Sunday, March 22, 2009

Amateur Radio Licensing Exams

The ham radio sector of the blogosphere is all agog over the FCC's recent decision to deny KI4NGN's petition to expand the minimum question pool size from 10 times the number of questions on each exam, to 50 times. As best as I can figure, the argument for expanding the pool is to make it impossible for candidates to memorize the entire question pool, enabling them to pass the exam without actually learning anything. The FCC denied the petition because the petitioner did not provide evidence that the current practice was, in fact, leading to there being a large number of operators who did not possess the necessary skills to operate their stations correctly.

First, I'm one of those people who at least partially memorized the question pools, with enough success that I managed to pass all three elements at one go. I also publish study decks (derived directly from the NCVEC published pools) that, used in conjunction with readily available software, will assist others in using this technique to prepare for the exams, if they so choose. I don't have a problem with people using this approach; I want to see as many interested people in ham radio as possible. (Unlike some people. More on this later.)

Second, many of the commentators state that Mr. Mancuso's proposal would make it much harder for candidates to pass. That is something of a reach. While it would make memorizing the pools much harder (the Technician and General pools would be forced to increase to at least 1750 questions from their current 392 and 486, respectively, and the Extra pool to at least 2500 questions from its current 738), it would be entirely up to the NCVEC to decide whether the additional questions would actually cover additional material, or merely involve more permutations of the same material requiring no additional study.

Either way, however, it would have a dramatic impact on the NCVEC. The NCVEC, as the body currently charged with producing and maintaining the question pools, is, as far as I know, organized entirely on a voluntary basis. Accepting Mr. Mancuso's petition would have forced the NCVEC to come up with four to five times as many questions on each of the three pools; this would have strained the capabilities of what is presently an all-volunteer organization and would likely have required the NCVEC to hire staff, which would force them to solicit contributions from, or even charge dues to, their member organizations, which would in turn force them to increase testing fees. There are enough people complaining about the ARRL/VEC charging $15 per session; just imagine how loudly they'd screech if that went up to $30. And, of course, any increase in fees will incrementally exclude some candidates.

More importantly, however, is that Mr. Mancuso's petition underlies a misunderstanding of the purpose of the licensing examinations. The exams do not, by any reasonable standard, test candidates on whether they have the skills they need to successfully operate an amateur radio station. Some of the questions are, in fact, somewhat relevant, but a passing grade on the current exams, whether obtained "honestly" through actually learning the material, or merely by memorizing the pool, does not in any way insure that the licensee has any clue how to successfully operate their station. There is no practical test of operating procedures anywhere in the licensing process.

And that's a good thing. Why? Because such a test would invariably involve subjective judgments by the examiners. The nice thing about the examination structure we have now is that the tests are objective. The answer is right, or wrong, and there is little wiggle-room for a VE to fail a candidate for "inappropriate" reasons. There's no real way to do a practical operation test that doesn't involve subjective judgments by the VEs on whether the candidate passed or not, and as soon as you allow subjective judgments you allow for the possibility of prejudice. And that's something we just can't afford to have.

And anyway, the purpose of the tests isn't to ensure that every ham radio operator is competent at electrical engineering. The purpose of the test is to filter out people who aren't willing to take the time (in some way or another) to learn the minimum material required for the tests. The idea is that we put on the test material that we want newcomers to our hobby to be at least passingly familiar with, so that they will learn at least some of it, and by passing the test they demonstrate at least a passing commitment to learning these things as well as an understanding that our hobby has rules that everyone is expected to follow. For this purpose, the size of the pool is almost completely irrelevant; all that matters is that the questions cover the range of material that we want newcomers to be exposed to.

In short, the exams are, pretty much explicitly, a barrier to entry. Barriers to entry are always a challenge. Set them too high and you don't get enough participants; set them too low and you get people who lack sufficient attachment to the common goals and purpose of the community to feel bound to follow its rules. I think for the moment we set a pretty good balance on this issue with the current examination practices, both with the Technician (entry-level) license and with the somewhat harder General license. The more rigorous study required for General increases the likelihood that the licensee will have come to understand the importance of following the rules, which is more important for licensees with access to HF because of the worldwide propagation and much more restricted spectrum to share.

The much higher requirements for the Extra actually make sense because this is a "prestige" license; not being able to get Extra doesn't substantially exclude the licensee from much of anything (mostly, access to short callsigns, access to some of the more valuable contesting spectrum, full privileges as a volunteer examiner, and broader reciprocal privileges when traveling abroad), and so making this license substantially harder will limit it to those who show substantial commitment to the hobby, which is exactly what we want.

