Measuring Amplifiers - Page 52

So is Chuck OK?
Anyone that knows him check in?
Or did he get tired chasing time slips. Er. Watt reports.

He posts over on HTGuide a lot more then here any more. I suspect his speaker projects are top priority these days and maybe AVS does not offer him much anymore.

DC means no variation with time, not everlasting.

By your definition batteries aren't DC.

Yes; mathematically, batteries aren't DC, or 0Hz, or 1/(infinite time). Hardly anything is. Frequency is inverse of time, so strictly speaking, any given battery has a low-frequency cutoff frequency of 1/hours, 1/days, 1/months, etc. But practically, for all our purposes, batteries are approximately DC.

HOWEVER, and this is the main point, symmetrically-clipped zero-mean sine waves have no more DC content than their unclipped counterparts. Lowest frequency content in both is the fundamental. A symmetrically-clipped zero-mean 20kHz sine wave has NO frequency content below 20kHz. If you want to call the flat tops and bottoms DC, you're approximating 20kHz and up content as DC.

Don't know how else to explain this, so I'll let it go. Will simply say that clipping and DC offset are two distinct phenomena, and leave it at that.

EDIT: Oops; just thought of one thing. Some may think that DC is required for square waves to stay square, but actually, it's that the LF corner has to be low enough to keep the phase response good enough with "simple" filters...

"A symmetrically-clipped zero-mean 20kHz sine wave has NO frequency content below 20kHz"

See how your tweeters like 20kHz square waves at rated power.

Maybe you're definition is correct, but I'm interested in the engineering concepts.

The fact is that the flat portions of the waveform don't have the di/dt that would let inductance reduce the power input.

Okay; this is a somewhat different argument, so I'll try to address it.

Tweeters don't like high-amplitude 20kHz square waves because of HF content, not DC. Pass that same tweeter-busting 20kHz square wave through a 1kHz HPF first. Happy tweeter? Highly unlikely. Same signal? Pretty much. Square waves have more HF content than the corresponding sine wave, but no more LF content. You gotta see the mathematical expression for a square wave, is all it boils down to. Nothing but the fundamental sine wave, plus smaller and smaller fractions of all the odd harmonics from there on up. Nothing at all at lower frequencies.

Actually, must also point out that the original tweeter argument is fallacious. Tweeters don't like DC: agreed. Tweeters don't like large-amplitude square waves: agreed. Therefore, square waves contain DC: does not logically follow at all. Tweeters don't like fire, either, but fire doesn't necessarily contain DC.

The fact is that if you have any inductance, the steep portions and "corners" will be the first to go. Low pass a square wave "enough," and you end up with the fundamental sine wave.

Much as I personally don't care much for math (anymore), Fourier came up with some extremely elegant stuff. Not immediately intuitively obvious--at least not to me--but satisfying once you get the basic idea. Kinda need to get into Fourier Series if one is interested in this sort of thing, and Taylor Series approximations if interested in distortion.

Could this thread be stickied?? I refer back to it quite often and I'm sure others do also. Seems like it would be a good sticky.

A sticky would be good as I refer back to it too. Ever since Chasw added the direct links on the first page it is a quick to get the info.

Figured I would add in the request to sticky this thread also.

I second that. Somebody go to the other forums to get Chuck back over here. I have an amp I need him to test

Bump... can we sticky this thread?

And yet another request for a sticky.

Is there anyone who has the equipment to take over this thread? It is very nice to have.

Oh yeah, another request for sticky.

Not a bad idea as a lot of us use this thread on a regular basis.

This seems like a decent place for a discussion I'd like to have.

There always seems to be an argument over just how long an amplifier should be able to maintain output, ratings and some other things. Some of us would prefer that an amp could sit there and do a sine wave for an hour without cooking itself. Others feel that 30ms is enough duration. Most are probably somewhere in between. Do you use a real unregulated AC line? Lately it seems that manufacturer's are siding with marketing and the bigger the # you can claim the better, so we are getting burst ratings of 4000w (EP4000), 6000w, 18000w etc...A lot of people suggest that you "test the amp in the real world" and that's what I'd like to discuss.

Exactly what would be a valid real world test procedure of an amplifiers performance that could be used for nearly any amplifier? I'm not talking about evaluating SQ differences in some kind of listening panel, but I guess subjective comments could be a part of it. I'm interested in determining real world output limits both dynamically and long term, overload characteristics, reliability, long term performance with heavy demand, etc. That kind of thing. I'm thinking it would be something involving a bank of high power speakers probably limited to the bass range mostly as the default load? I don't think anyone wants to try max output or overload testing with tweeters and mids involved...
Perhaps a normal high power full range speaker to gauge the subjective performance of the amp with music?

Ideas?

I am with you guys, sticky this thread please.

Valuable information that should be readily available............sticky please.

There will be many different answers, since "real world" is subjective. The whole "music power" thing has led to endless debates since "music" means different things to everyone, and as a matter of fact amp manufacturers don't actually rate "music power" based on music.

In my opinion RMS power is the only meaningful number because it's the only one which everybody tests in the same way. But of course, let's go into the whole "music power" thing.

A number of people mentioned the 1/8 rating, that 1/8 peak power represents average energy in a music piece. Okay let's take that as base and rate according to that. But wait... With the whole "loudness war" trend still going on, the average power of music is increasing. So what do we rate with then? We invent a "burst power" which can be everywhere from one second to one cycle of a 1kHz sinewave. And thus chaos ensues.

