Anatomy of a modern Audio/Video receiver
( disclaimer - I am not an electronics engineer. This information is based on what I have learned from reading and studying schematics. I mention Yamaha fairly often because that's what I know best, I am not promoting Yamaha over other brands.)
The modern AVR is a complex device, perhaps one of the more complex electronic devices in the home. It has a computer, amplifiers, audio signal processing chips, signal converter chips. Many AVRs have video processing chips as well.
The main circuits in an AVR are -
* Power supply
* Analog Audio
* Digital Audio
* Analog Video
* Digital Video
Each of these groups can be one or more circuit boards, or some could be combined onto a single circuit board. Care must be taken in the design such that circuits don't interfere with each other. There's also the need to deal with physical needs such as connector placement and air flow for proper cooling.
Audio Routing and Processing
To apply bass management, room correction, or surround processing, analog audio is first converted to digital using ADCs. An ADC, or Analog to Digital converter, converts one or more analog channels to a digital signal. The digital signals represent the original signal as a series of numbers.
The ADCs used in AVRs can usually handle two channels. AVRs typically cannot handle conversion of multi-channel sources for economic reasons, and therefore, processing on multi-channel analog inputs is non existent. This means you don't get surround sound, bass management or room correction processing on multi-channel analog. That's why you will see bass management settings on players with multi-channel analog outputs.
Before the analog audio is converted to digital, the receiver has circuitry to select among the various analog inputs. This circuitry can be quite complex if the receiver has audio zones. The receiver may also have routing so that analog audio can be sent to a recording device. Phono inputs have special circuitry to deal with their much weaker signal, and to apply RIAA equalization.
Digital audio is usually serialized so that it takes up the least space on a circuit board. A serial connection only needs a few traces ( electrical paths,) on a circuit board.
There are multiple digital audio sources, so the receiver has to have circuitry to switch from among the possible digital audio sources. This is made more complex by the fact there are three different digital audio connectors in common use, S/PDIF coaxial, S/PDIF optical (TOSLINK,) and HDMI. Of course, the receiver is not only selecting digital audio originating outside of the receiver, but also digital audio converted from analog.
Once analog is converted to digital, and the receiver has selected a single source, the digital audio is sent to a DSP. First, the AVR has a circuit to select among all of the possible digital inputs. HDMI is a bit more complex as it also contains video. Suffice to say, HDMI audio is routed through the DSPs just like any other audio, but the path is more complex.
To process the audio, one or more Digital Signal Processing (DSP) chips are used. As DSPs are limited in processing power, multiple DSPs may be used. This is why budget receivers can't usually apply surround processing in all modes, because their processing ability is limited. Multiple DSPs can be chained together with the output from one being the input of another.
DSPs are specialized processing chips which can be programmed to perform operations on their input and then output the results. DSPs can be programmed to perform any of the following tasks -
* Decoding of audio codecs such as Dolby Digital
* Application of post processing such as Dolby Pro Logic IIx
* DSP mode processing such as Yamaha's Cinema DSP ( typically can't be done at the same time as post processing such as DPL IIx is done)
* Bass Management ( For routing low frequency signals to a subwoofer)
* Room Correction processing; e,g, Audyssey, YPAO or MCACC
From the DSP, a number of channels will be output. For example, in a 7.1 receiver, a left, center, right, left surround, right surround, left rear surround, right rear surround and bass/LFE channel will be output. Yamaha's presence channels, and the upcoming Dolby Pro Logic IIz require even more channels.
After the DSP or DSPs, the digital signals will be fed through DAC chips, to produce an analog channel for each digital channel. DAC chips typically handle a pair of channels, so multiple DAC chips are used. The output from the DACs are sent through op amps, which is why some people seem interested in what op amps are used in the receiver. How important the DACs and their associated analog circuitry are to the sound is often debated.
Depending on the receiver's capabilities and design, the receiver might need switching circuits to select from multi-channel analog inputs, stereo analog which bypassed digital processing, and analog audio converted from the digital audio circuits. This is done before the volume control circuitry.
Volume control is handled by one or more volume control chips. For example, the Yamaha RX-V3900 uses four volume control chips. One is a higher quality stereo volume control chip used for the Left/Right channel. Another handles the Center, Surrounds and Bass/LFE channel. Two more handle zone volume (one for each zone.)
At this point, further switching may be needed for zones or biamping. Amplifying a zone takes amplifiers away from the main zone. Biamping also takes away amplifiers from the main zone. There are only so many amplifiers to go around.
Once all this is done, each channel of selected analog audio is sent to the amplifier circuitry.
(This article sticks to explaining class AB amplifiers)
An AVR has five, seven or even more amplifiers. These amplifiers share from a single power supply in most designs. See the section on power supply for more details.
