How to accurately set a brick limiter, to protect drivers?

As in the title, I'm trying to figure out how to properly determine the maximum voltage / watts, or whatever that my various drivers can handle, and then understand how I would use THOSE numbers to set my limiters. (For both subs & tops.)

I of course want the option of maximum safe volume, without burning up coils.

At a minimum I'd like to understand the basic steps typically taken for this. I'm pretty sure it has to do with voltage, yes? But that's where my knowledge ends.

I do own a signal generator, a dB meter, a volt meter (Fluke 87) and some RTA software that can also somehow function as an oscilloscope, though I have no idea yet how that part actually works.
================================================


More specific helps would be even better, of course, so FWIW:

My processor is an EAW UX3600, and both tops and subs are driven by Lab Gruppen C48:4 4 channel amps.

The amps have a few different settings for maximum voltage, but no fine-tuning. The highest three choices are 100v, 118v, and 141v. With a typical input gain, this equates to about 630w, 880w, and 1200w per channel. I believe that's at 4 ohms, (the manual doesn't specifiy) as that's the maximum 1/8 power at 4 ohms. Everything I'm running now is 8 ohms per channel, so MY GUESS is that these voltages equate to about 525w, 760w, and 1,000 at 8 ohms.
These amps also have protection circuits that hard-limit the output to those voltage settings.

The processor has brick-limiters for each output. The parameter of concern here is the threshold setting, which is in dBu, and defaults to 6 dBu.
---------------------------------

Again, if anyone has the time to respond with specifics, then here are my driver specs:

TOPS: EAW KF394, currently running in full-range mode (not bi-amped) and each rated at 1100w / 8 ohms.
I'm not using that amp's outputs 3 & 4 yet, but will later when I experiment with the KF394's bi-amp mode

Two diffferent subs types, depending on the venue:

A: Bag End sealed 18's. I run four of them, and use one amp channel for each, so at 8 ohms. They are each rated at 400w / 8 ohms

B: Two custom-made dual 12" ported boxes, each with two Eminence Kappalite 12" drivers. (Many thanks to Art W for helping me fine-tune this design.)
Again, I'm currently running each driver off of a separate amp channel, so 8 ohms each. They are each rated at 450w / 8 ohms.

And additionally - I may build a third 2X12" sub, for outdoor gigs. At that point, I'll probably change the subs' internal wiring to parallel / 4 ohms per cabinet, so I don't have to carry another amp with me. So I need to know how THAT affects maximun voltage and limiter settings as well.
=====================================================



ANY info on this at all would be massively helpful, even if just the basic concepts.

Thanks in advance, guys.
 
This is also harder to do than the simple math suggests because it is program material dependant, and with only peak limiters available you have to sacrifice some performance for genres with low crest factors as they will put more heat into the drivers quicker.
With my system and the music genres it sees most of the time the subs are the limiting factor, this actually simplifies things as you don't want the tops out running the subs because that makes the system sound thin, so all limiters should activate when the subs run out of gas. In my case my sub drivers are rated 700w continuous and I have 1 amp channel per driver capable of 1100w, and I have found that simply preventing amp from clipping is safe. With a bigger amp like yours you want to start conservative and work up, so start with the drivers continuous/program rating(2xRMS) and move up only if it looks like there is more untapped potential.
To calculate an output voltage knowing the power level and impedance you want the formula is V = Square Root(PxR).
 
Thanks, Conanski.

V = Square Root(PxR). - OMG, I knew there'd be math involved ! 😀

------------------------------

I'd still like to learn the technical end of this (maybe someone will elaborate on this) but otherwise your "seat of the pants" method seems reasonably safe. SO:

I know that the Bag End subs, when two are run parallel / at 4 ohms, can take 1600w, because I use to run them off a Lab Gruppen fP6400 set to that, and regularly slammed the snot out of it. (Lit up those red led's like it was christmas.) The surrounds occasional got little tears in them, but the voice coils never went, bless their little copper hearts.

I currently have that 4-channel subs amp set to 880w (again, probably more like 760w @ 8 ohms. - I think.) so that must be safe.
If I have the tops balanced, then raise the whole system until the subs amp's protection LED flickers, I guess that's where to set the limiters on both amps, yes? (Unless that blows the tops, of course. 😳 - which is mostly why i started this thread. ) Well, the tops do take 1100w, and those 4 amp channels max out at 1,000w / 8 ohms, so I guess I have nothing to worry about other than possibly overloading the tops-amp's input stage.

So this makes sense?
 
I know that the Bag End subs, when two are run parallel / at 4 ohms, can take 1600w, because I use to run them off a Lab Gruppen fP6400 set to that, and regularly slammed the snot out of it. (Lit up those red led's like it was christmas.) The surrounds occasional got little tears in them, but the voice coils never went, bless their little copper hearts.
Yeah 800w/driver sound about right, that equates to 80v at 8ohms according to the math.
I currently have that 4-channel subs amp set to 880w (again, probably more like 760w @ 8 ohms. - I think.) so that must be safe.
If I have the tops balanced, then raise the whole system until the subs amp's protection LED flickers, I guess that's where to set the limiters on both amps, yes?
Yes. I suspect once you have this setup the mids amp channels will never get close to full output, with the bass boost everybody likes to hear and the fact that mid hi cabs are usually more efficient than subs it doesn't take much power to produce balanced sound.