Myth buster attempt: POWER AMPS

[sounzgud]

Two questions have been bothering me about power amps that I can't figure out unless I buy a bunch of different kinds of amps and blow a bunch of different kinds of speakers. So hopefully some of you have blown enough speakers and used enough amps to anwer my questions.

Question 1: What is the deal with head room?

I use extremely powerful compression/limiting tools before I send any signal to my power amps, so if I were to actually power my speakers to their program rating and run my amps hard, but not clipping, I would blow speakers. I end up using amps that are about 1.5x rms and I still haven't seen a clip led in about a year. Is head room mostly for systems that don't use much if any compression?

Question 2: Is there a difference among power amps?

I realize that there are different power consumptions, weights, and some amps break more easily than others. What I am talking about is performance, man! What is the difference between a nady 2400 watt power amp and a crown ma 2400? Besides the nady is lighter and draws less power(in that case I would be willing to wadger that the nady would break more easily [and I appoligize to crown, the i-tech series might as well have been hand crafted by god!]).


E-Rock!

[DAS]

In layman's terms headroom is simply the difference between the nominal and maximum operating levels of a device.

It gets considerably more involved when you start thinking about RMS versus peak and duty cycles, etc. (maybe when I have more time....), but the basic concept is as above.

As for the qualitative differences in amps...the bottom line is that you will find sonic differences that in some cases can be quite significant. Damping factor and slew rate are a couple of clues, but then looking at how the amp performs under various load conditions can weigh quite heavily on its sound in real world applications. That's a start. Maybe some others can chime in here.

[jeffmac]

I am a representative of QSC amplifiers so I hope you don't mind if I jump in and attempt to answer some of these questions...


Question 1: What is the deal with head room?

Headroom is the ratio of the largest undistorted signal possible through a unit or system, to that of the average signal level. For example, if the average level is +4 dBu and the largest level is +26 dBu, then there is 22 dB of headroom.

Headroom covers the ability of the amp to deal with the amplitude of transients, which are very quick spikes of high output. While high quality compressors will control some of this, it really takes a very high quality look ahead limiter to eliminate all transients. Any compressor has an attack time that is typically controllable but most have have a minimum attack time no matter what the controls are labeled as. If a transient shoots through, it takes a finite amount of time for the compressor to kick in. For that short period of time, the signal is actually outputting above the threshold.

Headroom is there so the amplifier still has a range of gain to be able to represent that signal rather than "square waving" the signal as it crosses over the maximum amount of amplitude it can output. When the amplifier does this, no matter how briefly, it is outputting huge amounts of high frequency content, which is potentially damaging to the speaker.

Keep in mind that depending on the quality of the sensing circuits and the reaction time of the amp, very short transients may not cause the clip light to go on...

If you are compressing and limiting heavily you also have to consider that you could be eliminating transients but that part of how we hear and localize sound is based on those transients. While you may be able to get more average signal out, you may make the music you are reproducing sound less than real. This could be a good or a bad thing depending on your intention!

Keep in mind that the simpler the audio you are reproducing, the more critical to realism those transients are. A solo piano or acoustic guitar desperately needs those transients to sound like it does when you are in the room with it...

Sooo...headroom is really for all systems, but more so for applications where you do not have COMPLETE control of the audio that is being amplified.

Also, keep in mind that most speakers are actually blown based on a DC pulse. This pulse can last less than a quarter of a second and still melt a voice coil. This DC can be created by clipping any where in the chain even if you are not clipping the amplifier. Heavy compression can actually accomplish this and the amplifier will never show this clipping. The speaker still dies a horrible death. This makes gainstaging through out the system critical.


Question 2: Is there a difference among power amps?

Whoo! This is a big one! I will try not to blather on too long!

1. Power supplies

They are a range of different power supplies for ampliers. These are critical for the accurate operation of an amp for several reasons.

First of all, filtering. Noise can come into an amplifier most easily through the AC inputs. Good amps have good filtering to deal with this...bad amps don't have much at all on this front. They are going to tend to perform worse when the power is not optimal. This could mean noise, distortion, or not getting the rated power out of an amp depending on where you plug it in. Better power supplies mean better audio...period.