AVR amplifier specs often don't take into account driving all amplifiers at the same time, which is one good reason to not place to much faith on power specs. Some benchmarks on AVRs show less than 1/3 of rated power with all channels driven.
AVRs often provide some flexibility as to how amplifiers are used such as to amplify one or more zones, or for biamping. It should be obvious that using amplifiers for biamping or in zones, will limit how many channels can be amplified in the main zone.
Amplification is done with transistors. You can think about a transistor as a valve. Rather than being controlled manually, the valve is connected to a pipe. Change in water pressure in the pipe control the valve. The valve controls the flow of water through another pipe. By having the pipe controller by the valve carry more water with more pressure than the pipe controlling the valve, you can "amplify" changes in pressure in the pipe controlling the valve.
An amplifier uses multiple transistors to increase a small signal, usually less than a volt to a larger signal with a peak voltage of many volts.
Each amplifier has a number of stages. A typical design has the three stages -
* Differential Input Stage
* Voltage Amplifier Stage
* Output Stage
The differential input has two inputs, the audio signal, and a signal fedback from the amplifiers output. The feedback reduces distortion.
The voltage amplifier stage is responsible for most of the amplification. It also has feedback, but it's local feedback - it's own output is fed back into it's input.
The final stage usually provide little amplification, but buffers the voltage amplifier stage from the low impedance of the speakers. The standard AVR amplifier uses a class AB or class B design, with two sets of transistors, one set handles the positive portion of the input signal, the other set handles the negative portion of the signal. In a class AB implementation, both sets of transistors will stay on at lower signal levels. At higher signal levels, class AB will work just like class B, and one set of transistors will switch off while the other is operating.
Besides the limitation of the supplied power, the power transistors, have definite limits, which is one reason why receivers have protection circuits. These circuits will monitor conditions such as temperature, and shut down the receiver if programmed limits are exceeded. Another common problem, shorting out your speaker leads, is detected and the receiver is shut down to avoid damage. It's prudent not to rely on the AVR protecting you, so avoid shorting out connections and playing too it too loud.
Receivers may also have video processing circuity. This circuitry must deal with different types of analog video as well as digital video. It can often convert video in various useful ways. If the receiver has zones which support video, it also has to route video (usually restricted to composite analog video.)
I will describe how the RX-V3900 handles video. Analog video is first sent through a series of chips which select from the current input. These chips deal with composite, S-Video and component video signals. The analog video is then sent to an analog to video converter (there's also analog video bypass, but let's ignore that.)
HDMI video is sent through two 2-channel switches, which are hooked into a two channel HDMI receiver chip. This allows for four inputs. The HDMI receiver sends digital audio to the audio circuits and sends the video to it's video processing chip.
The digitized analog video can also be sent to the VP chip. In the case of the 3900, a single video processing chip handles both deinterlacing and scaling operations. This could be split into two different chips on some receivers. The video is then sent to a chip which can overlay the GUI on top of the video signal.
FInally, the processed digital video signal is sent to a digital to analog video converted to for output to analog video monitor jacks, and is sent to two HDMI transmitters.
AVRs have a CPU to handle many operations. Common operations -
* Processing user inputs from remote controls or front panel controls
* Providing a user interface for changing receiver settings
* Controlling DSP chips
* Controlling volume control chips
* Controlling chips used to switch signals
* Controlling the operation of video processing chip(s)
Connected to the CPU is a ROM chip containing the firmware for the CPU. Many receivers allow for the firmware to be updated, but in some cases, the manufacturer requires updates done at the factory. Firmware updates may include updates for more ICs than the CPU - DSPs and other sophisticated chips also have firmware.
The CPU communicates to and controls other chips such as DSPs using standard methods such as chip to chip serial communications.
Seeing how I am not an expert on amplifiers, this will be a simplified overview. A typical amplifier design used in receivers is class B. Class AB is closely related.
A class B amplifier uses a push-pull design where one power transistor (or in some cases a set of them,) amplifies the positive portion of the audio signal, and another the negative portion of the audio signal. It should be understood the input to our amplifier is an audio signal which consists of a voltage level which varies from a fairly low negative voltage to a fairly low positive voltage.
We need this voltage to be much larger to move the drivers in our loudspeakers back and forth.
An amplifier has multiple stages as it must multiply both voltage and current. We will talk mainly about the final stage, and will focus on class B amplifiers.
At the final stage we rely on power transistors to do most of the work. The power transistors are driven by a low voltage signal. The power transistors are connected to a relatively high voltage provided by the power supply (see below.) One transistor is connected to a positive voltage and the positive portion of the audio signal is connected to it. The other transistor is connected to a negative voltage, and the negative part of the audio signal is connected to it.
"Handing off" the audio signal from one power transistor to another every time it switches polarity creates crossover distortion. A well designed amplifier should be able to cope with this according to Douglas Self in his book on amplifier design. Class A amps avoid this, but have very real drawbacks - they are heavy, hot and expensive due to inefficiencies in how they work. Class D amplifiers avoid this by working in a totally different way, but their operation is outside the scope of this brief article.
