• Welcome to The Audio Annex! If you have any trouble logging in or signing up, please contact 'admin - at - theaudioannex.com'. Enjoy!
  • HTTPS (secure web browser connection) has been enabled - just add "https://" to the start of the URL in your address bar, e.g. "https://theaudioannex.com/forum/"
  • Congratulations! If you're seeing this notice, it means you're connected to the new server. Go ahead and post as usual, enjoy!
  • I've just upgraded the forum software to Xenforo 2.0. Please let me know if you have any problems with it. I'm still working on installing styles... coming soon.

Department of Dispelling Stupid Ass Amplifier Myths

soundhound

Well-Known Member
Balanced amplifiers. Beamed down from the heavens as the answer to everyone's amplification nirvana. The touted features of these wonderful machines are, according to ATI:

1) Lower Noise
2) Double the slew rate
3) Apparent gain in volume
4) Immunity to stray fields
5) Lower Distortion
6) Reduction in all types of amplitude distortion
7) Immunity to hum
8) Better transparent performance

Here, I intend to rip these "features" a new asshole.

First, some important background. Whatever you might hear from manufacturers, and whatever your mommy told you, "balanced" amplifiers (really, the correct term is differential, and I will use that from now on) were created mainly to reduce the heat generated by multichannel amplifiers which have as many as 7 channels stuffed into a single chassis. Without differential design, the multichannel amplifiers we have today would not be possible because they would generate far too much heat. As a consequence, they would need so much expensive heatsinking and/or fans as to render them either far too expensive for consumers, or too noisy from the fans.

Differential amplifiers pull off this heat reducing trick by reducing the + and - rail voltages by as much as a half. Where a traditional 200 watt class A/B amplifier might have rail voltages of +/- 80VDC, a differential class A/B amplifier would typically have rail voltages of +/- 40-50 VDC. By having lower rail voltages, there is less voltage potential across the output transistors, and as a consequence, less power dissipated by the output transistors in the form of heat (remember - class A/B amplifiers are at best 60% efficient).

A 200 watt differential amplifier is really nothing more than two 50 watt amplifiers in bridge mode. Yes, that's right...nothing more than that. If you have any old stereo 50 watt amplifier which could be put into bridge mode, then you would have exactly the same thing as a "balanced" amplifier.

Let's do some math. A 50 watt amplifier puts out 20 volts into an 8 ohm load. If you take two of those 50 watt amplifiers, invert the polarity of one of them and connect the 8 ohm load across the "hot" terminals of the two amplifiers, then you would see a total of 40 volts across the 8 ohm load. Using Ohm's law, E (voltage) squared divided by 8 (ohms), equals....viola....200 watts!

That is how a differential amplifier works. The only other requirement is that each channel, and the power supply, must be robust enough to support that 200 watts output. So in practice each channel of a differential amplifier has more output transistors than would be required for a "normal" 50 watt amplifier, and the power supply is capable of more amperage.

Differential amplifiers usually have "balanced" (that's really differential) XLR inputs. When you are inputting a signal which is already differential (one phase normal, and one inverted), those two polarities are sent directly to the two halves of the amplifier. When you input a single ended signal into a differential amplifier, the signal goes directly to the "+" side of the amplifier, and also, via an inversion stage, to the "-" half.

Now for the alleged "benefits" (which are really bullshit marketing to make something which doesn't sound all that sexy (heat control) into whiz-bang "features"):

1) Lower Noise: A differential amplifier behaves basically in the same way as any differential circuit - it has common mode rejection. If the cables going to your power amplifier are picking up hum and noise, then that will be largely canceled. Unless the cables to the power amplifier are really long, are routed near large motors, or are in an area where EMI/RFI are prevalent, then there is no reason to buy a differential amplifier for this feature.

