'Very' OB Midrange Baffle Width Study

[cuibono]

So, some measurements.

I EQ'd the raw driver response a bit, getting it approximately flat. I was a little surprised by the off axis measurements, not being better than the TB W4-656sc, so I dug a little deeper. What I found is inline with what StigErik has been getting into - that the mounting method matters (I haven't tried his suspended driver method yet, but I think the following supports his position).

First, a graph and photo of the original On/Off axis FR and mounting method. Note there is 3dB/division vertically, and the graphs are 0, 30, 45, 60 and 90degs.





Now, look at the improvement with a more minimal mounting method (particularly at wide angles off axis):







Overall, the new TB driver (W4-1320sj) looks good. Pretty much comparable on and off axis behavior (with the more minimal mounting method), definitely lower distortion (measured, not necessarily audible), but still a little more expensive. I listened to a single driver in mono (with a healthy dose of EQ) for several hours last night - actually not bad! No buzzing, which is nice. My only concern is the drivers low sensitivity combined with its low power handling - I may burn it at some point. We'll see - if that happens, I go back to the Alpha6a, or with a 4-way...

Enjoy!

[keyser]

At what side did you do the measurements with the initial mounting? On the left there is an extra wooden bar against which there might be a reflection. I'm am curious what exactly makes the difference here.

[cuibono]

Hi Keyser,

I did the off axis measurements with the extra wooden piece turning away from the mic (if that makes sense). Either way, the less stuff behind the cone, the better.

[cuibono]

SO! Here is what I've been trying to get at for some time.



This is the Neo3PDR tweeter and TB W4-1320sj, fully EQ'd. The lines represent 0, 30, 45 and 60deg off axis. I've set the listening axis at 30deg (the brown line), hence it is flattest. The XO is LR4 (acoustic) at 1700Hz. Ignore the regions below 250Hz and above 20k - they are invalid.

The goal of this thread has been to get the smoothest transition between mid and tweeter possible. I feel like I'm there.

Any comments?

PS - here is the individual drivers and then combined (at 30deg):



Soundeasy really makes it easy to achieve target functions...

[gainphile]

Cuibono,

Thanks for sharing your TB investigation ! I happen to have a pair of 3" TB so your experience has given me something to do on the weekend !!

You seem to cross to Neo at about 1.6kHz. Is this not straining that tweeter?

Cheers.

[keyser]

This looks near perfect! I think I'll reconsider the AL170. However, power handling probably is an issue with this TB. How are you planning to integrate it to a woofer?

[brucemck2]

Can you post the EQ settings you used? How much EQ is involved?

[cuibono]

Thanks everyone! I'm glad you enjoyed seeing it.

Gainphile - I've been using the Neo3 XO'd at 1700Hz since I've had them - I haven't noticed any strain yet. We'll see though (I'm more worried about the mid).

Keyser - Yes, there is a woofer involved. I XO between woofer and mid at 275Hz, LR4. The woofer is my own design, which there is a thread here somewhere ("another sensitive dipole woofer" or something like that)

Bruce - I can post them, but there is a bit of EQ involved. I use a computer and some studio software, but anyone else can do the same. I'll have to give a good write up on it, it took me a while to develop the software/hardware end of things. One thing though, is that even if someone replicates what I've done exactly, it is probably necessary to do final in-room measurements to get things really tight.

[Rudolf]

Beg your pardon, cuibono

but since everybody is showing his dipole tweeter measurements in your thread, I dare to throw in another one:



I placed two Dayton ND20FA-6 back to back on a small baffle (8 x 4.5 x 1.2 cm) and got this horizontal response (0, 15, 30, 45, 60, 75, 90 deg), measured at 45 cm distance (not calibrated):



There is no EQ involved apart from a 12 dB/oct HP at 2.5 kHz. For sure the vertical response is a mess and this botch will not find its way into 'production'. But it shows how mounting two very small tweeters very tight together can take you a loong way.

[cuibono]

Thanks, those do look really good! I hadn't thought of mounting them that way, I guess because I think the figure of 8 pattern would be pointing at the floor somewhat. But tell us how it sounds, it looks really good so far!

I like your 'turntable' too.