μblog: engineering from the trenches

It looks like Siemens Energy and Automation has a has made their Seimens Technical Education Program (STEP!) courseware available (PDFs). Topic range from basic electronics, to motor control, to power distribution. What really caught my eye was the abundance of easily understandable safety information in the power distribution course section. The surge protection section is easy to understand and seems to contain enough information on properly wiring a residence to avoid certain catastrophes. Basically, this place seems like a good place to start for those who would like to do some minor work on their home’s power system but don’t have the required training. (This is not to say that STEP has all of the required training either!)

On a side note, I am in the middle of writing a document up to overview the Yagi antenna. I am more than half of the way done, however, I am going to the Neural Interfaces Conference next week. Consequently, the full version should be up next Thursday. If anyone else is there and has a desire to meet up then send an email.

del.icio.us | digg

Whenever I get a shiny new copy of Electronic Design magazine, I eagerly read the section called Pease Porridge, at the very end of the magazine. This section is written by Bob Pease and typically contains whatever material Mr. Pease decides to include that month. In my view, this is the only page in the magazine that actually includes electronic designs and is in contrast with the rest of the magazine’s content (advertisements that look like written text).  Sometimes there are designs for circuits that detect dropping temperatures in boots, sometimes something on taxes, and sometimes the section contains responses to letter (or emails) that readers have sent in. This month’s porridge dealt with sent in letters and the first one, by Anonymous Engineer, has provoked some thought.

Hello Bob,

A note concerning electric cars and plugin hybrids: Consider that politics has little to do with engineering and/or science. It only pays lip service at those altars. So, somebody has to do serious planning for the immediate future.

I’ve been working on some serious battery-charger designs. One of our planners (an engineer) did some research in good old California. We learned that your utility companies have problems with metering even small numbers of plug-ins, nor can the California infrastructure absorb many cells of even 5000 plug-ins. You’re running at about 81% capacity, and without a smart meter and control, we would easily overload the electricity capacity on two peaks every day. That’s not politics. It’s business.

American, Japanese, and European manufacturers were contacted, and none of us can really do this without the cooperation of the utilities. Oh sure, we can sell a few and look green. The press wouldn’t even know who to blame when the grid broke. Some of our competitors have been doing that, but without that smart meter, it’s the wrong thing to do. They know it. But that’s business.

All of the automakers easily agreed on the meter and protocol. The utilities did not. They already have contracts on meters that aren’t smart. That’s business. The real greenies were there, too. They’re part of politics. They want California to be energy neutral in 10 years—sorry, no data on how to do it. Industry must be hiding it.

Back in the Midwest, we run about 1/2 the total power per person aggregate (at 740 W per person continuous) than you do out west, but that’s because of our low transport and air conditioning costs. Perhaps just targeting our levels would be a better starting point. There is no magic bullet in the next 10 years. So the utilities answer to the greenies and business, not to the engineers. Our charger is going to be great. We will get patents. It will be used all over, but not in volume in the west.

-An Anonymous Engineer

Prior to reading this, I have been all for plug-in electric vehicles, although I have never worked out the power load requirements to support such a population. Now thinking a of the power blackouts that crippled the North-Eastern United States a few years back, I am no longer sure that switching privately owned vehicles to plug-ins is the right thing to do right now due to obvious issues with the power infrastructure in the U.S. Although the need to address environmental issues related to automobiles is apparent, I doubt there will be a serious motivation behind it unless there are economic benefits involved. The switch over to electric (or hydrogen for that matter) vehicles should start with the sector that can be converted most efficiently which would allow for the highest return on investment.

In my mind, the public transportation system should make the switch first. I am mostly experienced with the Metro fleet in Washington D.C. and can note that all of the buses are made by Orion Bus Industries (owned by Daimler Trucking) and fall into two models (larger and smaller). Given that the buses operate on a pre-determined route and are equipped with GPS (to automate the current location display inside the bus), the energy demand per vehicle per day can be predicted with high accuracy. Furthermore, most of the vehicles are at the depot during the late night and early morning, when electricity is cheaper than the daytime, which would make for an ideal charging period.

Over all, this would lead to a very predictable off-peak energy demand that can be negotiated with the local power generation utility in advance to ensure that the power demand is met economically and without sacrificing the power grid’s integrity. An on-site generator may be employed to charge batteries in case of a power emergency where public transportation service is required. A more clever (and more difficult) design may involve modifying the buses to have a modular power system that can be swapped between conventional diesel and electric power-plants.

Again, most of this is mostly my intuition as I have not worked out most of the mathematics behind it, however, it seems that power demand predictability might be a mitigating economic factor for public transportation conversion. Unlike us, the bus driver rarely gets sleepy while reading at night and takes the bus out for a spin to the coffee shop.

del.icio.us | digg

“…hello, is this tech support?”

