How Many Watts Available for Short Peak dB?

More like, the denon or pioneer might do the 100 watts briefly.

Dynamic capability, as it's sometimes called, is mainly due (IMO,) to the capacitors.

Note that the total demand during movie playback involves multiple channels, so it might be hard to get an exact idea.

In theory 3 dB of headroom gives you 3 dB more SPL on peaks. On paper that does not seem that big of a deal, but it seems nice to have.

If a 100 watt amp had 3 dB of headroom, I would think that implies 200 watts of short term power.

I don't see how an typical AVR can state that for more than 2 channels. I don't think they typically have much in terms of filter capacitance.

I am actually interested in the exact electrical scenario when you become power starved in an AVR. Do you run up against the limiter circuits, or run into the situation where the supply voltage drops a lot? Does the power supply drop because the filter caps become discharged enough to not be able to hold the peak secondary voltage? Or does the secondary voltage itself drop under heavy load? I have never found an explanation of what happens.

So I can say is you will eventually run out of power and the sound will suffer when that causes a high enough distortion from clipping.

p.s. you asked about the XPA-3? I own one. It's been trouble free. I drive my LCR speakers with it. I have no idea of how loud it will go, because I am not about expose my ears and neighbors to that level. It's got something like an 800 VA transformer with lots of capacitors. I will just assume it can handle anything I will ever throw at it (my reason for buying it, but I bought it for less than $600, shipped, so it felt like a no brainer then.)

I also threw in a discounted Crown XLS 1000 into the system. Just in case.

For me, it's about toys and not having to question whether I have enough power. I built my system over time, so it was not some huge bill at once. I can't even imagine putting together what I have now on one bill. I would guess over $5000 in speakers, receiver and amps.

My problem is that I am not in love with the speakers at all times. I plan on looking into different speakers, but only had these for two years, so we will see.

Well, seeing how I don't trust any power ratings, I would rather not worry about it. If your system does not sound good at some volume, reduce the volume.

If you say, bought some cheap AVR that was rated 100x7 watts, it seems likely it can't manage 100 watts on peaks in loud scenes in a movie. A guess. Based on what I suspect happens. I suspect there's too much load on the receiver during the loud scenes, and the cheap AVR won't get it done. But how can we know precisely what is going to happen?

I am kind of a fan of the upper end of receiver lines. Having played around with quite a few receivers, the cheap ones just don't seem to get it done at SPLs I am likely to use at home.

But say you buy something starting around $600 MSRP. Seems you are getting a better power supply, and it's more likely to not impede your enjoyment.

Just one man's opinion.

You can get kind of an idea of different models by reading a lot of bench tests. You will see the cheap models really don't do well at all.

We can apply some logic. A cheap AVR might weight 18 pounds. We know that a lot of the weight of an AVR is from the transformer. You heft one of those things, then look inside through the vents, and you see this small transformer. And you might suspect that the small transformer will limit the AVRs power.

Or you might pick up a more expensive AVR weight 25 pounds. The transformer looks more promising.

Not science, I know, but unless it's a class D amp based model, you can take some educated guesses. (I know some people will disagree.)

Caps only supply filtering/smoothing from the power transformer, thereby removing the ripple (or noise) that occurs from the rectification of the AC supply. You would need literally whole Farads of capacitors to form some kind of short term supply. But that could be supplied from the transformer anyway. In any case, the transformer is almost always the limiting factor in a power amplifier. Electrical supply travels just as fast whether its being discharged out of a capacitor or transformer.

Even 166,000uF (I think the most that's found it's way into a receiver, the Denon 5805) will give you no more than 19.92 watts of power by itself. It's not something you can use to generate hundreds of watts.

I would guess better receivers hold sufficient capacitance to supply some dynamic needs.

I noticed large spikes when I had a wattmeter connected with hard drum hits. The cap is drained, causing a high temporary draw from the wall to refill the caps. There's a delay between the hits, and the spike (could be the meter though on the delay.)

In my estimate, that shows how it's going to work....

That's not to say all receivers are going to be able to have any noticeable dynamic power.

I don't have time to work out the math though, and I admit I could be wrong. If I was not so busy with my new job, I would go research the topic.

1 farad = 1 coulomb / volt
1 coulomb = 1 amp/second

10,000 uf is .01 farads?
Does 60 volts give you .6 coulombs?
.6 amp seconds?
Does that mean for one second we have 36 watts?
.6 * 60 volts = 36 watts?

