PNW AES Banner

n.b. Chrome users need to refresh their browsers to ensure they have the latest content.

Meeting held June 1, 2013 at Raisbeck Hall, Cornish College

Everything You Need To Know About Using Ethernet Cable for Portable Audio
an All-Day, In-Depth Workshop
Presented By
Dan Mortensen and the AES PNW Section

image linked to pic1.jpg
(L-R: Karen Sitzberger, Dave Tosti-Lane, Lindsay Smith) Running over the cable with a scissor lift.
image linked to pic2.jpg
(L-R: AES Executive Director Bob Moses and Dan Mortensen) Torturing a cable by pulling it thru a small aperture.
image linked to pic4.jpg
The Audience: 13 AES members, 13 non-members

Media

Video Recordings

Video Recording Part 1 
Video Recording Part 2 (additional parts to follow) 

Audio Recordings

morning session, 64k MP3 (49.1MB)  
afternoon session, 64k MP3 (33.1MB)  
afternoon torture session, 64k MP3 (47.9MB)  

PowerPoint and Text Links

Steve Lampen's PowerPoint deck (4.8MB ppt)  
Cable Testing Definitions (225k pdf)  
Rick Rodriguez' paper "Ethernet Cable for Digital Audio" (442k pdf)  

Photos

Cable Twist Photos (2.4MB pdf)  
Photos and Audio Recording by Gary Louie.
Audio Recording mixing and editing by Rick Chinn.
Video Recording and editing by Dan Mortensen.
For June 2013, the PNW Section held an all-day workshop on CAT-type cabling for portable audio. Organized by PNW Committeeperson Dan Mortensen of Dansound, Inc. after some personal experiences with such cabling, several experts spoke and demonstrated, and attendees had the chance to try creative abuse to several types of cable and look for failures. 26 attended (13 were AES members), including AES Executive Director Bob Moses. The meeting was also webcast (a simple webcam and room mic) for anyone unable to physically attend.

Elections

This was the PNW Section Business Meeting, so its election ballot was presented. With a quorum present and no motions for write-in candidates, Chair Dave Tosti-Lane asked for a show of hands to accept the ballot by acclamation. All yeas, no nays, the motion carried.

2013-1014 PNW new electees:
Chair - Dave Tosti-Lane
Vice chair - Steve Malott
Secretary - Gary Louie
Treasurer - Greg Mauser

Committee 1 - Bob Smith
Committee 3 - Scott Mehrens
Committee 5 - Daniel Casado
Committee 7 - Dan Mortensen
Committee 9 - Dr. Michael Matesky
Committee 4 - Lawrence Schwedler

The Meeting/Workshop

As is our custom, PNW Chair Dave Tosti-Lane had everyone briefly introduce themselves. Dan Mortensen introduced the day's activities and what led to this meeting. Dan recently encountered a problem on the job, with a continuous white noise sound emitting from the FOH system when he was positioning his CAT 5 cable transporting digital audio packets between the snake head and console. The positioning involved pulling on the cable while it was somewhat constrained by other, large cables causing both a stretching and pinching of the cable. The noise went away when he stopped pulling. This aroused his curiosity.

Steve Lampen of Belden presented a detailed history of "CAT" cabling and Ethernet, along with many obscure facts and details of cable construction.

Next, Dan showed photos of the different twist rates of different pairs in several brands of cable. Steve noted this was all "secret sauce" by manufacturers to meet the stated "CAT" specs. Steve also showed an example of an interesting "instant snake" adapter for use with CAT 5/6 cable (ETS InstaSnake), and he showed Belden single pair CAT 5 cable, which his bosses thought was a dumb idea, but is a big seller.

Kurt Denke is the owner of Blue Jeans Cable in Seattle, and a vendor of assembled AV cables for home and industry. Blue Jeans is a Belden customer and can actually sell "made in USA" network cable assemblies, although he also sells some import AV cables. He demonstrated and spoke about cable termination and RJ-45 connectors, and performance testing. Steve noted that Belden has the patent on bonded pairs in CAT cable, and he'll send you a Belden pair splitter gizmo if you email him.

Mac Perkins (Pacific Studios) joined in and the three discussed terminology of CAT cable, such as permanent links, patch and channel cables, and how testers treat each differently. Field-installed RJ-45s usually won't meet CAT 6 specs, said Mac. Kurt continued with info on the $9k Fluke Networks cable testers, its adapters and testing criteria for different cable types.

