In this Power Quality Newsletter:
• Upcoming power quality seminars in Texas and Oregon
• Answers to shield grounding question
• Revisions to two standards: IEEE 1459, and SEMI F47
• PSL's Power Instrument Museum
• The Leonardo Power Quality Initiative

------------ Upcoming power quality seminars in Texas and Oregon -------------------
I will be teaching my annual seminar on practical power quality on Wednesday, January 25 in Austin, Texas, and Thursday, February 23 in Portland, Oregon. This year we're emphasizing harmonics, process stability, and some new ideas about voltage sags, all with a very down-to-earth hands-on approach.

If you're an engineer and you're interested, please sign up soon - the seminars in November and December sold out, and we had to turn some engineers away. More information at www.PowerStandards.com on the right side of the page.

The Austin seminar will be at the University of Texas - if you're one of Prof. Mack Grady's students (past, present, or future), send me an e-mail and we'll get you in for free, including lunch. Despite what you may have heard, there is such a thing as free lunch. Not often, though...

------------- Answers to shield grounding question ----------------
In my last newsletter, I asked for opinions about which end of the shield on a signal cable should be grounded: the source end, or the load end. Eighteen engineers from 12 different countries offered interesting ideas.

• Several engineers argued that the cable shield should be grounded at both ends - especially engineers with experience at electric power substations - for the following reasons: high frequency radiated interference from arcs, cell phones, etc. are not attenuated by single-ended grounding; grounding both ends effectively introduces a common-mode choke around the signal conductors. (Marc points out that this function used to be performed by ferrous conduit, but that modern plastic conduit has made the problem worse.)

• On the other hand, there are engineers who argued that cable shields should only be grounded at one point, especially at electrical substations (!) - otherwise the shield itself will carry earth currents between locations that are at different earth potentials. During earth faults, when a major power conductor accidentally touches the earth, these currents can rise to thousands of amps, and will quickly destroy any cable that happens to be connected to earth at both ends. One of the engineers suggested solving this problem by breaking shields on long cables into shorter segments of shield, with each segment locally earthed.

• Yet another argument: engineers in the telecommunications industry apparently ground both ends of shields, and they've been doing it this way for more than a hundred years, so who can argue with them? On the other hand, engineers who construct petrochemical plants always ground just one end of the shield, and their equipment works pretty well, too. And I'm told that the elevator industry, which runs signal cables long linear distances near power cables, follows the single point ground rule, too.

• An engineer with great practical experience wrote that initially he followed the "always ground the transmitter end" cook-book instruction until he constructed a steel mill in Australia in the 1970's, and found that his feedback signals to DC drives were swamped by noise. He then switched to the philosophy of grounding the end of the shield that is closest to the sensitive electronics, which fixed the problem, and he's stuck with that solution ever since.

• Bill Moncrief's answer was so clear, as usual, that I will reproduce it here: "To protect the interior wiring from having a current induced, you connect both ends and let current flow. That makes the inside of the shielded cable a net-zero space; an electromagnetic shield. To protect the inside wires from collecting a voltage from a transmitter, you ground one end and let the voltage appear on the shield. The inside wires are a net-zero space again from the electrostatic shield. And we used tri-ax cable in EHV substations to carry one-watt radio signals. Double ground the inside, single ground the outside, but I think that the order was just convenient."

• A different idea was suggested by a couple of design engineers, who work on signal grounding inside equipment: you must ground the source end, because for any signal there is only one source (but there might be many receivers). If you're going to ground just one end, you don't have any choice.

Thanks to everyone who wrote in, with special thanks for the very clear explanations from Patxi Pazos of Iberdrola (Spain), Gaetan Ethier of Hydro Quebec (Canada), Christ in Hong Kong, and Bredhold in Kentucky. My own conclusion from all this: there probably isn't a single correct answer. The best solution for shielding a cable that runs six inches between two circuit boards is probably different from the best solution for a 200-meter cable in a steel mill, so (a) find out what the conventional practice is for your particular application, (b) start there, and (c) change it if it doesn't work. I'm not really satisfied with this answer, but I can't come up with anything better.