I am not sure any of this is correct. But the power is delivered not over a second, but over a much shorter time. I believe that should increase power (which is energy / time.)

[@MJH] - You're kinda taking the transformer out of the equation there, the capacitor and supply transformer are coupled to the one supply, they will discharge into the load as required. The cap's are able to keep the supply voltage above the RMS voltage of the transformer, as they will charge to the peak voltage during the peak's of the Rectified AC waveform. So they discharge into the load when the rectified waveform is not at its peak. On a 60hz supply, the caps will be charged at 120hz. If the caps fully discharged during a transient of less than 1ms, the supply voltage would fall back to the rectified AC and become noisy, unless this occured during a AC peak.

Here's a great pic I borrowed, the black lines are the transformer supply, and the capacitor is the red lines.


Don't forget that a transformer is effectively a large inductor as well, which stores electric current in a magnetic field around the coil of wire. This will also discharge into the load as quickly as a capacitor will.

Yes, I do know a transformer is part of the power supply And the purpose of the caps in the power supply

I also know that caps hold energy. And that energy can can obviously be released under load. Your power supply voltage is therefore defined by a complex relationship between the filter caps, the secondary voltage at the transformer, and the load the power supply (and I know there's a non trivial charging cycle with the capacitors.)

I can't give exact math, not an EE. I just know what I have read. That large cap banks are used to help supply power during periods of high demand. They certainly don't need to be anywhere near as large as they are in some amps if they are being used simply to smooth out the rectified secondary voltage from the transformer.

For a while, Stereophile included burst testing in their amp tests. They don't anymore AFAIK. The short answer that I'd give from looking at a dozen or so such teste is that every amp is different. Unless somebody has measured the amp's burst capability, you cannot guess it from other specs, AFAIK.

Home Theater includes 1ch peak ratings, on the higher end AVRs they usually beat the ACD results (and remember that ACD never happens in real life). Remember though, 100W vs 150W is <3 dB; it's probably not audible. With your proposed scenario, we're talking about 98.7 dB vs 100 dB (all theoretical). You're unlikely to put the amplifier to "max" in either scenario.

Anyways, if you're really worried about it, get a bigger amplifier. 200-300wpc will do no harm. I'm skeptical you'll ever see that much power coming out though (the XPA-3 or XPA-5, for example). Alternately, are more efficient speakers an option?

Interesting, to me, is whether the ability to cleanly reproduce the very leading edge transients is audible. There was a semi big blow up in the bass guitar community over an amp manufacturer whose 450 and 750 watt amps both actually put out a max of under 240 watts. "But they sound just as loud" because of limiting and compression applied ahead of the amplifier stage. Looking into the tech, and from things I've seen posted various places, the initial transient of a plucked note may be 12 or even 20 dB above the level of the actual note's tone. These transients last one half a cycle, so something like a twentieth of a second. If they're clipped or slightly compressed, how much can we notice?

BTW, seems to me there were more than a few of the Stereophile tested systems with less than a dB of headroom for short bursts, but several with at least 3 dB, and perhaps more, I just can't recall with clarity.

Unfortunately for those to whom it matters, asking how much burst power amps have is like asking how much RMS power amps have. Depends on the amp.

You guys can play with yourselfs over the numbers. I can tell you for sure that the low end light weight denons are not going to sound any better no matter what number for peak power you come up with.

URL is a dead link.

I'm very skeptical that this needs as much attention as you're giving it. It doesn't appear to be a "problem" at all.

The kinds of peaks you're looking for are theoretical at this point, you can go off and figure out some transient from a given movie to get a better idea of the actual demands, my guess is that you won't see any of these very brief peaks that require 40 dB (or some equally insane value) of momentary gain.

And there is no "RMS watts" - there's power. I think what you're looking for is either dynamic power (and there are ratings for this, not all manufacturers publish them, who knows how honest or even comparable they are), or PMPO (which is a magical value created by putting numbers on a page). Both are fluffy, "feel-good" values in most cases.

See here:
http://www.audioholics.com/education...-power-ratings

Here's an example of an amplifier that gives dynamic power ratings:
http://www.integrahometheater.com/mo...=Amplifier&p=s
And here it is measured:
http://www.hometheater.com/content/i...-labs-measures

It meets the output specs (150x2), so I'm primed to believe that it may deliver its dynamic claims into a single channel. But the question then becomes, under what conditions? And does it matter? Aside from being able to stamp "400W/ch" on the box, what's the point of those numbers?

