2226, 2445 2405/Dayton AMT upgrade?

[danvprod]

Currently running a pair of 2226s in 4 cu ft enclosures tuned to 45 Hz, 2445s mated to 250 Hz Tractrix horns and 2405s with an active/passive setup. Running a MiniDSP, which lets me bi-amp the 2226 and 2445s. X-over is 500Hz 24/12.

2405s are passive 12 dB/octave at 8 kHz.

I've not really been happy with the handoff between the 2445s in the big horns and the 2405s.

I've been considering tri-amping the whole thing and running the dayton AMT ribbons (http://www.parts-express.com/dayton-audio-amtpro-4-air-motion-transformer-tweeter-4-ohm--275-094). The thought would be to run the 2445s from 500 Hz - 3500 Hz and have the AMTs take over after that.

Either get a MiniDSP 4x10 HD or something similar to do the x-over. Also run a 47 uF on the 2445 and a 20 uF on the AMT for protection.


Anyone have thoughts on this? Would the AMT be able to keep up with the 2445? (Home listening levels).

I've also recently changed from powering the 2445s with a pair of 300B amps to a 6-channel SS amp.

64589

[Lee in Montreal]

So, you run a 2-way active and a 3rd way to the tweeter passively. Am I right? If you don't like the transition from midrange to tweeter, maybe it is a problem with phase/timing or filter slope. If you run the midrange until it dies up there, and added a tweeter from 8KHZ, it might indeed sound funny.

Whay not use a 3-way active crossover? Maybe be w/ 12 or 18db LR slope

64592

[danvprod]

Hi Lee -- that indeed may be the good first step.

I've found that running the 2226 LR4 @ 500 Hz and the the 2445 LR2 @ 500 gives me an acoustic LR4 and really good integration between the two. My thought was just that I might be running the 250Hz horn too high?

I don't think I could run the 2405 much lower unless I am wrong? Maybe 6 kHz...

The 250 Hz horn and 2445 starts giving up the ghost at about 7 kHz.

I know you run your big horns all the way up. I forget what CD you are using.

[Lee in Montreal]

The 2405 should indeed be crossed at 8KHz

If crossed at 7KHz,something is wrong ;-)

On the big 2360 w/ 2445, I run a 12db correction (or is that 15db? Don't remember) of 6db/octave starting at 3.5kHz