The term transistor is short hand for transfer resistor. The way they are used in an audio amplifier is to have a small voltage signal control how much of a large voltage is allowed to pass through the transistor. In this way the small signal is amplified many times.
The transistors must have the ability to handle the voltage and current which passes through them. The main limitation on a receiver's output power is not the transistors though. It's the power supply.
The power supply consists of one or more transformers, and associated circuitry. The power supply design puts limitations on how powerful the receiver is.
A transformer takes AC from the wall, and provides one or more taps with different voltages. These outputs are then fed to various regulatory and conversion circuits to provide the needed variety of DC voltages.
A standard amplifier design used in receivers is class B. There's also class AB, which works like class B except at low power.
The main feature of a class B design is that one power transistor (or transistors) is used for the positive part of the signal, and another is used for the negative part of the signal. This often referred to as a push-pull design. This design requires a fairly large positive and negative voltage supply (my receiver has +70/-70 volt rails). These voltages are often referred to as voltage rails. (or a number of similar terms.)
The main transformer in the receiver will supply a suitable AC voltage for these rail voltages. This is not suitable because it's an AC voltage. We need a positive and negative voltage rail. Not only must it be positive and negative, but it must maintain a fairly steady voltage. If you ever see terms like power supply ripple or PSRR (power supply ripple rejection,) they are talking about the ability of the supply to maintain a steady DC voltage.
First the AC from the transformer's output is fed into a bridge rectifier. This slick device using nothing but diodes makes it so the AC voltage is always positive or negative. The positive and negative outputs from the rectifier are connected to large capacitors. These smooth out the voltage which would otherwise continuously vary from 0 to max voltage (or min voltage in the case of the negative side.)
These filter capacitors also give the amplifier the ability to handle brief demands.
If the world were a perfect place, the power supply could put out the designed rail voltage all day long no matter what. Unfortunately, the world is not perfect. There are definitely limits on this. If the amplifier demands too much from the power supply, it will be unable to keep the voltage at the rated level.
If you look at some receiver reviews with benchmarks, you will see receivers differ quite a bit in their ability to provide power when five or seven channels are driven at the same time. A receiver rated at 100x7 watts is likely a stretch by the marketing department. It has 7 channels. And those channels can do at least 100 watts (in many cases more than that.) But it can't push 100 watts into all all channels at the same time. The power supply voltage sags under the heavy load, and clipping occurs at a lower than rated power.
A really budget 100x7 watt receiver might do 35 watts with all channels driven. A more expensive receiver might be able to do 80. It's all about transformer size. To some extent, more expensive receivers use larger transformers and put out more power. You might read about people discussing receiver weight. That's because transformers account for a lot of a receiver's total weight. Receiver weight can be an indication of transformer size, and transformer size is directly related to receiver power.
People probably need less power than they think. Realistically no one needs 100 watts continuous power to all channels. That would be stupid loud, probably over 100 dB - pushing rock concert levels. If you don't have neighbors and don't value your ears, and you have speakers which can handle that kind of power, you might be an exception.
What you do want is sufficient power supply capacity to handle peak demands. Movies especially are very dynamic as you likely know. Many people have the remote near bye when watching movies as they are maybe TOO dynamic - hard to hear in a quiet scene, then very loud in a follow action scene.
A reasonable goal is to avoid excessive clipping at comfortable listening levels. A little bit of clipping is likely not audible and not damaging to your speakers. Too much clipping and it will impact the sound, and could be hazardous to your speakers ( there are some good articles on clipping you can find with google.)
Unfortunately there's no simple rule to determine whether a given receiver will fall short. If it's a light weight budget model, it's fair to say it's probably going to have some real limitations. If it's a mid range model it can likely better handle peak demands without clipping.
By convention, a receiver would not be a receiver without a tuner. Tuners have the ability to tune in either AM or FM broadcasts, or possibly HD Radio.
Typically, the tuner works like any other source, and the same processing options are available for it.
HD Radio tuners allow you to tune in digital broadcasts in the same band as conventional AM/FM. The tuner will automatically lock into the digital signal if possible. HD Radio allows for sub channels, that is, multiple channels on the same frequency by using multiplexing.
A conventional tuner uses the superheterodyne technique, to tune in channels. This was created by Edwin Armstrong. Basically, when two waves are combined, you get both waves, plus the sum of both and the difference of both. By carefully tuning a local oscillator, you can get the same difference every time. For example, FM radio tuners are designed such that when the local oscillator is combined with the signal from the antenna, a 10.7 Mhz signal is always created. The tuner circuitry is then designed to amplify a 10.5 Mhz signal.
"But this one goes up to 11"