Point is, if you're designing a hypothetical theater room and you want 100 dB at the listening position, you need speakers and amplification capable of putting 100 dB at the listening position; doesn't matter if you only "need" it for 2 seconds or if it's going to be "needed" for 2 hours. We can sit and invent PMPO values for your amplifier all you want, but it's a very touchy state of affairs (especially because now we're getting into "oh well, if the peak is Xms it won't be okay" and all of that since you've introduced the hedge words "momentary" and "dynamic"). It doesn't need to be complicated.

Generally there isn't a material advantage to a "bigger" amplifier, unless the current amplifier is unsuited to the load it's trying to drive. That's also part of my point. If your current 100W/ch amplifier isn't hacking it, 150W will not do it, 200W might do it, but 1000W is where you'd need to look "next." Most speakers won't survive 1000W. If we're talking about theoretical 20-40 dB peaks, well, the numbers just got a lot bigger. My normal "rule of thumb" is that you won't need more than a few wpc in a domestic setting; this is usually true. With your theoretical setup of 90 dB/W/M speakers, and a 12 ft seating distance, a 100wpc amplifier is perfectly suitable. Before we argue about "can we hear 20ms or 50ms" lets get back to "is 98.7 dB SPL vs 100.00 dB SPL really that big of a difference?"

The hearing damage question is interesting as well - it doesn't matter how you characterize the sound, loud is loud, and can damage your hearing with sufficient exposure. This does not mean your 100 dB peak will damage your hearing, but it does not mean that it "doesn't matter because it isn't the right kind of loud." (I've seen this argument perpetuated in the past, that OSHA/CCOHS data don't matter because "we aren't dealing with industrial machines" or whatever).

millisecond burst wattage available from an amp is less important than the speakers' ability to continue to respond linearly across their frequency range at higher levels. You will not see specs for that. AFAIK, it's safe to assume that speaker power input levels are thermal, and that compression and distortion occur at significantly lower levels.

Bottom line is you need equipment that can actually hit the levels within its linear range, and frankly a little extra whatever from an amp's burst power is almost certainly the last thing you'll care about. Seriously, I don't know that people could hear the difference between a 5 msec transient at 100 dB or at 103 dB.

I promise you that no commercial recording in the world captures a 12 dB transient from an electric bass. Back in analog tape days, you'd be monitoring to get as hot a signal as you could. I'm not sure the typical VU meter can respond to a 5ms transient, and anyway nobody's going to leave all that headroom up there to foul up the signal to noise ratio. Turn it up and the tape, via saturation (and maybe gracefully clipping the input stage), magically limits the peak for you. Voila. Modernly, you'd control those peaks, sooner or later, with compression/limiting (likely when recording) because it's easier to get a good mix when you hold them down. Taking 6-10 dB off the top is just no big deal, and even with relatively mundane equipment is fairly to entirely transparent.

After all, it's not that the transients are gone, they're just not as much louder than the rest of the sound as they would be if not limited. I mean, if you can catch the imaging from modern popular music, which tends to be mastered with a peak to average of 3 dB or less, you know that it's not the relative volume of the transient that gives you the ability to localize the sound.

Aha! Thank you.


Depending on *what* we're listening to, it can be much lower. I've seen 20 dB stated as the guideline for Dolby/THX reference, but I've also seen music material with only a few dB. But I think this is relatively fair. 20 dB is not that insane though, that's not kW level unless you've got the most inefficient speakers known to man, or are out in a big field/stadium.


Again, there is no RMS. There's just power. The continuous power handling rating on a speaker has to do with mechanical and/or thermal limits - the peak input is either a mechanical limit or a thermal limit (you can either put the driver into over-excursion which risks damage, or you can overheat the driver, which risks damage; depending on the design and the setup either is plausible).

It can't handle the higher output over a long period because you've exceeded the limit of the device. Those values are stamped on packaging just like "Dynamic Power" ratings on amplifiers - remember that 400W looks a lot nicer than 100W, and when you're competing with the 10,000W HTIB solutions you've gotta milk things a bit (remember that for the vast majority of customers, "watts" is the metric by which all quality and performance of any audio component is judged).

Burst/dynamic/momentary power is a worthless measure because it's so relative. It also isn't something you want to "plan in" - like I said earlier, if you want X you need something that can do X continuously. These "temporary" values are there to look nice on the box and assure the customer that the equipment has some tolerance before it fails. The same is true of anything else (you used cars and guns as examples, it applies there too) - you don't intentionally over-load an apparatus as part of your design.




