Hard to say. It's easy to get generic off-axis measurements, and they can indeed tell quite a tale. A lot happens off-axis.

C

[Mudjock]

Thanks for your answer.

If I understood correctly, crossing over low means that off-axis dispersion is flatter than when crossing over high... I'd assume a 6" crossed over at, say, 4 KHz would suffer from "suck-outs" from 2 Khz and up.

If this is correct, and the ear is indeed more sensitive from 1KHz ~ 3 KHz, would it make sense to cross over low and use a bit of a BBC dip?

I think you understand correctly. The simplest example is to simply look at driver curves that have on and off axis frequency response. Larger drivers off axis response tends to fall off at lower frequency. Driver size is not the only factor, however. Voice coil inductance and driver geometry (cone and dustcap/phase plug) also play significant roles.

cjd has it right about off-axis response being significant. Many speakers do have a built-in BBC dip due to off axis nulls around the crossover frequency. Tweeter dispersion can vary quite a bit and significantly affect target frequency response shapes. Back before I had any measurement equipment, I designed a pair of MT's to have a flat on-axis response and they sounded incredibly dead (like I had ear plugs in). I had to take out about 4 ohms of tweeter padding to get them to sound balanced. I later confirmed with measurements that the tweeter response had about a 5 dB upslope, much like Zaph's more recent ZD1 design.

In theory if you had a low crossover point and your drivers were not quite in phase on axis, you might come to some point off axis where the drivers would sum to a peak if your dispersion was good enough that your response was still strong out there. A well-designed speaker probably shouldn't have this problem....