Amp / Speaker Matching & Clipping

[warpfile]

Hey,

I am at a bit of a loss here, due to my lack of understanding of physics.
Sorry if I'm asking questions that have probably been asked & answered a 1000x on here. This seems to be a rather contested subject, and I want to make sure I'm getting some good answers for my specific questions rather than try to make sense of related topics. Having said that, I don't mind if you send me links to articles and such that answer these specific questions.

Thanks! thumbsup


My problem:

Normally I would set up a system so every piece of equipment runs at or slightly below 0 db. That way I can be pretty sure that if I don't clip the mixer / EQ at the FOH, nothing will clip down the path and I'm pretty much safe from any distortion or danger to the amps and speakers.

I would usually choose amps which have a RMS rating that is (1.5x-2x) higher than the RMS rating of the corresponding speakers, because I believe it gives me more headroom and thus is safer for the speakers overall and sounds cleaner. I usually turn the amp's input controls all the way up.


Now. If the amp severely overpowers the speakers (let's say a 5000W amp driving 500W speakers), what would be the smarter and healthier choice: turn up the input gain on the amp fully clockwise, and send less signal to the amps by turning down the mains on the mixer OR turn the gain controls on the amps down way low, while keeping the 0db gain structure in place before the amps? Or is it ultimately the same, because the speakers get sent the same "power", only attenuated at different stages?

Also if in this scenario I have the amp inputs turned down, and I produce clipping on the way to the amp (by running the mixer / EQ way too hot) will that harm the speakers, will it harm the input stage of the amp or none or both?

A 5000W amp with the inputs turned down low will still be able to deliver 5000W if requested by the speakers (in other words, the input controls don't really alter the output behavior), right?

[aussie_techie]

no matter where the problem is the answer is always to turn down the output of whatever processor you are using simply because of ease of set up and recall its one one thing that you dont have to turn up/down every time. its only the stage to use limiters if you need to have them.

[warpfile]

Thanks for the answer, I don't really think it answers all of my questions however.

By processor, you mean the EQ I'm running my mixer through for example?

BTW I guess I couldve explained it better, I'm talking about Live Sound Reinforcement here. But the same principles should apply anyhow.

[aussie_techie]

By processor i mean the speaker processor, crossover, drive rack, whatever you have.
You didnt say exactly what you have but as your talking 5000W i cant imagine not having a processor of some kind. then again I should of also said if you are using passive crossovers then turn down the FOH comp output as it will be your final stage of protection.

[warpfile]

OK, gotcha. thumbsup

Yeah, I totally forgot to mention those My 5000W example was a fictitious one, but sure enough the last rig I operated had a speaker processor running. I was just trying to make up an example with an amp that's way too powerful for the speakers.

So you're saying: attenuate the signal at the last stage (whatever it is) before the amps, and turn the amps input up all the way (=no attenuation) (which is basically what I "instinctively" do).

Is there any technical reason why that method is "better" than turning down the amp inputs?

[aussie_techie]

I figured 5000W was just an example, but anything getting up to that kind of range will be the same.

In live sound its one area that is less about being technically correct and more about ease of setup and recall and to a lesser extend matching each side of the amps/speakers.

If its a digital speaker processor then ideally you want a strong enough signal in and out of the processor to maximise the A/D and D/A converters but you have half a dozen amps its easier to turn them all to max and set a lower output in the processor. In being easier it also minimises the risks of damage as the amps them selves cant be turned up any more, any limiting in the processor remains an upper limit rather than something someone could get around by turning up the amps further. When looking at the gain staging of the entire system and depending on the amps you may choose to set a lower voltage gain while still keeping and front attenuation at '0' (assuming the switches arnt easily assessable for anyone to change).

[warpfile]

Ok! Good point about the converters, I hadn't really considered that. So from that standpoint it would (sonically) be pretty stupid to attenuate the signal before the processor. While it may be easier (for me personally) to do so (also it's probably not the "technically" best way to get the best possible SNR, I know).

OK, so let's say I set up the rig so with the main faders at unity, the signal leaves the mains at 0db, goes through EQ, compressors etc, and gets attenuated in the speaker processor. Everything runs smooth & safely. Limiters in the speaker processor keep the signal from running any hotter. I go eat some pizza and the guy that does the sound for Band X decides they need more volume, goes to the amps, sees they're turned up all the way. Is totally ignorant of speaker processors. Goes back to the mixer and jacks up the signal there. Clipping LEDs at the FOH blinking left and right.

Will that be harmful to the amps or speakers, or will it just result in a very nasty distorted but harmless sound? Like I said in my first post, I'm lacking some fundamental understanding, here I think I know that a truely clipping amp sents out up to twice it's normal wattage, but an amp that's merely amplifying an already clipped signal doesn't, right?

[warpfile]

(first and final bump)

Anyone else care to chip in / answer my stupid remaining newbie questions?