RMS limiting and amp choices

[kipman725]

I'm running my PA at a party at the end of the month and it will be the first time I am using my new design:

https://www.diyaudio.com/forums/subwoofers/345807-isobaric-sub-based-pyle-plpw15d.html

TL:DR >115dB/2pi/1m continuous from 30Hz.  I have four of them which I will arrange in a coupled line configuration across the front of the stage.  This is indoors so I should have some boundary reinforcement.

Each sub is wired as 16 ohm nominal but there are the usual double peaks due to the reflex tuning.  The drivers have 2.5" coils on an aluminum former.  I never burnt a coil in the previous sealed boxes even when running into heavy limiting on nu6000dsp and having hot boxes (boxes wired for 16ohm).  The cooling is much improved in the new cabs.

During the power compression testing I see about 2dB compression in the excursion limited portions of the sweep applying 94V to the cab and about 1.5dB of compression at port tune.  As this was right on the limit of the amp I was using (nu6000dsp) I'm not sure if its actually less.

I have two amps available:

NU6000DSP, P7000S

In bridge mode the P7000S will do 200Vrms which would seem like it could be worthwhile as the cabs seemed to have a bit more to give at 94V.... however the amp is only rated at 8ohms bridged.  Although tantalizingly there is a short duration power rating for 4 ohm bridged  in the manual.

As per usual I need everything I can get from all my get from my gear so I am wondering what combination of amps would work best.  My current thinking is perhaps risking loading them all bridged onto the P7000s as actual impedance will (most of the time) be over 4ohms or running the center subs off the bridged P7000s and the outer ones off the inuke.

My other question is what to set my thermal limiting to, I think 300w per coil is reasonable as they have an aluminum former so should cool better than typical non conductive formers.  This gives 600W per cab and 97Vrms.  With my limiters I can set attack and hold times up to 3s I would guess I want a attack time less than the thermal time constant of the voice coil and the longest release time possible?