Technical question about power amps

[David Houck]

Since I am not technically inclined I do not at the moment have it clear in my mind how much power I'm actually sending to my speakers at any given point.
 
My power amp is a QSC PLX 2402.  It is rated at 700w per channel into 4 ohms.  The front panel gain controls are shown in db, the maximum setting being 32db.  The manual states that this maximum gain setting is also 40x.  What is the relationship between the db setting on the gain control and the amount of power going to the speakers?  What does the 40x refer to?
 
The front panel also has LED indicators showing Signal, -20db, -10db, and Clip.  The manual states that Signal indicates 0.1% of full power, -20db indicates 1%, and -10db indicates 10%.  With my Acme Low B2 I'm frequently lighting the -10db LED but never the Clip LED.  What is the relationship between these levels and the power going to the speakers?
 
Relatedly, what is the relationship between changes in signal level from the preamp and the amount of power going to the speakers?
 
With 700 watts per channel I want to make sure I don't fry some of my cabs that are rated significantly below that level.  Thus I need to have some level of understanding where to set the power amp's gain control for a given cabinet.

[musikill]

Dave,
the best thing you can do in a situation where you have more power available than your speakers cna handle is to put inline fuses in your speaker cables.  I would not be too concerned with how much juice you are sending as transient spikes cause all bets to be off anyway.  The fuses will protect your speakers and that is what you are really after.  I have the formula for figuring out what size fuse to get around here someplace and will post it when I find it.  I think it is in my old Crown DC 300a manual.  My preference has always been the Crown DC series.  the old ones are all hand built and are rugged as can be.

[David Houck]

Thanks Greg; actually the Acmes and Edens have built in light bulb type fuses, so I'm not too overly concerned.  (I also have a limiter in the rack to get a handle on those spikes.)  However, if the fuse goes in the middle of a hot set in front of a full house it can kind of change the mood a little.  So I feel I really need to improve my understanding of the situation.  I should probably do a web search; I imagine there's a good discussion of the subject out there somewhere.

[bigredbass]

While hardly an electronics wiz, I do know this:
 
Speakers are MUCH more likely to fail with a too-small amp pushed to clipping constantly than a seemingly too-large amp that is comfortably loafing along.
 
The current going down your cables from the amp to the speakers is AC: The +/- swings are directly driving the voice coils, etc., IN and OUT with this swing.  Just like God and JBL intended.
 
Push an amp to clipping and those moments when it's clipping are essentially DC: In those moments you're sending + or -, but NOT the swing between the two states.  The driver will push OUT. . . and stay there in these moments, absorbing this DC state as heat and stress, and you can see where I'm headed with this . . . BOOM!
 
I've seen amps where the output transistors failed and fed the speakers DC indefinitely . . . until the paper cones literally caught fire, the grill cloth went up, a real pyro show.
 
I've always read that the Acmes were relatively inefficient (not real loud relative to a lot of cabinets) and generally require a LOT of amp.  But they seem to be very high quality, so if it was me, I'd strap in and blow . . . because I doubt your cabinets will.  
 
J o e y

[bob]

Thanks, Joey, I never had a good understanding of what clipping really did to the speakers - despite the fact I'd just written up some other thoughts about it :-)
 
Greg, the fusing info would be great information to have. You've summed it up nicely, in that something like catching a fingernail could be disastrous, but as long as you don't actually do anything extreme, my experience has been that having an amp rated well above your cabinet's peak handling will be better for sound.
 
I'm not quite sure why this is, but I think that using a more powerful amp - for the same volume - usually gives you better current delivery, and probably better damping, resulting in a tighter and cleaner tone. In fact, to my ears it seems that all things being equal, you get better sound at the same volume by turning down your instrument or preamp and cranking the amp (assuming it's clean enough). So my preference is to have an amp rated above the cabinet peak (by as much as 50-100%), and either live dangerously, be careful, or maybe fuse it somehow.
 
(I should note that I play mostly jazz, fingerstyle, not going for killer volume. If I were slapping or otherwise looking for huge dynamic spikes, I might feel differently.)
 
Dave, amplifier power is usually measured as how much the thing puts out just before it starts to clip. Clipping is bad, because it distorts (sort of chops off or flattens out) the tops and bottoms of the usually graceful waveform, and for various reasons, speakers don't seem to like this. However, if you only drive it to clipping infrequently, and not grossly so, then you can usually get away with it.
 
