Power Amp Output: AC or DC?

[JingleDjango]

I hope this is the appropriate section for this question:
I'm studying recording in school. In the live sound course, in our discussion of power amplifiers the instructor was adamant that the output of a power amplifier is a DC current. This is giving me a bit of cognitive dissonance because my understanding of audio signals is that they are fundamentally oscillating between + and -. I believed that the input stage received an AC signal of very low voltage/current and the output stage ideally spit out the same signal, but with greater amplitude (More positive voltage on the upswing, lower negative voltage going down).
http://www.prosoundweb.com/article/whats_under_the_hood_power_amplifier_sections_connectors_classes/P3/ <- This page suggests that the output can be either pure AC or a varying DC signal depending on the output class. But how would the loudspeakers respond to that DC offset, which would have to be quite great to lift the entire audio signal above 0Vground. Naturally if the output is transformer balanced then none of that DC energy should make it out of the box anyway!

Any help in wrapping my head around this conceptually would be appreciated!

 

[JohnRoberts]

Audio power amplifier outputs are just higher current, higher voltage replicas of the input audio waveforms,  so definitely AC...

Ask your instructor what DC audio sounds like (answer a fan blowing air at you  ;D )

JR

PS: Those who can do, those who can't do... teach, those who can't teach... criticize. (GB Shaw)

 

[JingleDjango]

That has been my suspicion! It's been very confusing trying to integrate what I hear in that class with my experience modding and building stuff.

So then, if an amplifier stage needs to be biased totally positive to perform its function, then that DC offset would have to be subtracted from the AC signal on the way out the amp? I assume this would be using capacitors if not transformers.

 

[Andy Peters]

http://www.prosoundweb.com/article/whats_under_the_hood_power_amplifier_sections_connectors_classes/P3/[/url] <- This page suggests that the output can be either pure AC or a varying DC signal depending on the output class.

There was a running argument on the old Live Audio Board (the precursor to ProSoundWeb) where one person insisted that an amplifier which had only a positive output (Class A, for example) was a DC amplifier. He based that opinion on a strict interpretation of the words that make up the initialisms "AC" and "DC."

Since that Class A amplifier's output current wasn't alternating in direction, it was direct current.

The truth is that the terms AC and DC are misnomers, and when we talk about signals we talk about the static component (DC) and the time-varying component (AC). All signals have both.

The amplifier's job, as JR says, is to take an input signal and embiggen it. It can be made larger in voltage and/or in current. (Power amplifier is a misnomer, too, as power just falls out of the relationship between voltage, current and resistance.) Nowhere in the definition of amplifier does it say that the signal (in or out) needs to be time-varying, in which case the AC component is 0.

The DC bias on the output of a single-ended amplifier is just a detail of the design. In the case of the push-pull amplifier which has both positive and negative power supply rails, you still have a DC bias. It's just at 0 V so it's ignored. 

Put another way, in both cases, the single-ended and the push-pull, your output signal has both an AC component and a DC component. In the former case the DC component is not 0 V.

-a

 

[JohnRoberts]

That has been my suspicion! It's been very confusing trying to integrate what I hear in that class with my experience modding and building stuff.

So then, if an amplifier stage needs to be biased totally positive to perform its function, then that DC offset would have to be subtracted from the AC signal on the way out the amp? I assume this would be using capacitors if not transformers.

Yup back decades ago some amplifiers using a single supply voltage would bias the output half way up (or down) and then use a huge capacitor in series with the output to block DC.

DC fed into a speaker can damage it.

Transformers generally don't like DC either, but some tube circuits operate with DC in a winding using a special transformer designed to accommodate the DC flux.

JR

 

[JingleDjango]

This has been really helpful. Thanks, guys.

 

[JohnRoberts]

http://www.prosoundweb.com/article/whats_under_the_hood_power_amplifier_sections_connectors_classes/P3/[/url] <- This page suggests that the output can be either pure AC or a varying DC signal depending on the output class.

There was a running argument on the old Live Audio Board (the precursor to ProSoundWeb) where one person insisted that an amplifier which had only a positive output (Class A, for example) was a DC amplifier. He based that opinion on a strict interpretation of the words that make up the initialisms "AC" and "DC."

