I'm interested in the Crown XLS 1500 suggestion. I can't see a bandwidth spec or anything around 100KHz for it though?

www.crownaudio.com/en/products/xls-1500


Another consideration / feature is that I could be able to feed 1 volume controlled output from Najda and then chain them with the output / link feature - Only useful if each amp has signal delay for time alignment and of course X/O capability. The Behrigner Inuke has.
Trawling the Crown XLS drivecore series blurb - it appears not! The XTi series does though (quoted as 50ms range) - bit more money of course!

Background to this is:

The Tapped horn subs are something like 9ms or so behind the rest of the system, due to horn length etc. So they are zero timed and the rest of my channels are delayed in DSP to be in alignment.
Alignment is done by measuring a signal (first peak positive) through each driver and delaying until they are aligned. Of course perfect alignment is only in one spot - where the mic is (ideal listening spot). But the coherency is vastly improved anywhere in the listening zone I have found. 

As Najda (my DSP ADC / Pre / DAC X/O), is only 4 channel I am currently using the upper channel feed - which is open (no upper x/o slope) feeding the Raal Ribbon Tweeters on their own amp via a passive X/O. Due to the short wave length at upper and tweeter frequencies, physical alignment is possible - it is fractions of a mm movement though to get it spot on.
With the bass amps only needing one feed I could put the tweeter on it's own Najda output and then physically move the tweeter forward / inline for less soundwave reflections...

ho hum! Glad I bought the Inuke DSP version now, but it is the mid bass amp that needs to the delay of course...

At sub frequencies alignment would require something like 10m+ physical difference:)

Yes the Crown Xti 1000 has it but!

"Delay: For signal alignment of driver; 5 mS total delay"

So, Xti 1002 and up then...

Class 'd' as usually cited for bass duties as they have a high power output for small psu size, neat and small design, very efficient around 82% + Duty cycle etc.

These are usually Chinese made, banged out by the thousands and do a great job of hanging onto bass drivers, big damping factors etc. Plenty of dynamics too.

Look at one under a decent scope or spectrum analyzer and you will 'see' a large amount of harsh sitting ontop of the sine wave form being sampled, though this is usually out of pass band and you will not hear it! (again depends on design parameters)

Most class 'd's tend top to have a switching frequency around 1/2 meg and conveniently tie these in with the switching point of the SMPS so you obtain an cancelling out effect. (Theoretically)

I find moving the switching frequency at which the amplifier 'works' much higher up generates far better texture and a more nature style of sound compared to these usual suspects they give, power, drive, detail, and clean sound. Though this does generate other issues that need addressing.

A simple low pass filter say 80Khz to 120Khz depending on your zobel network and feedback arrangements is all that's needed to bring the signal back into the analogue domain.

There are always exceptions to the rule.

For example some people feel the Devialet's first sliding 15/20W into class 'a' helps remove that 'switching amp' sound for myself it just means it takes a little longer to go digital hash lol though other opinions are valid.

Watch out for the Crown amps - most of them (if not all) are fan cooled - even the Class D ones. I cannot tell you how wonderful it is to have convection-cooled amps that are dead silent, after years of fan noise from Class A Krells, Audio Research valved power amps and (yup) a Crown to drive the sub.

Your not really going to tax a crown with 110dB horns I believe

Still feel the Crown is way OTT for your big horns, dynamic yes for sure, wall of sound quite possibly, but really a Caribbean gentleman speaker sound Steve?

Had you pegged for more subtlety, have you tried a Sugden Class 'a' power amplifier at all, they have the texture and richness just not the speed or drive, however 110 efficiency would cure that and then some!

What about using one of Colin's designs if I recall correctly they are around 20 watts, seriously would be more than plenty they are very fluid sounding too.

I used to use the trick of wiring 12V cooling fans in a PC with 5V to make them run virtually silently but still move the air around.

Any updates Steve

Just bought an NU3000 to try on my OB's.