Types of power supplies. There are linear power supplies and switching power supplies. Any power supply works at it's most efficient at the very top of its range. Right below max is where it is wasting the least power. Unfortunately, we are only using the full output of most amps at most 2or 3% of the time that they are running. When they are not at full output, that additional energy turns into heat and heat kills electronics. This is one of the reason that cheaper amps do break down more often.

One of things that QSC does on some of our our amps (PLX series for instance) is that we install "switching" power supplies. That means if the amp does not need all of the power to output the amplitude it needs, it switches to what is in essence a lower power supply. Since it is doing this it is always operating the amplifier at as close to max level as possible. This mean the amp is more efficient has less distortion and creates less heat. This means the amps sounds better and last longer. A QSC amp can make these power supply changes 230,000 times a second to react to the changes for the needs in power.

2. Heat dispersion: The ability to get rid of heat is critical to an amp. Becuase of this the quality of the heat sinks and the fan circuitry is a very big deal. QSC does this very well. We have huge sinks and temperature controlled fan speed. All QSC fans draw air from the back of the amp and blow it forward. This is very important, because if the amp is in a closet or rack, if the amp blows to the back, that heat builds up and can often not dissipate...killing you amp more quickly.

The ability to deal with heat also means that QSC amps deal with low ohmage load much better. 2 and 4 ohm loads make an amp work much harder and therefore build up much more heat. A quality amp will always run well at 2 ohms, so look for a spec on this when you are looking at amplifiers. You may not run your system at thes ohmages but speakers physical characteristics change over time and particulary as they are getting ready to die ohmages can swing very low. This can take the amp and the speaker out.

3. Amplifier circuits:

When you are comparing amplifiers,it is important to compare apples to apples. Take a look at the type of amplifying circuit in the amp.

There are different types of circuits that are used in audio amplifiers. The most common are A, AB, D (this does not mean digital) and H.

Different types of amplifier circuits are applicable to different kinds of amps.

Type A amps are accurate, simple and very wasteful of energy. For every watt you generate you are throwing one away. This makes them only applicable for smaller power amps...or very expensive and heavy larger amps.

Class B amps only amplify half of the signal. This makes them more efficient but very high in distortion. Not terribly useful for audio.

Class AB is a a compromise between A and B amplifiers and is sort of two B amps strapped together...on for each polarity of the output. THis is the most common amplifier format, but it suffers some distortion at the resting point of the amp which means it performs its worst at the lowest amplitudes. Negative feedback improves the performance of this type of amp, but at the cost of output power.

Class D amps are very efficient, but can introduced the same kind of quantization errors that we heard in early digital recordings. Many people criticize the audio quality of class D amps in higher frequency ranges so these often get used for subwoofer applications.

A class H amp (most of the High power QSC amps fall into this category) is basically a combination of a D and AB amplifier. It combines the switching power supply and modulating output of both amps and is a great combination of effieciency and quality. It is also more technically challenging to design and build.

If you want to go deeper into this subject, this is a pretty good, unbiased reference with some diagrams and circuits...http://en.wikipedia.org/wiki/Electronic_amplifier.


4. Fixed gain versus variable gain

All QSC amps are fixed gain which means they are applying 32 dB of gain. The knobs are actually input sensitivity not amp gain.

Most cheaper amps tend to be variable gain. This means the knobs on the front of the amp actually controls the gain. This is a noisier approach.




5. Quality of Specifications;

This is much harder to quantify. It is almost impossible to compare products by purely comparing specifications because the marketing and engineering departments almost always fight over who gets to write these. Quality companies (Crown, QSC etc.) tend to do their level best to deliver accurate specs. When comparing guitar amps you often hear people talking about "Marshall" watts because a Marshall 25 watt amp would often get much louder than someone else's 100 watt amp.

Audio amplifiers are no different. Wattages are specified by picking an acceptable level of distortion at a certain frequency range...so if a company is willing to accept 2% distortion in testing to get 2500 watts at 1 K they can make it look as good on paper as a Crown or QSC that is doing that same work full bandwidth at .01% distortion.

Both are 2500 watts...sort of. Which one is a USEABLE 2500 watts is a different question.