I'm not sure how to quantify this, but as sort of an order of magnitude thing, if you play for half an hour and see the clip LED flash briefly half a dozen times or so, and you don't hear anything really gross - and you amp and speakers continue to function - then you're probably fine. A little clipping is okay, you just don't want to make it a habit.
 
But having said that, it's not just an issue of whether you drive the amp to clipping a lot. You might also have problems if the amp is consistently putting out more current than the speakers can handle.
 
Most of the time, you don't use anywhere near the full rated power of your amp. And speakers are generally rated for both continuous and peak capacity (though I don't see that for the Acme's), because it's okay to push them hard for a while, as long as you give them some chances to cool off a little (literally, a heat thing).  
 
I downloaded the QSC manual (these things look pretty nice, I can see why they're so popular). My interpretation is that if you have the front panel gain control maxed out (40x, giving 32 dB gain) the 2402 can put 700 watts per channel into 4 ohms, just before the little red light comes on.
 
If you were to play exactly the same (intense enough to clip at full gain), and had the front panel gain turned down to half, then I'm pretty confident the amp would not put out more than 350 watts (and it wouldn't be anywhere close to clipping either).
 
Unless you drive the amp to clipping, with the gain turned all the way up, you aren't getting anywhere near the stated maximum power. There is definitely a log function related to sound level in here somewhere, and as you've noted, the -10db indicator means you're only using 10%. I'm pretty sure this would translate into a mere 70 watts.
 
Regarding your question about preamp level, you can sort of lump a whole bunch of things into this category (everything contributing to gain before the signal gets to the amp): preamp volume, instrument volume, even the trim pots on your pickup preamps. Obviously, this stuff has an effect on overall volume, but you don't really have to think about it too much - just watch those little LEDs on your amp, because they are showing you what the final output level is, going to your speakers.
 
Look, I wish I knew more about this myself, and I'm definitely not an expert, but I think you're going to have a hard time finding practical answers by searching the web (maybe, or better yet, someone here could straighten us out).
 
My suggestion would be that you take an experimental approach. Play for a while, at home if your neighbors permit it, or at your next practice, and carefully watch the LEDs as you play louder and more aggressively. See just how much it takes to get the clip indicator to flash. My guess is that if you're as worried (i.e. conservative) as you sound, you probably have way more headroom than you imagine. But take along a spare light bulb, just in case :-)
 
Your amp is only twice the continuous rating of your cabinet (suggesting it can probably handle higher peaks), and I think you're going to have a hard time driving it that hard even for brief moments. If you normally get by with the amp gain turned down a little, then you're almost certainly safe, and you might as well enjoy it.
 
Oh, one other thing: I see that there is a clipping control on the QSC, which looks interesting but I don't fully understand how it works, or what their warning about it really means. But if you're already using a limiter earlier in your chain, then I would probably leave this thing turned off. Use one or the other if you're worried (or just like the limiter), but both would likely be overkill, and probably degrade your sound.

[David Houck]

Thanks Joey and Bob; where do I begin ?  There is a lot of useful information there; and Bob thanks for going to the trouble of downloading and reading the manual!  (One of the appealing qualities for me of the QSC was the high damping factor; I tend to think that a high damping factor will give a clearer more musical tone on the low end where an amp with a low damping factor has a harder time keeping up.  And the speakers will live longer too.)
Well, I guess I'm more concerned about the Edens and a Bag End I use on smaller gigs than I am about the Acme.  I employ a wide dynamic range in my right hand technique including slap style with thumb attacks on the G string (which still needs a lot of practice ).  With the Edens rated at 350 per channel it seems logical to set the QSC gain control at half power.
 
Bob; you said the following:
 
My interpretation is that if you have the front panel gain control maxed out (40x, giving 32 dB gain) the 2402 can put 700 watts per channel into 4 ohms, just before the little red light comes on.  If you were to play exactly the same (intense enough to clip at full gain), and had the front panel gain turned down to half, then I'm pretty confident the amp would not put out more than 350 watts (and it wouldn't be anywhere close to clipping either).  Unless you drive the amp to clipping, with the gain turned all the way up, you aren't getting anywhere near the stated maximum power. There is definitely a log function related to sound level in here somewhere, and as you've noted, the -10db indicator means you're only using 10%. I'm pretty sure this would translate into a mere 70 watts.
 
You've said a lot of interesting things here.  With my limited and confused understanding I tend to think of db as being exponential rather than linear.  So what does it mean to turn the front panel gain control down to half?  If full is 32db is half 16 or some number greater than 16?
 