A lunch break was held, with attendees free to find their own lunch and bring it back for socializing. After lunch, some door prizes were awarded:

-Mackie zippered accessory bags, courtesy Rene Jaeger/Mackie - won by Aaron Gates and Brooks Shepperd.
-Meyer Sound Labs tote bag, courtesy Dansound Inc, won by Rick Rodriguez
-CAT 5 cables after today's final stress tests, courtesy Belden - won by Steve Willkins, Gary Louie, Greg Mauser and Lindsay Smith.

Continuing after lunch, Mac Perkins spoke and demonstrated CAT cable testing instruments. He described TDR (Time Domain Reflectometry), a time-tested technology where a test signal is applied to a cable and precise measuring shows locations where impedance changes and discontinuities occur in a cable, and showed his vintage CRT Tektronix 1502 TDR tester. Also shown were several Fluke Networks testers (about $9k), and less expensive testers from Fluke Networks ($400) and Greenlee ($200). The high-end Fluke unit performs quickly, automatically and does extensive cable tests which can fully certify a cable to industry specs. Apparently, a simple continuity tester is considered next to useless!

Steve Macatee (Rane) joined in the discussion, with comments about Ethernet, latency and transport timing with audio data protocols like SuperMAC, AES50, Dante and AVB.

Lastly, for the rest of the afternoon, some space was cleared so that attendees could physically abuse many types of cables while connected to digital mixers and digital audio snakes, looking for data interruptions. Abuses included extreme pulling, yanking, rubbing with live power cables, stomping on, and even driving a 2 ton stage scissor lift repeatedly over a cable. This torture testing did lead to some discoveries that were not always as expected. Some extreme forms of abuse did not induce data disruptions, and some cable constructions did prove to be better against abuse.

Moderator Mortensen summarized his conclusions:

THESE CONCLUSIONS MAY ONLY APPLY TO THE SUPERMAC (AES-50 compliant) PROPRIETARY POINT-TO-POINT DIGITAL AUDIO CONNECTION TECHNOLOGY; relevance to other Ethernet-cable-based audio transport is neither proven nor alleged, except that measurement device results (in this case from a Fluke Networks DTX-1800) will say whether or not a cable passes acceptability testing for use as an Ethernet CategoryX cable . We used Behringer X32 Digital Consoles and Behringer S16 Digital Snake heads for our tests.

Data was considered disrupted if the console data screen flickered, and/or any anomalies were heard in the audio outputs.

