Is my Phase too fat?

[dcibel]

Interesting. That's kind of where I'm headed right now. I feel like I'm almost there but need to drop the tweeter a hair. Your directivity comment has me thinking now because currently one of the bonuses of my current XO is when I stand up in the listening position (~2m) the imaging is fantastic, extending behind the speakers almost 3-D. I haven't really experienced that before with previous builds. In sitting position the tweeter has just a little bit too much energy up top. Just a hair on some tracks (Cymbals, Aja for example).

Ah good, I didn't realize you had built this already and were listening. If that's the case, you can always apply some EQ, whether from a PC source or DSP or receiver, etc. to determine what needs to change in the crossover to make the sound perfect, but I would start in the 2.5-3.5kHz range. Too much energy here will make the speaker sound overly forward, and you'll only have to drop the level 1-2dB to make a world of difference. You can also try dropping the entire tweeter level a bit if you find its just a bit too much overall (cymbals, etc).

[Wushuliu]

You'll probably be happy with what you have, though you might consider adding an LCR filter to address the lump in the 2.5k to 3.5k region. The region just above the crossover freq will have a wide directivity compared to the frequencies below the crossover due to the transition from the woofer to tweeter. I find that lowering the level here often makes the sound more natural, basically you are compensating for the overall radiated energy (power response).

You will see something to this effect in Troels SBA-10:

Interesting. That's kind of where I'm headed right now. I feel like I'm almost there but need to drop the tweeter a hair. Your directivity comment has me thinking now because currently one of the bonuses of my current XO is when I stand up in the listening position (~2m) the imaging is fantastic, extending behind the speakers almost 3-D. I haven't really experienced that before with previous builds. In sitting position the tweeter has just a little bit too much energy up top. Just a hair on some tracks (Cymbals, Aja for example).

[dcibel]

You'll probably be happy with what you have, though you might consider adding an LCR filter to address the lump in the 2.5k to 3.5k region. The region just above the crossover freq will have a wide directivity compared to the frequencies below the crossover due to the transition from the woofer to tweeter. I find that lowering the level here often makes the sound more natural, basically you are compensating for the overall radiated energy (power response).

You will see something to this effect in Troels SBA-10:

[Wushuliu]

So a third order cap would be needed even though I'm already more or less at LR4 at 1.5khz and ~30db down at 600hz?

I just noticed Troels crossed his TW29 second order electric/LR4 Acoustic at 1.7khz (attached along with my XO response)...

[dlr]

I have used the original SB STC25 tweeter and now have the STC26 in its place. In my auditioning I tested it LR4 @1500Hz and higher. I detected no audible stress, but extended use might show a heat problem. I ended up using the LR4 @2K. It's part of my design hosted by Zaph. He provided the drivers and requested that I post a design. I took that a step further and documented the design and build fairly thoroughly. That was eight years ago!

The Chameleons

dlr

[isaeagle4031]

Agree, Jeff crossed the Kairos at 1.8k, 2nd order electrical.

[Wushuliu]

JeffB: "I've said this a number of times in discussions about my Kairos, but the Satori can cross lower naturally (without the aid of waveguide) than any other tweeter I have used"

Troels: "from my SBA-10 construction I learned the TW29RN can go really low without problems."

[isaeagle4031]

You see even the much cheaper Seas (and the Transducer Labs) has a lot less distortion at 1K.
The SB tweeters are some of the best but I would never cross them below 3K at 12 or more dB
per octave.

Have to disagree there. I've crossed the sb26stac at 1.8k, 3rd order electrical in a series network. No issues with distortion or otherwise. They have the ability easily.

[isaeagle4031]

Why? To protect the tweeter?

Yep.

[dcibel]

Here's the TW29 soft dome:
http://www.audioxpress.com/article/t...gh-end-tweeter

And here's the Be edition:
http://www.audioxpress.com/article/t...m-dome-tweeter

You have increasing distortion below 2kHz, but it is purely 2nd order. I wouldn't hesitate to use the TW29 with a 2nd order 2kHz crossover, 4th order acoustic rolloff, but no lower.

[Wushuliu]

Huh, I thought one of their advantages was that they could be crossed low.

Can't cross below 3k?

[xmax]

You see even the much cheaper Seas (and the Transducer Labs) has a lot less distortion at 1K.
The SB tweeters are some of the best but I would never cross them below 3K at 12 or more dB
per octave.

[xmax]

Which version TW29? None of them like to go that low, they have a natural roll off that starts around 4K,
but then a large peak at around 1500, even if you get decent measurement it will not sound very good.

Also what are you basing your cap/coil ratios on?

All the SB stuff can sound fantastic but the tweeters can get naughty below 2K if not very sharply
filtered.

[Wushuliu]

Understandable. As far as 3rd order, given the somewhat low xover point, I would use that as a minimum anyway

Why? To protect the tweeter?

[isaeagle4031]

Yeah, but I have the LCR parts on hand and just kept them there as part of the learning experience.

@decibel, will do.

Understandable. As far as 3rd order, given the somewhat low xover point, I would use that as a minimum anyway