Analog

Well, being that I mostly work with the digital side of things and my professional exposure to analog electronics is minimal, I thought that I could fulfill that part of my experience via delving deeper into amateur radio.

Part of the problem with longer-distance radio communications is the need for lower frequencies (at least below VHF), which requires more expensive equipment and antennas.  Some people build their own antennas, but for the HF noob, perhaps pre-built ones are more accessible.  Buckmaster has an off-center-fed dipole antenna that covers seven bands, which may be where I should start.

One thing about ham radio is being able to find active frequencies.  To that end, I've poked at using Airspy with an RTL-SDR to find conversations to eavesdrop on.  Ultimately I'd like to get a portable, self-contained setup on a Raspberry Pi with a display to run all the time, or at least on a timer.  On that note, RPi projects for ham radio, I found a project site that has some useful information.

Lastly, while looking for SDRs, I found Analog Devices' learning toolkit for SDR, ADALM-PLUTO.  More interestingly, they have a 25 MHz oscilloscope device called the ADALM2000 that plugs into a USB port, for only $200, along with an analog electronics learning package.  That would sure be fun to poke at.

Decoupling Capacitors

Had occasion to read up on the appropriate way to add decoupling capacitance to printed circuit board assemblies, and found a useful explanation.  One conflicting element with what I've heard at work is that the paper mentions that power pins should be connected directly to the power plane via a via (ha!), whereas here we tend to put the cap near the pin connected by a trace which then goes to a via.  Plus our component pads aren't shaped optimally according to the paper.  Why can't it be simple?

Addendum [2015.03.15]: By the way, "Manhattan routing" is routing horizontal traces on one plane, and vertical traces on another.  "Manhattan distance" is the corresponding length of a trace routed this way.

PiPhone and Others

Some random RPi stuff:

• PiPhone (and its maker's site)
• The RPi B+ 
• Arduino-powered LED Lighting 
• 3D-printed Gameboy look-alike

Programmable Power Switch

I've been thinking about one of these for a while...need to be able to power off devices programmatically.  Some of the comments are useful.  To wit:

1. You might consider a "zero-crossing" detector however, as switching AC during the high portion of the sine wave will eventually cause your relays to pit, scar, and fail from the internal arcing. By sampling the AC wave and switching the relay on/off only during the zero crossing of the sine wave period - that will make your relay last for years.

2. You should add a base-emitter resistor, something like 10K. If you leave the base unconnected noise(50/60Hz) might turn on the transistor - and the relay. It happened to me a while back while switching some lights with a very similar circuit.

3. Try something like a MOC3043 zero crossing detector optocoupler. If you couple it with a triac you basically get a good switching method, similar to a SSR. Don't forget the snubber for inductive loads.

Need to get my workbench in order....

GPS over SDR

The thought occurred to me that perhaps the RTL-SDR dongle could be used to pick up GPS signals.  GPS L1 is 1575.42 MHz and L2 is 1227.60 MHz, which are within the range of the R820T tuner's 1700-or-so MHz.

Why, you ask?  Because an RTL2832U + R820T dongle is $10, which is a lot cheaper than most of the pluggable GPS modules out there (usually $40+).  If the base system is an SBC anyway, then this route is much more economical and flexible.  This would also have applications in an SBC-based telemetry system.

Wikipedia has the basic details on GPS and the signals used.  The C/A code is 1.023 MHz and the data is 50 bps, so when the SDR is tuned to the carrier this is well within the bandwidth of the RTL2832U.  Oddly enough, the R820T spec notes that it only supports up to 1002 MHz.

The University of Colorado has a useful GPS overview that includes more specifics, such as how to calculate the PRNs.

And of course there's already a guy who's been working on GPS over SDR, one Michele Bavaro.  Development seems to have stopped, though; and the GPS-SDR site has been converted into a lunch scheduler.

OSH Park

Was looking at the HopeRF mini-transceiver boards, and in the comments of a Sparkfun site, there was a link to a breakout board on OSH Park, which puts in daily orders to a USA-made PCB fab.  2-layer boards are $5/in for 3 copies of the board, with a 12 calendar-day turnaround.  Cool!

I'm going to need to learn how to layout boards....  Incidentally, the mini-boards are cheaper on eBay.

Displays for the RPi

I was bothered by the DSI port on the RPi that's going unused.  So I went looking for DSI displays and found via a forum that the RPi Foundation is working on one.  The guy who came up with HDMIPi did a preview.

Supposedly one will be able to use both displays at once.  Awesome!!

The "DSIPi" is to be 7" at 800x480, good enough for DVD video.  The HDMIPi is 9" at 1200x800; 720p, but not full HD.  I think the DSIPi will be good for a car video player project, but the HDMIPi will be good for keeping me from hogging the family TV!

Power Switches

Sparkfun, an online hobby electronics store, is selling an AC power switch that can be controlled with TTL logic, as well as the Sharp optoisolated solid-state relay itself.

I kind of wanted to run a low-power server to make available lots of storage which I could switch off to save energy when not being used.  It could also be useful for security for preventing sensitive data from being accessed, should the server be exposed to the Internet.

Raspberry Pi seems like an ideal candidate for such a system.