Audio and video signal susceptibility classes

My article Signal reference grids in the data center shows that different types of data communication interfaces are classified into susceptibility classes:

Data communication interface classification

Susceptibility class: Low immunity
Interfaces in class: Parallel ports, RS-232 ports, Proprietary back-plane, Video cables
Characteristics: Copper cabling with ground referenced signals. Any shift in ground voltage between the interconnected equipment is superimposed on the data signal. Inter-system ground noise of 0.1 volt or even less can interfere with communication (depends on interface).

Susceptibility class: Partial immunity
Interfaces in class: Modbus, RS-485, SCSI
Characteristics: Use balanced or differential signal transmission that is not ground referenced. Inter-system ground noise of 10 volts or more can interfere with communication when it exceeds interface “common mode range”.

Susceptibility class: High immunity
Interfaces in class: Ethernet
Characteristics: Copper communication interfaces that have a balanced or differential signal transmission with full transformer isolation at both ends (Ethernet over 1000 volt isolation). Communication protocols have built-in error correction.

Susceptibility class: Total immunity
Interfaces in class: Fiber optic, Wireless
Characteristics: Inter-system ground noise does not have any effect on communications.

In that article I also added: My own addition to list would be that analogue audio signals if your system uses then go to low immunity class (some properly implemented balanced line level audio signals would go to partial immunity class).

After some later thinking I think it could be useful to make similar susceptibility classes listing for signals commonly used in audio/video systems. Here is my audio/video susceptibility classes for audio/video signals (mostly analogue signals):

Audio and video susceptibility classes

Susceptibility class: Low immunity
Interfaces in class: Unbalanced analogue audio signal, analogue video signals on coaxial cable (composite video, S-video, RGB, VGA), audio over USB, many coaxial SPDIF implementations, DVI, HDMI
Characteristics: Copper cabling with ground referenced signals. Any shift in ground voltage between the interconnected equipment is superimposed on the signal. Inter-system ground noise in millivolts to less than volt level can be heard or seen on the signal.

Susceptibility class: Partial immunity
Interfaces in class: properly implemented balanced line level audio signals, AES/EBU digital audio, MIDI
Characteristics: Use balanced or differential signal transmission that is not ground referenced. The effect of inter-system ground noise is very much reduced (but not usually completely eliminated due limited common mode rejection capabilities in practical interfaces).

Susceptibility class: High immunity
Interfaces in class: Audio over Ethernet
Characteristics: Copper communication interfaces that have a balanced or differential signal transmission with full transformer isolation at both ends. Communication protocols have built-in error correction.

Susceptibility class: Total immunity
Interfaces in class: Fiber optic, Wireless, SPDIF over fiber, ADAT fiber connection, video over fiber
Characteristics: Inter-system ground noise does not have any effect on the signal.

Note that this is my first attempt to make this classification. Hopefully I got the different signals as close as possible to the right classes. This classification could be improved in the future. Feel free to comment if you have improvement ideas or questions on the classification.

Notes on audio/video classification

Balanced audio interface is a huge improvement over unbalanced signal interface. Instead of passing the wanted signal voltage in relation to a fixed zero-volt reference, a balanced interface generates the wanted signal voltage between two wires, neither of which is connected to ground. Since the wanted signal is conveyed as a varying voltage between those two wires, no ground reference is required at all, and the input circuitry simply looks for the voltage differential across the two wires. For this reason, a balanced input is sometimes referred to as a ‘differential input,’ and in ideal case it doesn’t care what is happening on the screen or ground connection, because it’s not looking at it.

Real-world implementations of balanced interfaces have practical limitations how well they can receive the signal between the two signal wires and leave out all the noise (CMRR). Also cables have some practical limitations in their noise performance. Practical cables and some equipment have some routes how some part of a ground loop current flowing on the balanced cable’s grounded screen can get into the signal (causes much less noise than with unbalanced connection).

The low immunity of unbalanced interfaces is the main reason why you need to understand different ground loop solving tricks (like ones described at UNDERSTANDING, FINDING, & ELIMINATING GROUND LOOPS IN AUDIO & VIDEO SYSTEMS document). If all audio/video interfaces were in partial immunity class, there would be a lot less ground loop problems to solve. There are different kind of tools to improved the noise susceptibility of unbalanced sensitive interfaces. For audio the most typical solution is to add audio isolation transformer. For analogue video signals the normal solutions are video isolator and humbugging transformer. Those additional tools can usually increase in the best case the susceptibility class nearly one class up.

How ground loops can affect digital interfaces

It is worth mentioning that ground loops can affect digital interfaces too, but in a far less obvious way. Unlike traditional analogue interfaces, where you can hear the result of a ground loop hum directly, a digital interface similarly afflicted rarely produces any hum-like sounds at all. Instead, what tends to happen is that the digital link becomes very unreliable and unstable, often producing dropout glitches and splats at the slightest provocation. Different digital interfaces have different characteristics on how they can handle ground loop.

SPDIF on coaxial cable is has about same signal level as video signal, and ground loop noise can disturb the signal when it is strong enough (typically causes output to be silenced for few seconds for every error). The most problematic digital connections tend to be S/PDIF coaxial (phono) ones, which, although unbalanced, should, in theory, have transformer-coupled outputs. Unfortunately, this is an area where corners get cut on budget audio interfaces, and if your S/PDIF phono sockets are bolted direct to the metalwork, no transformer has been used. On the other hand SPDIF signal over plastic fiber (Toslink) is in total immunity class.

AES/EBU that uses balanced signals, so it considerably less sensitive to noise than coaxial SPDIF. AES/EBU normally uses balanced connections with transformer isolation. It’s very rare to run into AES/EBU ground-loop problems, since these sockets generally appear on professional gear where no corners are cut.

MIDI inputs also incorporate opto-isolators, so plugging in a MIDI cable should never result in a ground loop. The most common cause of MIDI ground loops is using an incorrectly wired cable.

USB interface can be disturbed by ground loop. The USB bus is a Differential Bi-directional serial interface cable bus. Signal levels used on the bus depend on the USB speed used.

Digital video interfaces like DVI and HDMI can also have ground loop noise problems as they use a combination of balanced and unbalanced signals. Ground loop problems is one very possible source of intermittent problems over long distance in HDMI installations.

3 Comments

  1. HDMI and ground loops « Tomi Engdahl’s ePanorama blog says:

    [...] and ground loops In my Audio and video signal susceptibility classes I wrote that digital video interfaces like DVI and HDMI can also have ground loop noise problems as [...]

    Reply
  2. Ayesha Scioneaux says:

    I do not agree with some things, having said that I did liked the post in general… The article was in fact highly recommended to me by a buddy at stumbleupon and this lady was right. It is actually such a fantastic read! Thank you for sharing the facts along with us.

    Reply

Leave a Reply to Tomi Engdahl Cancel reply

Your email address will not be published. Required fields are marked *

*

*