Ok; I just played a riff using an aggressive thumb style on the low end of the neck.  With the gain set at 32db I had the -10db light lit with every note but never lit the red clip light.  I then turned down to 22db, played the same riff, and never lit the -10db light.  If the -10db means you are only using 10% of full power, or 70 watts, then if that light is never lit it would seem to indicate that at a gain level of 22db I'm using less than 70 watts; far below the 350 the Eden should be able to handle.
 
Last weekend at a gig the drummer and one of the two guitar players were both asking me to turn up (and it's a rare occasion when the guitar player asks me to turn up ); but I felt like I was at the limit of what was safe.  But now I'm thinking I probably had a good amount of room left.
 
So where is halfway?  I'm starting to think it's some number greater than 22db, and that there must be some curve that should provide the answer.  Does the 3db increase equals 2x the power formula apply here?  If so then working backwards if 32db = 700 watts, 29db = 350 watts, 26 = 175, etc.  And 23db would equal 87.5 watts, which seems to fall in line with the observation above that at an output setting of 22db the -10db light was not lit suggesting a power level of less than 70 watts (whereas in the same experiment a setting of 23db did occasionally light the -10db light).  It doesn't seem sensible that I could set the output level of this amp at 29, when the max setting is 32, and safely run it to my Eden 350 watt cab.  Plus working backwards is probably the backwards way to approach the curve.  But the backwards curve does seem to come close mathmatically; 8db = 2.73w, 5 = 1.37, 2 = 0.68.  So the real curve must be something close.
 
(Bob; on the QSC's clip limiter.  Since it only comes on during actual clipping, if I'm never lighting the clip light I'm guessing the clip limiter never comes on.)
 
Well, I've quite confused myself and I'm starting to get concerned about the fact I'm not getting any work done and that I haven't had lunch .  I still don't yet feel comfortable setting the output to the Eden at 29db or even 26db; but shouldn't I?
 
Thanks again for the help.

[musikill]

Bob,
You are quite right about running the power amp full while keeping the preamp low.  This produces a much cleaner sound.  One thing to keep in mind (and I believe this is true for all power amps)is that the level controls on the power amp are input gain only.  The power amp is always delivering full capability - you regulate the output by regulating the input.  Clipping is an overload condition of the input - your description of the affects of clipping are spot on. I'll get the fuse formula posted; I need to look for it somemore.

[bigredbass]

dave:
 
Another clue to all of this is in those damn specs,
 
Somewhere, QSC will list a spec entitled something like 'input required for rated output'.
It should read somewhere around 1 volt or so at a low impedance to deliver that 700w at 4 ohms.
 
Your preamp should have a spec sheet as well listing its output (which will probably be, guess what, around 1 volt at  a low impedance), so will your limiter, etc.  If you compare these outputs to the inputs of the following components as you go down your signal chain, you will probably see that you're in the ballpark down the line.  
 
If you know a very good sound reinforcement guy, they are used to long signal chains where everything MUST remain clip-less.  Or I would imagine Q S C would gladly walk you through this.
 
You know, Bob, the AC nature of amp output was one of those things I'd never realised until it was pointed out to me.  Later, I was at the CREST website, and ran across a page for a 5kw per channel amp that they build for massive concert subwoofer arrays that put out 125 VOLTS AC at full output.  Zalytron doesn't sell an banana to AC plug adapter, though . . . .
 
J o e y

[David Houck]

Greg; for some reason I did not focus on whether the gain controls regulated input or output.  That significantly alters my previous thought process.  I guess I need to understand more about how much voltage I'm inputing to the amp.  So the 32db setting on the gain control is not so much related to the 700 watts of output as it is to the measurement of the increase in voltage input?

[David Houck]

Thanks Joey; the QSC manual states that the input sensitivity at full rated power at 4 ohms is 1.3v(+4.6dBu).  The Alembic F-1X operating guide does not indicate it's output.  All my other components, including the limiter, are in the effects loop; which compounds my not understanding the effect on output to the power amp .

[bob]

Sometime after the first cup of coffee this morning, I felt a bit uncomfortable about a few things I said late last night (I should know better...). Also found a little note I wrote to myself just before turning in that says 3db = 2x watts.
 
Musikill caught one of my mistakes (thanks): turning down the level control on the power amp does *not* limit its maximum output. I found some stuff on the Crown site which says it is best to think of them as input attenuators. Ignoring where exactly you set the dial, if you dropped the amp gain to half, and then doubled the preamp gain, you'd be right back where you started, with the same amount of power going to the speakers. So you shouldn't think of it as a protective move.
 