As long as you're not running whatever amplifier into clipping, the distortion it's producing is going to be very low relative to other problems (like the speakers, which produce substantially more distortion even when they're not being over-extended).

It's a very interesting topic, I will give you that. I'm just not convinced that it "matters" in the sense that not enough data is (or likely will ever be) available, and that even if data were available, it would go along the lines of "this is a hail-mary kind of scenario" in terms of output capabilities. Remember that you're asking about dynamic/peak power as a way to address transients in cinema or movies. What happens when there's lots of transients right after each other?* There's just too many variables to contend with, in order to want to consider this in planning out a system. It's easier to just assume that your peaks are your maximum continuous output, and be done with it.


* What I'm thinking of here would be higher tempo action movies, where you'll have a lot of "rapid-fire" peaks during some scenes. In some cases those peaks occur very quickly, one after another, and most meters won't even "fall" during the sequence (things like explosions or other bass-heavy effects). There's music that does the same thing. Is this still a situation for "dynamic power" to save the day? Perhaps I'm getting off-topic, I'm just seeing this as a "what about a situation where an amplifier that would normally be lighting the "clip" light up would have more power for a half second" and thinking "yeah but usually in those situations, the "clip" light will turn on and stay on until you (or the program material) bring the levels back down."

When you talk about handling the dynamic range of movies, i think we are talking about buying the proper receiver/amp based on it's rated power not it's ability to handle short term peaks.

When they do rate amps for dynamic power, it's usually something in the general range of 3 dB.

Dynamic peaks in movies are much higher than 3 dB. Because quite scenes in movies are pretty quiet, in comparison. A scene may just involve someone speaking with no background music, and then something loud might happen.

IMO, when looking at movies, all you can do is to use some guesswork. Figure out the peak SPL you want, and use some power calculators (you will find them online,) to get some idea of how much power is needed.

But as i think I said before, power ratings may not be very helpful. If they rated a receiver at 100x7 watts, it may not have the performance you would expect of a 100 watt amp.

I would agree with everything said here.

Here is the easy answer to calculating wattage requirements for a home theatre:

Most people listen at no louder than Reference Volume (85db average, with up to 20db peaks = 105db peaks)

Get your speaker's sensitivity rating, mine are 98db for example.

Go to http://www.crownaudio.com/apps_htm/d...ct-pwr-req.htm

Enter your speaker's sensitivity, listening distance, and 105db for the desired SPL, and 3db for headroom

It will then tell you the wattage requirements for a single channel.

This will give you the wattage needed to drive your speaker to 3db higher than the loudest peaks in a movie. From an Amp standpoint, you should never be able to clip your amp if it can send out this wattage.

In my case, I need 90 watts per channel to reach 105db with 3db of headroom (or 45 watts to reach 105db and be right on the verge of clipping).

What does this tell me ? It tells me that my Pioneer which is rated at 110watts per channel (2 channels driven) may be able to handle short peaks of 105db if the front 2 channels are being driven. If the movie mixer is also sending loud peaks to the center channel, I may run out of power early and clip the amp.

If I want to be able to run 105db peaks and not worry about clipping, I will need an amp that can do 90watts continuous.

Now, in my condo, I usually listen to movies no louder than -10db, as do most people. That puts my average listening volume at 75db, and peaks at 95db.

My wattage requirements are now only 9 watts to get 95db with 3db of headroom. So at -10db from reference, any receiver out there should have no trouble giving you good clean sound with no clipping, with efficient speakers.

Nice thing though, is that, in theory, on paper, the 400 watt amp only gives you a 3 dB gain in SPL. Which is not much.

So if finances were tight, one might simply choose to live with 3 dB less SPL. your ears might even appreciate it

I have never been a fan of THX reference levels at home. About as loud as I go is 10 dB below that. Or 95 dB peaks. I understand everyone is different, and I am not telling you what you like. Just pointing out that not everyone will need a lot of power, because they don't listen that loud.

A -10 dB change in volume is quite a bit, but it reduces power needs drastically from what reference level requires.

And don't forget, that if your sub is not up to 115 dB peaks, you are not truly reference level capable in any case.

Clipping the amp hard enough to kill the tweeters IS overdriving the speaker. Too much power to tweeters kills them, not the shape of the wave they receive.