2) Double the slew rate: This one is true. However it should be noted that every competent "regular" power amplifier has more than enough slew rate to reproduce the highest audio frequencies. If an amplifier has a specification that it can output 200 watts all the way out to 20kHz, then it has enough slew rate to do the job. Having "more" slew rate is not buying you anything.

3) Apparent gain in volume: BULLSHIT, BULLSHIT, B U L L S H I T ! ! ! ! ! Basically all home theater amplifiers have a standardized gain of 28dB, which is the same gain specified in THX specs. The differential (XLR) and single ended (RCA) inputs have buffering which reduces the voltage of the differential input by 6dB, which makes the net gain of the amplifier the same, whether the differential or single ended inputs are used.

4) Immunity to stray fields: Nope, bullshit. This is because differential circuits rely on the fact that the "+" and "-" polarities of the signal are intimately placed next to each other - i.e. in a cable, the two wires are twisted tightly together. This assures that any interference which impinges on the cable equally effects both wires (both polarities). The input of a differential circuit sums the signal on the two wires together. That which is different between the two wires is passed (i.e. the audio), and anything which is common and equal on both wires is canceled (i.e. the noise). The relatively widely spaced physical layout of the circuitry inside a power amplifier makes it impossible for any real cancellation at any frequencies which would be useful.

5) Lower Distortion: Yep, that's true...but be careful what you wish for. When talking about distortion, the percentage of distortion is not nearly as important as the spectra of the distortion. Even order harmonic distortion is both easily accepted by the ear (it's simple octaves of the original frequency), and its presence tends to mask the presence of the odd order harmonic distortion components. Odd order harmonic distortion is of a much more complex relationship to the fundamental frequency, and as a consequence is more grating to the ear. Differential amplifiers get their "lower" distortion spec by eliminating the even order harmonic distortion components and leaving the odd order ones laid bare.

6) Reduction in all types of amplitude distortion: To be honest, I don't know what the fuck they're talking about here. Amplitude "distortion" is non-flat frequency response, and differential amplifiers are no better than regular ones in this respect. This is an example of what happens when marketing people take too many drugs...

7) Immunity to hum: Bullshit for the same reason as #4 above.

8) Better transparent performance: Bullshit. Transient response is directly dependent on bandwidth, and differential amplifiers don't have bandwidths any better than conventional ones. A "transient" wavefront is nothing more than a fundamental frequency with many harmonics riding on top of it - just like a square wave, the most "transient" signal of them all is nothing more than a fundamental with every harmonic component present at equal amplitude, theoretically up to infinity. If the bandwidth of an amplifier is sufficient to reproduce all the frequencies up to and slightly beyond the audio spectrum, then it has good enough "transient performance" to reproduce the steepest waveform the human ear can hear.

Wow...I feel much better now.... :p
 
And it was due to all the bullshit that I purchased a single-ended amp over the differential version.

Welcome back Soundhound.
 
soundhound said:
Wow...I feel much better now....
Kick back, put your feet up and bask in the knowledge of a job well done!
:eek:bscene-drinkingcheers:
 
Thank you for the explanations, soundhound.

Now, how about one on bias point of amp and sound difference? :twisted:
 
DIYer said:
Thank you for the explanations, soundhound.

Now, how about one on bias point of amp and sound difference? :twisted:

This one's relatively easy.

As part of the heat control in multichannel amplifiers, manufacturers must run the standing bias current through the output transistors somewhat less than optimal. This means that they are more prone to exhibiting crossover distortion which occurs at the point where each output transistor "half" passes the signal to the opposite set of transistors. To avoid crossover distortion, a standing bias current is run through the output transistors so that the amplifier effectively runs in class "A" up to the point where the amplifier is producing several watts, where it then transitions to class A/B. The longer the amplifier can stay in class A, the less the chance of exhibiting crossover distortion. In the old pre-home-theater days, amplifiers usually stayed in class A up to the point where they were producing several watts (or tens of watts in some cases). Most home theater multichannel amps nowadays are only in class A up to maybe a watt at the most -thus more crossover distortion.