In my opinion, the testing system we have now is pretty close to the best one we can really hope for. We could lower standards further in the hopes of getting more hams, but I suspect we'd not get that many more active hams, just more people with licenses who don't actually use them. We could "increase standards" further by increasing the number of questions on the tests, increasing the size of the pools, or broadening the content being tested for, but that runs the risk of merely excluding people who might otherwise become licensees, without actually improving the competency of current licensees. We could institute "practical testing", but that introduces a huge opportunity for prejudicial administration of the examinations that would exclude people from the hobby for illegitimate reasons. It's certainly important that the NCVEC consistently revise the pools to ensure that the questions asked continue to expose licensees to the issues they need to be aware of, but I do not believe that any significant change to the process as it exists now would materially benefit the hobby.

Other posts on this topic that may have inspired this one:


Saturday, March 14, 2009

Running a PC off batteries

This (well, actually next) month's QST has an article on DC-to-AC inverters targeted toward the specific purpose of running a shack computer off the shack battery backup supply. While DC-to-AC inverters can be useful, there's a better solution to this problem and the QST article completely ignored it: replacing the PC's supply with a direct DC-to-DC converter.

Powerstream (a company with which I have no experience or connection) offers several DC-to-DC power supplies in the ATX form factor for very reasonable prices. Laptops and other devices with "brick" or "wall-wart" supplies typically have low-voltage DC inputs that can be accommodated either by adapting a car adapter (recall that 13.6 VDC "shack" power is essentially the same as automotive accessory power) or by homebrewing a DC-to-DC converter (buck/boost converters are one option for this). About the only devices that cannot be easily run by a DC-to-DC converter are CRT monitors and laser printers. CRT monitors should be replaced by LCDs anyway as the latter have far lower power consumption, and laser printers have such high startup currents that they probably can't be run off a DC-to-AC inverter either unless the inverter has a very high surge capacity. I've known more than one IT operation that had UPS problems caused by plugging a laser printer into the UPS.

Of course, most DC-to-DC converters use pulse-width modulation to generate the desired voltage(s), and hams will have to be careful to ensure that any RFI generated by such approaches is managed with care, and it might be difficult to do this. My main beef with the QST article is that it presented DC-to-AC inverters as the solution to the "how to power a computer off batteries" problem without even mentioning in passing that there are other solutions that at least deserve some consideration.

Wednesday, March 11, 2009

Battery backup for ham radio gear

The second issue I heard on last Monday's TechNet (a weekly net held by the club I'm a member of) was on the topic of battery backup for repeaters and other ham radio gear.  This issue was raised by Ryan (KC9OFF), and as it happens his specific situation is not going to be that interesting to hams as he is trying to provide power to an Icom FR3000 series repeater, which is not ham gear (it's business band gear) and that particular repeater already has a built-in battery controller so setting up a backup battery on it is just a matter of connecting a battery to the terminals on the repeater in the correct polarity.  (I really should charge him a professional fee for researching this for him, but hey, whatever.)  However, us hams tend to use gear with slightly fewer "creature comforts", which means we have to roll our own (or at least put together from components) power controllers for such thing.  Not that that's a problem: rolling our own  is part of what we do, right?

During the net I asserted that there was an article in November's QST on battery backup systems for ham shacks; fortunately, I was right: the article in question is on page 76 in the regular "Getting on the Air" column.  That column describes two approaches to battery backup.  The first approach is use an inline charger and a boost/buck regulator to provide regulated 13.8 volts to the radio no matter what voltage is provided by the charger/battery combination (which could be as high as 16 volts when the charger ramps up, or as low as 10 volts as the battery runs down).  The second is to use a switching power controller that has separate inputs for charging power and the battery; the difference is that the power controller will not draw from the battery unless the main power is down, while the inline design draws from the battery and the line power (charger) to the extent that each can provide power at any time.

In the inline design, you will likely draw from the battery while transmitting, because the charger probably only provides 5 amps at most and cannot provide the full 20 amps or more that a typical rig pulls while transmitting.  As long as your transmit duty cycle is low enough, the charger will have enough time to recharge the battery between transmissions.  The main problem with boost/buck regulators is that the majority of regulators out there generate "noisy" DC; virtually all of them use pulse-width modulation for power control, which generates RF noise and will require significant filtration to avoid RFI issues.  The switching power controller does not do this, but will present a variable voltage to the radio when switchover occurs and as the battery discharges (unless it also contains a boost regulator, in which case all the issues with those again arise).

Several people on the net noted the need to include protective hardware (mainly fuses and diodes) to prevent the battery from draining through the power supply and to protect against various sorts of possible failures that could lead to shorts.  The inadvisibility of using an ordinary car batteries (which are optimized for short high drains, not for extended low drains) was also mentioned.  Car batteries should never be used for backup power.  There are a variety of deep-cycle options for this purpose which are designed for long-term moderate-power draws.  Many are also engineered to avoid venting hydrogen gas, which would create an explosion risk when used in an enclosed space.