Dave, I think you're also right that you might as well leave the clipping protection turned on. Most of these are pretty well designed nowadays, so that they don't really affect the signal when they haven't kicked in.
 
As to whether the gain control regulate input or output: I think you're thinking too hard, and you might hurt yourself . Most likely, it attenuates the input signal before running it through the main gain stage, but however it's implemented, turning down the knob will result in less power coming out the back, for the same input signal, and that's all that really matters.
 
Unfortunately, I can't help you with the math - this confuses the hell out of me, too. Just by looking at the numbers on your level control (and mine), it's clearly not linear. It seems to me the rule of 2x power = 3db should apply, and it seems to fit pretty well. But I think you have to be careful about whether you're talking just about the power function in the amp, vs. the net effect on the final SPL - I'm not sure that's so simple.
 
Again, turning down the amp gain will not prevent you from driving it to maximum output, by playing harder, cranking the preamp, whatever. But as your experiments confirm, you are usually using far less power than you imagine.
 
Maybe the best way to put your mind at ease is to check with Acme on the peak handling capacity - I only saw continuous ratings, and many cabinets specify double that for peak. There's almost certainly a considerable margin, maybe all that you need.

[David Houck]

Unlike other speaker manufacturers, Acme recommends driving their speakers beyond the continuous ratings.  For instance, for the B-2 the rating is 350 watts RMS continuous.  But here's what Acme says about the cab, 350W RMS is a good starting point due to the relatively low efficiency of the cabinet. 500W RMS is ideal. The more power, the better the sound. (Within reason - It's only a 2x10!).
 
Acme doesn't publish a maximum number for reasons that are shown here:
http://www.acmebass.com/products/sensitivity.html
 
As I mentioned earlier, I'm actually more concerned for the Eden than the Acme.  I agree with your observation that the experiments confirmed that I was using less power that I thought.
 
At this point I now know a lot more about how much I don't know!

[dnburgess]

Further on Joey's point about the level controls on power amps controling input rather tha output level, Andy Lewis writes:
 
A little bit of information about the level controls on power amps:
We've had unfortunate conversations with an alarming number of people who have mode the some mistake:
 
A great many people, I've learned, believe that a 1000 watt amplifier becomes a 500 watt amplifier once the level controls are set to halfway up, 12 o'clock?, or only at 5, if you catch my meaning.
 
I would prefer not to have this conversation any more, because it usually is with some well-intentioned kid whose woofers are in tatters, and who can't understand what he did wrong.
 
So, read this twice if you have to: Lowering the setting of the Level, or Volume control on your power amplifier does not limit it's ability to produce it's full power! If a 1000 watt amplifier is only turned halfway up, it does not become a 500 watt amplifier. It is still able to produce it's full 1000 watts if it gets a ?hot enough input signal.

[David Houck]

Thanks David; and I read it twice .  (For those of you who don't know, Andy Lewis makes Acme speakers.)  So, one thing I've learned today is that focusing on the amp's gain control without regard to the strength of the signal being input to the amp will not achieve the goal of limiting power to the speakers at a specific level.  But what I have been unable to determine is how to effectively measure the input signal and then combine that with the gain control to find the desired maximum comfort level.
 
I'm wondering; I suppose I can take a voltmeter and measure the voltage at the output of the F1-X at certain settings while playing in an aggressive manner.  How would I then use this information in conjunction with the db settings on the amp's gain control to determine the power beings sent to the speakers?

[musikill]

This has turned into a mind bending discussion.  So much math YIKES!  It seems that many of you use the QSC amps that have some sort of level indicator on the knob graphic.  I'm not sure what the advantage is in that.  I use a few old Crown DC300a power amps and they have no level indication at all.  One DC300a is a series II and it has clipping lights but the other 2 DC300a's have nothing to indicate you are about to fry yourself into oblivion.  I have only fried 1 set of speakers - some JBL's that were in an old Dual Showman.  I didn't have a limiter/compressor at the time. If you have a limiter you probably are going to be fine.  Frankly, you will hear that something is not right if you are playing at levels that threaten your speakers.  You will hear the distortion and if you lower your volume you will be fine.  Unless of course you are pumping 700w into a speaker rated at 100w, then you will see instant results - like fire.  You're probably more safe than you think you are unless you are playing at very high SPL levels.  My suggestion: don't worry about it - if your ears are bleeding and your speakers are still in one piece -what more could you ask for?
greg