-------- Question: Power failures during forest fires ------------
I've recently been dealing with questions related to power failures in Brazil, Korea, and South Africa that were caused by burning vegetation underneath transmission power lines (in Korea, these are referred to as "mountain fires"; in South Africa, it was part of the sugar cane harvesting process). Tony Firth in New Zealand writes that, like me, he had always assumed that these short circuits were caused by airborne carbon particles - something similar to the problem with bird droppings on high-voltage insulators. However, there seems to be an argument that the heat from the fire actually ionizes the air, and it's this ionization that causes the short circuits. Of course, once the arc starts, ionization would be the mechanism that maintains the arc (here's a movie of a rather frightening example of this phenomena: Ions - Alex McEachern.mpg). Question: Does anyone know the exact mechanism(s) that leads from a vegetation fire to a transmission line power failure?

------------ Revisions to two standards under way ----------------
We have just completed writing the draft 5-year revision for SEMI F47, the voltage sag immunity standard. The new draft replaces both SEMI F47 and SEMI F42. The sag depths and durations have not changed. Here are the major changes: It's now very clear that the Scope covers both semiconductor tools and their modules and components. The draft has a grandfather clause - equipment that was certified under old version of SEMI F47 does not need to be re-certified. The draft requires test method according to IEC 61000-4-11 (less than 16 amps) or IEC 61000-4-34 (more than 16 amps). This means you must have lots of current available, which rules out most electronic sources - you should use the PSL Industrial Power Corruptor for all SEMI F47 compliance testing. The draft now has clear pass-fail criteria, and includes a specific list of what information must appear on a Certificate, and what information must be in a Report. A lot of the Japanese reports and certificates lacked this information, so we added it to the standard. Section R1-2 of the draft has a list of common errors, including a well-intentioned but wrong voltage waveform generated by one Asian test lab.

At the moment, the revision to SEMI F47 is just a draft, so we will have to see whether it gets approved in early 2006. Let me know if you would like to make comments about the draft.

Professor Alex Emanuel is leading a fascinating effort to update IEEE 1459 'Definitions for the Measurement of Electric Power Quantities Under Sinusoidal, Nonsinusoidal, Balanced, or Unbalanced Conditions'. What makes this process so interesting -- at least to me -- are the serious, thoughtful disagreements among the experts who are participating, even about the definitions of single-phase parameters. We're just launching into the three-phase discussions now.

If ever you thought you truly knew the correct definitions of power flow quantities, listening in on this e-mail discussion might change your mind. Doubts and uncertainties can be enjoyable, and Alex Emanuel's gracious guidance of this electronic discussion is a model of balance: while allowing everyone to have their say, he moves the discussion forward with great kindness towards the goal of publication. He hasn't set a deadline for finishing this update, but work is progressing well.

----------- PSL's growing Power Instrument Museum -------------------
PSL (well, me, actually) has a growing collection of classic power instruments. Most of them are functional and fully calibrated - in fact, we even use the beautifully constructed 50-year-old Leeds & Northrup Kelvin Bridge Ohmmeter 1425, at left below, for calibrating some of our current shunts. The center picture shows a gift from Francois Martzloff: a 1920's General Electric surge recorder, complete with a wind-up clockwork motor for advancing the recording paper. And at right is a 3-kilovolt calibrated DC supply, with a beautifully designed vacuum tube circuit.
   
I like to think of these old power instruments as coming from four generations: the earliest are made by watchmakers, and look like it; then we get the dove-tail oak boxes, like the two at left above; then we get vacuum tubes; and finally we get to discrete transistors.

These instruments, in addition to being beautiful, are great teaching tools: the instrument designers here at PSL always enjoy taking them apart, mostly to learn what endures and what doesn't (some hints: batteries are almost always a bad idea, and integrated circuits can't be repaired. And avoid sharp bends in insulated wire - thanks, Georges!).

Thanks again to everyone who contributes these old stalwarts - I appreciate it, especially if the instruments are older than I am...

----------- The Leonardo Power Quality Initiative -------------------
There's a new association working on power quality in Europe, mostly with an educational approach, called the Leonardo Power Quality Initiative. It looks like a good idea, and I respect many of the participants, so PSL has joined in as a Reference Partner. Take a look at the web site - there's some useful information, and interesting web-based seminars. For those of you who have asked, this has nothing to do with the brief essay I wrote many years ago about Leonardo da Vinci's power factor correction capacitor. The name is pure coincidence.

In January I will be in Taiwan, Korea, Japan, and Texas - perhaps we will get to say hello somewhere. In any case, you have my best wishes for a happy, prosperous, and healthy New Year!

Alex McEachern
Power Standards Lab
1201 Marina Village Drive #101
Alameda, California 94501 USA
TEL ++1-510-522-4400
FAX ++1-510-522-4455