  1. We were able to disrupt the data transmission of every cable we tried, including Cat5, Cat5e, Cat6, Cat6a, whether shielded (screened) or unshielded (unscreened) EXCEPT for the Belden 10GX32 Cat6a U/UTP. However, that cable does not have a robust outer jacket, and may not be suitable for portable live sound usage.
  2. The data transmission disruptions we were able to achieve at our event were of very short duration, as opposed to the disruptions I experienced on one occasion inadvertently and on another occasion on purpose with a specific 200' piece of Belden 7923A U/UTP. Those disruptions lasted as long as I kept pulling the cable and stopped only when I stopped pulling on the cable, and started again immediately when I pulled on the cable and stopped when I stopped pulling. Those disruptions also generated bursts of white noise from the audio outputs which lasted as long as the disruptions, and the data screens on the consoles blinked and sputtered simultaneously for the duration. I failed many other times to disrupt the data transmission when trying, using the exact same piece of cable. The disruptions at this event caused the screens to momentarily blink, and created "pops" from the audio outputs. I have no explanation for the discrepancy.
  3. The physical force required to disrupt the data transmission at our event was significantly greater than what I exerted on the two previously-described occasions. I have no explanation for this.
  4. The forces required to disrupt the data transmission are a combination of squeezing/pinching the cable while simultaneously pulling it through the pinching location. Objects used to create the pinching did not seem to be material-specific; we used parallel electrical wires and a wooden broom stick handle parallel to a length of metal conduit, all squeezed with one or two person's hands.
  5. We were also able to disrupt the data transmission by having 20 or more people stomping on the same piece of Belden 7923A cable while it laid on a flat hardwood floor. After that, we were able to disrupt the signal by having one person stomp on the cable on the floor while the other people were not on it, but another person trying the same thing could not disrupt the data transmission. We did not try this technique with other cables. Note that this specific piece of Belden 7923A registered "PASS" on the Fluke Networks DTX-1800 meter AFTER the abuse testing.
  6. The disruption of the data transmission can occur within a single pair, which we determined by stripping the jacket from the middle of a C2G CA19378 Cat5e U/UTP 150' cable and pinching and pulling each pair separately. All pairs exhibited the same behavior while disrupting data.
  7. We wrapped an inexpensive Richland E257448 Cat5e U/UTP 25' 40+ times around a 12" length of black-iron water pipe, and it still registered "PASS" on a Fluke DTX-1800 meter. From this, we conclude that CAT5e cables are likely to be unaffected by casual contact with or juxtaposition to random ferrous objects.
  8. Placing a relatively intense source of RF interference (AC cable to and from a barely-choked lighting dimmer and its 1KW lighting load) near, on, and across a Cat5e cable did not interfere with the SuperMAC data transmission.
  9. Things we learned from our Experts:
    1. The solid copper wire from high quality manufacturers in today's CATegory cables is vastly superior to previous generations of solid copper wire in terms of resistance to repeated flexing. Degradation of wire due to normal repeated use is much less of an issue than we'd expect. Low quality cable could still have issues.
    2. [Original text] From an American perspective, the European obsession with shielding cables is overblown, and a response to a problem that has not been demonstrated to exist at a significant enough level to need a solution. Unshielded cables work just fine in almost every application, except for specific environments with very high levels of specific RF interference. The vast, vast, vast majority of users will not benefit from using shielded cables, nor will they suffer ill effects from not doing so. That said, there are specific legal requirements to use shielded cable in certain situations in many legal jurisdictions, and the wise user will pay close attention to those requirements. [Update-11/2017] As a result of subsequent original research by Brian Wynn and later Dan Mortensen, consoles that use AES50 data transport need shielded CAT cable with Ethercon connectors to avoid being victimized by static electricity discharge (ESD, Electro-Static Discharge), such as happens in dry climates or in the winter when walking across a carpet and then touching the console, stage box, or microphone. Doing so without these precautions (STP with Ethercons on both ends) can result in a disruption of the sync between console and stage box which causes audio dropouts. This precaution is not necessary when using the consoles' analog inputs, as they seem to be immune from ESD problems.
    3. The RJ-45 connector was designed for a very different usage than with Category cables, and as the desired network speed increases into the future, it will be unsuited for those fast, fast speeds. At the existing ragged edge, it may be already unsuited, and performance may suffer.
    4. Many RJ-45 connectors are unsuited for use with shielded connectors, even those apparently designed for the purpose, and the grounding connection, especially for shielded cables without a drain wire, seems tenuous at best when compared to a soldered connection.
  10. To extrapolate from #9A above, we drove over the middle of a 137' Belden Brilliance Video Low Skew RGB7988P Cat5e U/UTP cable, placed on a solid hardwood floor and connected between console and snake head, with both of the front solid rubber tires of a scissor lift having a stated weight of over 900# per wheel, so there was damage in two places on the cable. This driving over was repeated at least 20 times. It permanently changed the shape of the cable from round to very, very flat, but did not disrupt the data transmission at all, and the damaged cable still measured "PASS" on the Fluke Networks DTX-1800 meter when removed and measured. This passing grade says nothing about the life of the cable, but it does indicate that you are unlikely to have immediate data transmission disruption from driving over the cable with a vehicle. We were going to test another cable by driving over it with a 10,000# GVW 14' box truck's tires, but felt that the scissor-lift test showed that we were unlikely to create damage using inflated rubber tires.
  11. Things we forgot/didn't get to:
    1. All Ethernet circuits balance each twisted pair, and the balancing is always done with surprisingly high quality balancing transformers on each end of each pair, thus providing galvanic isolation between the devices at each end.
    2. We intended to (but didn't) test the effect of having vastly different yet realistic ground potentials on
      1. the AC supplied to the console
      2. the AC supplied to the snake head
      and then comparing the effect on data transmission/audio quality output when the two devices are connected by U/UTP and then F/UTP cables.
    3. We only used the Behringer consoles to examine data disruption. We did have a Roland M400 digital console and snake head present, but
      1. there was no one advocating its use,
      2. I was told it exclusively uses a crossover Ethernet cable,
      3. the disruptive physical forces we were using could permanently degrade a cable,
      4. I didn't want to wreck the cable that came with it or spend time figuring out how to use one of our sample cables,
      5. ultimately time disappeared before we could test it. There were no other Audio-Over-Ethernet-Cable devices present. "

Special thanks to presenters Steve Lampen (Belden), Kurt Denke (Blue Jeans Cable), Mac Perkins (Pacific Studio), Rick Rodriguez (Fluke), who provided a handout of the essentials of CAT cabling for digital audio, and Cornish College of the Arts for making the Raisbeck Performance Hall on the Cornish campus available for this event.


Reported by Gary Louie, PNW Section Secretary

Revision History

6/17/2013 - creation 7/4/2013 - added cable twist photos 5/14/2015 - fixed busted URLS caused by move to media server. 6/30/2015 - add link to part 1 video, change link to banner (dtl) 7/28/2015 - add link to part 2 video, (dtl)
Last Modified 11/20/2017 19:20:59, (dtl)