The solution I'm still looking for I haven't found yet in commercially available gear, and involves using solar as primary power and a battery rack as backup during the night, with line power being used to charge the batteries only if they drop below the point that they aren't expected to recharge in the sun the next day.  I think I've seen a controller that supports multiple power sources, so if I put the solar panel on one input and a line supply on another, but use a switch/relay to only provide AC to the line supply when the battery voltage is below the threshold, (with some hysteresis to avoid chugging) then the system would perform in the manner I desire. 

Monday, March 09, 2009

MFSK Idle Tones

In tonight's TechNet (a weekly net held by the club I'm a member of) one caller (Ed, K9EW) asked about the idle tones that various MFSK implementations, given that he's noticed that different MFSK generators seem to generate different idle patterns. No clear answer was given on the net, although the discussion came to the conclusion that it was, at least to some degree, up to the programmer to decide how to do this.

The ARRL publishes formal technical specifications for some common digital modes, including MFSK. According to section 3.8 of this specification, compliant MFSK implementations must inject a non-printing character every 20 symbol periods, in order to avoid sending an extended period of a single tone. The standard does not specify which nonprinting character must be sent (although ASCII NUL, 0x00, is suggested). A programmer could elect to send any of several nonprinting characters, which would result in different bitstreams.

The reason for not allowing a extended single tone is that this would allow the symbol clock between the sender and the receiver to desynchronize. The specification requires that receivers maintain clock for at least 50 cycles, so sending a character every 20 cycles guarantees that sync is maintained as the receiver will resync its clock with each "diddle".

For more on the technical specifications of digital modes, see fldigi's excellent Digital Modes page.

Friday, March 06, 2009

The Right to Receive: Police Scanners and the Internet

Last night, I ran across this entry on a blog for the communities of Clayton and Concord in California.  In this entry, the publisher of the blog announced his intent to put up a live internet stream of his scanner, which is tuned to receive various police frequencies in the "ClayCord" (Clayton/Concord) area.  The comments on the thread are, to be certain, very interesting, as well as rather disturbing.

The idea of streaming scanners is pretty well established.  There's a whole organized system for picking up live ATC streams (LiveATC.net), which got some significant attention because the site was able to produce a recording of the communications related to the crash of Continental 3407 in Buffalo within hours of the crash.  There are also quite a large number of sites offering live and recorded police radio streams for much of the United States. 

And apparently the police aren't too happy about this.  Several states have laws that attempt to limit people's right to receive police radio, even though the Communications Act of 1934 granted a virtually unlimited right for any person to receive any signal they can hear.  Apparently the police don't want people listening to their radio signals, and rather than taking the obvious step of encrypting them to prevent people from usefully listening to them, they prefer to try to make it illegal, and when that fails, harass those who do listen to them anyway and make it easier for others to listen to them.

Some of this is evident in the linked thread: witness the several comments (probably all from the same person) declaring that the "FCC has regulations that prohibit streaming a police scanner" (they don't) and declaring other laws that make this sort of thing illegal, or which create criminal liability for the streamer should someone use their stream in the commission of a crime.  Why, one wonders, are the police so insistent that the public not listen into their conversations?

Then there's the recent issue with the Indianapolis Metropolitan Police Department, which recently got itself in hot water because citizens (reportedly, amateur radio operators) recorded IMPD officers using profane language to exchange vulgar and inappropriate comments, using frequencies the police do not have permission to use.  The use of unlicensed frequencies is especially concerning; the use of "unexpected" frequencies for such inappropriate commentary suggests that the officers in question knew that what they had to say was improper, but thought that, because it was being said on a "nonofficial" channel, would go unnoticed by supervisors and observers and thus not lead to repercussions.  Unfortunately for them, there are people who listen to everything, and someone took enough offense to report it to the FCC.

I've heard rumors that the FCC is withholding the identities of the people reporting the unauthorized transmissions because there is a real concern that the IMPD, or members thereof, will seek retribution against the reporters.  The fact that they were reported as "amateur radio operators" is especially interesting because of Indiana's mobile scanner law, which explicitly exempts amateur radio operators; by identifying the reporters as amateur radio operators that makes it impossible for IMPD to demand disclosure of identities so they can investigate whether Indiana's mobile scanning law was broken.  Those of us in Chicagoland know just how bad it can be when you cross a cop: just ask Mike Geinosky, who has received 24 parking tickets in 16 months, several of them for a car he no longer owns.  And that's mild for Chicago.

The clincher in that ClayCord thread was when the anonymous coward responded to my post explaining why the ECPA doesn't prohibit republishing police radio transmissions with posting somebody's address (not mine; I don't live in Concord) and encouraging people to go there and "protest" me.  Fortunately, someone (probably the blog owner) quickly deleted that comment.  Pretty blatant attempt at intimidation, there. 

So, while there's little question in my mind that while police scanners can be used for socially unredeeming purposes, the socially positive purpose of allowing the public to better keep tabs on how well the police (who are supposedly their servants) are performing their duties overwhelms that.