To increase the standing bias current in today's multichannel amps would entail more heatsinking and/or fans.
 
soundhound said:
As part of the heat control in multichannel amplifiers, manufacturers must run the standing bias current through the output transistors somewhat less than optimal. This means that they are more prone to exhibiting crossover distortion which occurs at the point where each output transistor "half" passes the signal to the opposite set of transistors.
How will this crossover distortion manifest itself in sound characteristics? Do people commonly interpret it as harsh sounding or anything like that?
 
DIYer said:
soundhound said:
As part of the heat control in multichannel amplifiers, manufacturers must run the standing bias current through the output transistors somewhat less than optimal. This means that they are more prone to exhibiting crossover distortion which occurs at the point where each output transistor "half" passes the signal to the opposite set of transistors.
How will this crossover distortion manifest itself in sound characteristics? Do people commonly interpret it as harsh sounding or anything like that?

Since the crossover distortion "glitch" is a sharp transient type of wavefront, it will have a rich harmonic series, both odd and even well up into the treble. It can sound harsh, but with sensitive enough speakers (like my horns) crossover distortion imparts a "grainy" type of sound. This is especially true of music which is mostly low frequencies (like organ pedal) with not a lot of high frequencies to obscure the distortion. Class A amplifiers avoid the whole problem since there's only one output device covering the entire waveform.
 
Wow! Once again SH demonstrates why he's such a valued member! :eusa-clap:

:doh: I used to think I knew just enough to be dangerous, but once again I feel like a dummy who knows nothing ......... :angry-banghead:

Thanks for the reality check.
 
Interesting information/knowledge.

Thanks, soundhound.


It'll take awhile for me to digest it all . . .
 
Unless I've missed something, this seems largely about multichannel amps and I'd assume little of this applies to stereo and mono amps. Regarding the latter, if it's not specified, how would a person go about determining whether a particular amp is a single-ended design?
 
Zing said:
Unless I've missed something, this seems largely about multichannel amps and I'd assume little of this applies to stereo and mono amps. Regarding the latter, if it's not specified, how would a person go about determining whether a particular amp is a single-ended design?

What I was discussing mainly applies to multichannel amps, however you should be aware that with some "stereo" amps, they simply use only two of the circuit modules from their multichannel amps, so the stereo channels would be no different in any way than the same modules in a multichannel amp, and have the same downsides. Generally, these amplifiers are built into the same chassis as the manufacturer's multichannel amps, with just a lot of empty space inside.

If an amplifier is "balanced", you can be almost certain that the advertising will crow about that fact, as if it were a big plus. If the manufacturer doesn't scream that an amp is "balanced" then it is probably single ended. Also, a sure fire way to tell if an amplifier is a balanced design is that the manufacturer will say that no two of the channels can be used in bridge mode. This is because, of course, a "balanced" amplifier is just two lower wattage single ended channels in bridged mode already. If the manufacturer says that the amp can be used in bridged mode, then it has to be a single ended design.
 
Alien said:
Is there a beginners book or website you recommend about amplifiers? I'm way behind.

I'm sure there are some articles on the web, but I don't know of any specifically. Try a search on audio amplifier theory. I'm pretty sure you won't find any information on the topic I originally started this thread with. I only know it from being an engineer at an amplifier manufacturer, and I'm sure they would not admit any of it on the record; their business is to sell amplifiers, and more amplifiers however they can. That's not necessarily a bad thing, but consumers should know if they really need to "upgrade" for a valid reason.
 
Two questions:

1. What's the most significant change in amplifier technology in the past 10 years? (My guess is the implementation of digital amplification.)

2. What's the most significant improvement in amplifier technology in the past 10 years? (My guess is that question is a head scratcher.)

John
 
yromj said:
Two questions:

1. What's the most significant change in amplifier technology in the past 10 years? (My guess is the implementation of digital amplification.)