Now, let me explain the pro audio world a bit. We won't talk about infrasonics since bass in pro audio usually stops at 50Hz. I was actually arguing with a DJ a few weeks ago, he insisted that you can't hear lower than 45Hz, that's why the highpass filters are set for 50Hz. I should have directed him to the IB Cult...

Okay, so we don't need high levels of very low frequencies in pro audio. However, things like impedance mismatch and the use of limiters (or even no limiters at all, hard clipping), are common in pro audio. So take a 2kW amp which is rated with the 1/8 ratio mentioned above. If a limiter is used to squeeze more power from the amp, while driving a 4-ohm load on one channel and a 2-ohm load on the other will it still perform as advertised or will it go up in smoke? Well, i've seen plenty of the latter.

But this forum is about home theater and tons of sustained LFE. In movie soundtracks, even if the long-term average level (let's say average of 10 minutes) is often lower than the 1/8 ratio, there are sustained bass effects 10 seconds in length and even more. So will the 2kW amp really deliver 2kW for 10 seconds when it has been rated for bursts of 100 milliseconds? Again, debatable. Thus i still feel that the RMS rating is the way to go.

But to satisfy the "music power" crowd i'd test in the following way besides RMS: The most demanding instrument in live music is the kick, correct? Well, let's consider a typical kick drum sample, with a sharp attack at 0dB and the bass part that follows the attack at -12dB. This fits into the 1/8 power rule. So let's make this kick drum say 300msec long. Now, kick the kick and see how far you can go without clipping the attack, and there you got "music power".

That'd be my opinion, anyone can comment or criticize.

I'd basically agree that the kick is usually the loudest thing in the mix and it's also a complex sound. That's one thing that I have in mind as a test of how much dynamic power an amp has. I'd also agree that rms is the best method of determining ratings and there is nothing wrong with a dynamic or burst rating, but it seems like there is no differenciation and a new 3000w rated amplifier is not the same thing as a 3000w rated amplifier from 15-20 years ago. Both types of ratings should be used. Exactly what duration is an rms test anyway? Seems to be too much gray area all around. We will never come to a consensus on what music is or what a real world test is as there are just too many variables. You are right that during some movies or some of the newer music there are heavy bass effects that can last for a sustained period of time of at least a few seconds at high level and it is on these that I feel the burst rated type amps may fall well short of their claimed capabilities and sag 3-6db. I have no proof of that though.

"... there are sustained bass effects 10 seconds in length and even more. So will the 2kW amp really deliver 2kW for 10 seconds when it has been rated for bursts of 100 milliseconds? Again, debatable. Thus i still feel that the RMS rating is the way to go."

I agree, depending on what you mean by rms.

To me the relevant spec would be how much power can it output for 30 s at low freq.

Yeah 30 seconds sounds like a good enough measure for the RMS rating.

Oh and throwing one more idea in the mix - peak power is basically a function of the energy stored in the capacitors. What should be defined as music power would be repetitive peak power. Say you got capacitors able to deliver 3kW worth of kick. Now all this energy isn't too useful if the power supply can't charge those capacitors back before the next kick drum hits.

30s at LF sounds good to me too for rms. Perhaps a dynamic test with a single complex and sharp transient like a kick drum hit for maximum dynamic output and a repetitive test with the same signal to test for recovery times and output? What sort of time between repetitions though, 6 beats per sec would be a moderate double kick rate? I'm also thinking that a complex music selection should be used to evaluate the onset of clipping with real material and perhaps you could loop a heavily compressed metal track for a few hours just below clipping to determine the long term ruggedness?

Maybe just set-up an array of subs that can handle a very large amount of power reliably and set-up a mic to record the peak and average spl levels during the dynamic and rms test. Something like spectrum labs could be used to look at the output from the speakers and compare the difference when different amps are used.

Wow, this thread is back from the dead once again.

Sticky please!!

Sticky please!

mmmmm sticky

ricci, a usable standard can't really be created because, as you know, the power drawn depends almost entirely on the impedance of the speaker/subwoofer system. if the test system is an re xxx sealed, it will easily handle 20-40hz all day, but compress 80hz because almost 3x power needs to be dissipated up there. conversely, a jbl 2242 ported, is almost the opposite, with minimal power required to pound out 50-70hz content, while huge power is required for 25hz content. then, add to that bit that content varies wildly across listener preferences. some care much for 15hz sustained (20s+ durations), while others care only for 50hz burstable (5 cycle each, 2 per second). while i don't have first hand knowledge here, my guess is the speaker/subwoofer design impacts the amplifier performance as much, if not more, than the amplifier design itself. does that make any sense?

30 seconds for RMS? You guys are way tough. I tend to listen to and watch a lot of media on my system, and I would say a real world indicator of peak power need not be longer than 3 seconds.

A simple REW sweep is longer than 3 seconds Steve. 30 is a good number I think, very harsh but good.

Who said that actual speakers are ever used for RMS testing? See, that's why Chuck's tests were done into resistive loads. Because, as diverse as loudspeakers are, you have a starting point.


I break in 200W+ speakers with my 2x 25W amp. I can run it at full RMS or even 10% THD all day. It's rated for 2x 25W @ 8 ohms, but i used to have 4 ohm speakers on it, and even tested 2 ohm woofers and it never complained. And it's over 20 years old. I think that speaks a bit about its quality... I replaced its output transistors a few years ago when i accidentally dropped my meter probe and shorted them. Had i not shorted them out it'd still be running on its original output stage.