Since that Class A amplifier's output current wasn't alternating in direction, it was direct current.

Not to quibble but while each Class A output driver is modulating a DC current, the net result output and seen by the loudspeaker is definitely AC.

The truth is that the terms AC and DC are misnomers, and when we talk about signals we talk about the static component (DC) and the time-varying component (AC). All signals have both.

The amplifier's job, as JR says, is to take an input signal and embiggen it. It can be made larger in voltage and/or in current. (Power amplifier is a misnomer, too, as power just falls out of the relationship between voltage, current and resistance.) Nowhere in the definition of amplifier does it say that the signal (in or out) needs to be time-varying, in which case the AC component is 0.

The DC bias on the output of a single-ended amplifier is just a detail of the design. In the case of the push-pull amplifier which has both positive and negative power supply rails, you still have a DC bias. It's just at 0 V so it's ignored. 

Put another way, in both cases, the single-ended and the push-pull, your output signal has both an AC component and a DC component. In the former case the DC component is not 0 V.

-a

There are any number of different amplifier topologies, but the important criteria is that the output signal is AC.

Back when I was working at Peavey I recall one competitor's amplifier that had the very undesirable characteristic of putting out a DC component along with it's AC (from asymmetrical clipping or current limiting). It killed a bunch of Peavey speakers before they got their act together.

JR

 

[PRR]

> the instructor was adamant that the output of a power amplifier is a DC current.

Listen to your instructors. Remember what they say, it may come up on an exam.

From 1% to 99% of what a teacher tells you is unclear, incomplete, or just plain wrong.

There is a story (apparently true) of a concert mixer power supply failed at a show. A Crown DC-300 audio power amplifier was connected to a 1.5V battery and "amplified" it to many Volts and many Amps of Clean DC.

(The DC-300 was originally an audio amplifier only incidentally. Who needed 2x150 Watts?? Test labs, shaker-tables, relay testing. So the subsonic/DC ability was important enough to put in the model name.)

Nearly all modern audio amplifiers have DC *protection*. They can make DC, but if they do more than a Volt for more than a second, a relay disconnects the speakers to prevent damage.

You can also connect a DC meter (volts or current, appropriately connected) and "see the DC" for yourself (don't bother the instructor). A "VU" meter withOUT the rectifier will twitch and jerk at high speaker power (may even jerk out of its pivots) but won't give a clear reading. Same for needle-meters. Digital meters may average-out audio to zero DC (with twitchings) or may get confused. (My otherwise good Fluke reads stray AC riding on top of its own DC Ohms test.)

> how would the loudspeakers respond to that DC offset

You can build a megaphone with a carbon mike, a transistor, a rugged speaker, and a 6V battery. True class A, and all the idle current (DC offset) flows through the speaker. This can only end well if the speaker has much more power-handling ability than the transistor. We typically used "15 Watt" mini-horns (stiff diaphragms) with 1-Watt transistors for sub-Watt audio (AC) output. The ~~Watt of heat in the speaker didn't cook it. Any more than a Watt into the transistor would drain the battery too darn fast.

Nealy all audio, acoustic and most electric, has big DC component. Atmospheric pressure! Without it, no sound! (and no listeners!) DC bias in most simple amplifiers to allow the _AC_ audio to swing both-ways. At the end, the loudspeaker already has atmospheric pressure, and it is conventional to ONLY send it the *variations* (AC), not the steady part (DC atmospheric pressure).

 

[JohnRoberts]

Now that we have blinded the OP with science I will try to distill this down.

Audio is AC so amplifier output is amplified AC. That AC output "could" have small inadvertent DC content but is dominantly AC. 

Crown made one amplifier model decades ago called the DC300 that had response down to DC and if you put DC voltage into it, DC comes out. Of course you can release the smoke from speakers that way, and many did from accidental DC present on it's input.

If your teacher believes amplifiers put out DC , and asks you that question on a test, the correct answer for him that one time is DC. Then work to remember that the correct answer everywhere else in the world is AC. 

Good luck.

JR