What it really comes down to is that "You get what you pay for" is amplified (no pun intended) when it comes to amplifiers! Because amplfiers are operating at much higher voltages than other audio components, the margin for error is much smaller and the consequences are much greater. Fire belching from a speaker or amp is certainly not unheard of.



A lot of cheap amplifiers will perform reasonable comparable to a better amplifier in perfect conditions...but how often is life and the conditions that we record and perform in perfect? Keep in mind that your neighbors Air conditioner or the lighting guy cross wiring his follow spots can hugely impact your conditions!


Sorry to write a novel but I hope there is some useful info in here.

[DAS]

Hey Jeff, really great to see you here. Great info. Thanks.

Everyone...Jeff is a great guy who REALLY knows what he's talking about, on much more than just power amps. Hopefully we can get him to post more on other subjects as well.

[dpd]

jeffmac - that rates a sure-fire entry for Post of the Year. Thanks. Very well-stated, especially the DC issue.

One thing about switching supplies that can cause audible differences - poorly-designed switchers can have artifacts on their outputs that can leak through to the audio chain and modulate the signals going through the amp. Holds for low/line level circuits, as well.

Power supplies can make night and day audible differences on identical audio circuits. Their impact on the overall sound is, IMO, not well understood by many audio gear consumers

[DAS]

Originally posted by dpd
Power supplies can make night and day audible differences on identical audio circuits. Their impact on the overall sound is, IMO, not well understood by many audio gear consumers

Amen!

[Dave Burris]

Good post Jeff! And a bit refreshing from the "hype" normally tossed around.

Let me add to the discussion "speakers" and "amplify" the specifications discussion.

First, you didn't specify what speakers you are blowing so there is no way to really determine if your problem is the amp at all.

That said, I would like to stress that unless you know electronics and can get companies to publish MEANINGFUL specifications, there is very little you can discern from "maximum" power ratings. For instance, as was mentioned already power supply capacity and design make a HUGE difference between maxiumum power capability and maximum SUSTAINED power capability, and neither is very useful without a distortion rating at that power.

Other things you need to know are sustained ratings at what impedance load and what ambient TEMPERATURE. Your speaker loads could be incorrectly designed for impedance, or you could simply be running the amplifiers too hot where their bias or protection circuits fail to operate as expected.

Properly designed high-power amps must be designed with adequate heat sink to keep inself cool at rated power. This requires either heavy massive heat fins or noisy fans moving lots of air or a combination of both.

I would start with a hard look at the speakers and crossovers. The amp could be the best designed in the world and will gladly supply its rated power until you overheat voice coils where any number of failures could occur. Take a close look at the crossover frequency, especially if you are blowing compression drivers rather than low-end drivers. You may simply be crossing over at too low a frequency and supplying power below the safe operating frequency.

Let me also add that using a compressor as a means of getting more amplitude out of an under-powered, under-designed system can lead to disaster (as you perhaps are seeing). We could argue all day about the "purpose" of a compressor, and my position would be that it is to control dynamics and not to "cheat" the system into supplying more power. Naturally they tend to be used for both because you can and it works to raise overall levels while being a bit safer. My personal view is that using a compressor this way is a sure sign of an under-designed system trying to do too much and that eventually something gives, usually the speakers but sometimes amps too or both.

My preference is to thoroughly review specifications for all the components and design for a sustained maxiumum power level approximately 10db (yes that 10X) above what you need to produce the required SPL over the required coverage area. Then use compression to control dynamics in the mix instead of a way to coax more average power out of the program material.

[Trumpetguy7]

Would it be wiser to get powered speakers? Like let's say the expensive Mackie one's? So my real question is: By getting a bi amplified or even tri-amplified speakers, do these problems still occur from passive one's with power amps?(the compression thing explained by jeffmac and other problems?

[DAS]

While there are certain advantages to amps built into speakers most of the same sorts of problems (with amps) still exist. Ideally the amps in a speaker such as the Mackie are optimized for those devices, and ideally it means you'd be able to safely get about as much out of those devices as possible. It does simplify things somewhat. The underlying issues are all still there, you just don't have to worry about them as much because you can't really change those relationships anyway. It makes it sort of "idiot proof" in a way.

[elsteve9]

Or at least, "Idiot-resistant"

-Stephen