Unknown to myself until today, it appears that the United States Bureau of Reclamation operates many of the nations hydroelectric power plants. As such, they also make available the Facilities Instructions Standards and Techniques (FIST!) manuals which are a wealth of interesting information. While I was initially looking around to find some good information on synchronizing dynamo generators, I think that their Abnormal Operations Generic Technical Guidelines for Power Stations manual is quickly becoming my favorite. The most interesting part starts with section 4.0 (p. 12) where the bureau outlines generic, step by step, guidelines for responding to various abnormalities in the plants operation. Although the contents are fairly straightforward and the messages clear, the name of the document is quite formidable. Be sure to check out the rest of the FIST manuals if you like this sort of information.

( vol1-12.pdf )

del.icio.us | digg

As some of you may have known, I am a bit of a FreeBSD enthusiast and have been using it regularly since around the time of the Linux fragmentation/teardrop vulnerability. I have always kept a FreeBSD system running at home so that I could perform various network tasks and automate things. In an effort to conserve power and reduce noise, I have been using my trusty Sony z505sx (pictured above). Recently, I switched to a Jetway MINI-ITX board and have been quite happy. The board runs a 1.2GHz Via C7 Eden processor, is fanless and consumes only a few watts of power while operating. It is loaded with 1GB of ram and a 750GB SATA (also low power) drive to provide ample storage. Since the system has RCA/S-Video out, I added a 5″ LCD screen on the top. So far, the machine performs well running as media/web/ftp/ssh server and is barely audible with no fan attached.

I am manufacturing a custom case for this machine and will post a full writeup once I have a cover that I like, for now, you will have to enjoy the single image below. As far as the z505sx is concerned, I decided to put it on eBay in a preemptive spring cleaning effort and to avoid clutter.

del.icio.us | digg

In analog circuits, it is not uncommon to see noise at 60Hz and harmonics of 60Hz. While some portion of this noise, especially in sensitive equipment, can be attributed to pickup of local electromagnetic fields, there is also an issue of noise coupling in through DC power supplies. Even the best regulators still have a finite power supply rejection ratio and let in a little bit of noise from the power lines. The problem is compounded when power circuits are specifically designed filter 60Hz noise but are unable to cope with higher order harmonics.

Reactive/non-linear loads are the core of the problem when it comes to many 60Hz harmonic noise problems. The supply cables have finite impedance, therefore, any harmonics on the current load will also be reflected on the delivered voltage, and vice versa, resulting in some level of harmonic contamination for all of the devices sharing the particular power circuit. A simple example is figure above. When the capacitor bank is charged, the bridge rectifier diodes will only conduct for the brief period of time when their output voltage is slightly higher than the capacitor voltage resulting in current spikes as indicated. Since the current spikes are much narrower than the rectified voltage signal, they will contain higher frequency harmonics.

This is where power factor correction and better design can come in to save the day. The primary idea behind power factor correction is to remove any apparent reactance or non-linearities from the load and make it look like a plain resistance. Another key feature of power factor correction is maximizing the power delivered to a device as the real power, related to the in-phase components of the voltage and current supplied, is what typically does the work in the system. Although this second feature is very important over all, it is rarely important to neurology related electronics as they are typically based on sensitive amplifiers and not heavy electrical loads. (The obvious counter example would be an fMRI machine.)

In any case, here is a short (11 page) primer on power factor correction from Fairchild Semiconductor followed by the more substantial (208 page) power factor correction handbook from On Semiconductor.

( an-42047.pdf ) ( HBD853-D.PDF )

del.icio.us | digg

If anyone has had their computer crash due to very brief power flicker, you can sleep better knowing that there Information Technology Industry Council is working hard to keep you on-line. I started reading a book on power delivery and quality and have found this figure above, the CBEMA/ITIC curve. The idea is that power supplies and systems in “information technology” are designed to comply with certain voltage tolerances, or rather tolerate certain deviations from the normal 120Vac@60Hz (in the U.S.A.) for specified amounts of time. For example, this type of equipment is supposed to tolerate a total loss of power for one half of a cycle without interruption. Conversely, the same equipment is supposed to tolerate a voltage surge of 500% of the nominal value for 1% of a total cycle period.

After looking deeper, I also located a SEMI F47 power standard which applies to manufacturers of equipment and subsystems destined for semiconductor manufacturing. This standard is not as rigorous as the one above and is only specified for voltage sags. It can be seen that this type of equipment must be tolerant of a reduction of line voltage to 50% of nominal value for up to 0.2 seconds on a single phase system or on one phase to neutral for a three-phase system. A presentation covering updates to the SEMI F47 standard is available here.

(  voltage_tolerance.pdf )

( semi_standards.pdf )

( 4overviewofsemif47-0706.pdf )

( iticurv.pdf )

del.icio.us | digg