2. What's the most significant improvement in amplifier technology in the past 10 years? (My guess is that question is a head scratcher.)

John

1) You're correct that it would be digital amplification. For serious audio work however, I still don't view this technology as an "improvement", at least in sound quality. The best that digital will ever do is to equal analog IMHO, which makes it a solution to a problem which doesn't exist. If power efficiency is taken into account, then digital makes sense.

When working at ATI, I had the opportunity to evaluate and dissect some of the best digital amps on the market (including Halcro). In all instances, the best that a digital amp could do was to come close to what analog already achieves. You will see specifications of digital amps which show very low distortion, however this is seriously skewed by the fact that the highest fundamental frequency that can be evaluated for distortion on a digital amp is 7kHz. This is because in order to eliminate the spurious clock frequencies from influencing the measurement, a steep filter known as an AES-17 filter, is inserted before the measurement set which chops off measurements at slightly over 20kHz - including the distortion components which reside there, and noise. The result is that the highest distortion component of even a 7kHz wave that could be measured is the 2nd harmonic, with a bit of the 3rd: this makes these measurements misleading at best and seriously flawed.

2) There have been no improvements. In fact it could be argued that the reverse is true; sound quality has gotten worse. Power amplifiers for home theater use especially run very lean bias currents in the name of keeping heat at manageable levels. The "new" classes of amplifiers like class G etc are also steps backward in sonics; they are more power efficient, but that's all.

The gold standard for audio amplification sonic quality is still vaccum tube amplification, and single ended triode (SET) amplifiers: technologies which date from the 1930s.

As long as basic speaker technology remains as it is, I don't see any fundamental change to any of this.
 
I would like to question request your opinion and review about Single ended tubes, against a Fet amp, and the solid state Transistor in class A.

using a class A transistor amp can generate heat and require heat sinks. What is the shortfall on the transistor amp compared to the Tube.

Same question on the fet design.
 
malsackj said:
I would like to question request your opinion and review about Single ended tubes, against a Fet amp, and the solid state Transistor in class A.

using a class A transistor amp can generate heat and require heat sinks. What is the shortfall on the transistor amp compared to the Tube.

Same question on the fet design.

While class A single ended transistor amps are better sounding than regular class A/B types, they still operate on radically different principles than tubes, and produce more of a transistor sound than a tube sound. FETs are voltage driven devices (much like tubes) where bipolar transistors are current driven devices, but they still do not produce tube-like sound.

These are generalizations since things like the specific topology of the circuits contribute as much to the final sound as the core devices; things like negative feedback and how its handled influence the sound a lot. This is where FETs and bipolar transistors are at a disadvantage compared to tubes; triode tubes can easily operate with zero global negative feedback, where FETs and bipolar transistors almost always have to utilize at least some, otherwise their distortion and listening quality would be unacceptable (also tetrode and pentode tubes need at least some negative feedback since their native distortion is higher than that of triodes, which is why single ended amps use triodes rather than tetrodes or pentodes...to complicate matters more, tetrodes or pentodes can be connected to emulate a triode...). A tube amp which uses massive negative feedback might sound worse than a transistor or FET amp which has less feedback: the devil is in the details.

Since tubes operate at such high voltages (typically 250 VDC to 350 VDC), the energy storage in their power supply in joules is considerably more than a transistor (or FET) amp can realistically have. Many people (including me) believe that this aspect contributes to the more liquid sound of a tube amp.
 
SH, what about tube amps vs transistor amps for planer magnetic speakers? Over at the Planer Asylum, somebody is always making the claim that their 80watt @8ohm tube amp can blow away some 200watt @8ohm transistor amp. This would be powering a speaker that has a sensitivity of say 85db and runs around 4ohms or lower and powering a large planer speaker like my Magnepan 3.6's.

One side says the transistor wins, via it's higher power and higher damping factor. The other says the softer clipping of a tubes and the "higher quality" (whatever that means) of their watts make them the better choice.
 
Back
Top