Copasetic Flow

  Squigily audio in our homebrew ham radio rig? Capacitors! This extends the capacitor post from last week. This week, while adjusting the Tuna Topper amplifier bias after a run-in we had with an unexpected rain shower, I put capacitors across the dit and day switches on the keyer. The dit capacitor was particularly useful because with the amplifier bias set above a certain level, the keyer chip would immediately reboot when the dit key was pushed. I'm assuming RFI transients were punching through to the little processor, befuddling it in the process. Today, the kid—Diaze—and I changed the battery from a D cell pack to a LiFePO4 battery. The extra bit of punch provided by the nicer battery meant we had to turn down the bias on the TunaTopper's onboard pot just another quarter turn or so. The resulting keying sidetone over the headphones was less than esthetically pleasing though. When Diaze sat down to key out a CQ, the buzzing and squiggly sounds were too much for her. A rec

 I've alluded to Project TouCans over the last several posts, but I've yet to introduce it. Today is the day! Project TouCans is the new, very compact, packaging of the RockMite  20 with an attached TunaTopper amplifier suspended in a halfwave dipole antenna. Here's the design drawing Diaze and I made In short, both components are suspended from the top of a pineapple can. The tuna-sized can the Tuna Topper shipped in is flipped over and used as a rain cover and antenna mounting structure. The antenna is mounted to the red and black banana plug binding posts at the top of the assembly. You saw a picture of the latest incarnation of TouCans in yesterdays post. Here's the original—and more colorful—project in its first incarnation The Ethernet port is the same one used by the Rockmite in the Flying Rockmite—for that matter, the radio is also the same one used in the most recent flying Rockmite. The first coil of Ethernet cable is attached to the pineapple can using elect

 The ionosphere at 5 AM in San Francisco wasn't set up to support long range communications, but I did get a pretty picture of the rig as Venus was rising behind it. The radio/amp pair reached across the Bay to both the Berkeley stations— W6BB and NU6XB — and nowhere else. References: The Ionosphere this morning Project TouCans RBN

 Had I set the power amplifying transistor's bias high enough, or did the radio always work this way with the amp? It's good to have data!  I've been collecting reverse beacon spots and QSO data since around the start of this year . The reverse beacon data is collected automatically, inspired by a productivity and documentation talk by Simon Willison. Last night, rain got into the Project Toucans radio. It's not waterproof enough yet, I guess. Turns out there are whole issues in using Tuna and fruit cans as radio cases that I'd never thought through. The lid is corrugated making it difficult to seal holes against water. Anyway... Earlier this evening, I had the radio/amplifier combination back up and running, but were they running well enough? Fortunately I had data for that. I queried the data using datasette . I asked about reverse beacon spots since the amplifier was added into the station mix a few days back: select rowid, id, tx_lng, tx_lat, rx_lng, rx_lat,

 OK, so this isn't so much one I learned, as one I remembered, and then empirically proved. Did you know that human thumbs are capacitative—as in they act like capacitors? Yup! They are! The project TouCans Rockmite/ TunaTopper stack suffers from the same issue the Flying Rockmite often did. Sometimes, the RFI from the antenna gets to be too much, and the memory keyer decides to stop keying out its memory. I noticed that when using the keyer with Project TouCans, if I put my thumb across the leads to the 'dah' keyer switch, the issue went away and the memory keyer ditted and dahed along to completion. I had brought a small capacitor out to the radio site to try placing across the headphone leads to solve an AM pickup problem inherent with having 25 foot long headphone wires. The same issue on a far lesser scale had been solved by Mark— N6MTS —as part of his work on the Open Headset Interconnect Standard, OHIS . The cap hadn't down the job for the Flying Rockmite or Pr

 I've been doing a lot of amplifier testing over the last week with the Tuna Topper as part of what has become Project TouCans. It's led to many more QSOs—meaning, more than one per week—from the house. I maybe haven't remembered to log them all, and so I wasn't too surprised when I saw that  N2AKJ had logged a Parks on the Air, ( POTA ), contact with me last Thursday when they activated K-2114 Nissequogue River State Park . Still, since I didn't remember the QSO, I wanted to make sure I was at least on the air. And.... There's an app for that! Using the RBN network data automatically accumulated by rm-rbn-history about QSOs and RBN spots from my callsign—KD0FNR— served through datasette , I was able to determine if I'd been on the air when the QSO was logged. Spoiler alert: I was. Understanding the SQL formatting to get the information was a bit tricky, so I'm documenting the SQL statement here so I can go back to it later. select rowid, id, tx_lng,

Project TouCans is well underway! I'll have more details soon. The key difference between TouCans and the Flying Rockmite is that it includes a Tuna Topper amplifier. The amplifier/radio combo chews through the 8 AA batteries more quickly than the Rockmite alone did. Consequently, Diaze—the 12 year old here—and I switched over to D cell batteries. Our keyer, however, is located on a AA battery case. Today, the design for the TouCans keyer became obvious when I looked at the Rockmite's current keyer next to the new battery pack: The two D cell battery packs will be glued together back-to-back. As you can see, the battery pack manufacture graciously included an indentation in the top of the battery packs that holds the on/off switch snugly—even more so after a drop of superglue is applied. The keyer video game switches will be attached in the same way they were on the AA battery pack in roughly the same location at the front of the D cell battery packs. The Ethernet breakout boar

I'm starting to be handy enough with Datasette that I can quickly get answers to my questions! This morning, the question is whether or not the Tuna Topper amplifier combined with the Rockmite is giving better propagation results than the we get with the Rockmite running alone. Since we have a database of all the QSOs and RBN spots of the Rockmite, this is exactly the kind of question Datasette can answer. Here are the commands to run datasette to get an sqlite database from the Rockmite's RBN and QSO data table . I'm listing the commands here because I've been to lazy to simply write a script, and thus far I haven't been able to get the whole assmeblelage to work on Windows, meaning I'm working on a github codespace. Codespaces are super-cool, but also—at least for me... so far—completely unable to keep command history around between launches python -m pip install csvs-to-sqlite python -m pip install pandas==1.5.0  csvs-to-sqlite -dt timestamp rm_rnb_history

While the gang and I were using the Datasette csvs-to-sqlite plugin, we learned something about modifying the code of installed Python packages: how to use pip uninstall to find the package. We knew we had issues , and we had begun to figure out how to fix them. What we couldn't figure out is where the actual code we needed to modify lived. We'd used pip to install the plugin, so finally I decided to re-install the plugin, and then watch the installation messages as they flashed by hoping to capture the pertinent directory So, I typed pip uninstall csvs-to-sqlite . pip immediately asked me if I really wanted to uninstall the package at the packages location. Problem solved! The kids and I found the code at that location, made our edits, and tried out our newly modified version of the plugin! References: csvs-to-sqlite plugin

A bit more about shielding the Tuna Topper amplifier with a can of Progresso soup. I tried a shielding experiment here at the home station today, and had great results. There was in fact no reductcion in output power. It's unclear to me if there's a reduction in general noise yet. There are roofers next door, and their propane heaters are laying down a nice, even, low-level ground floor. Here's a detail picture of the shielding: The soup can is opened on one end, the soup has of course been removed, and the can cleaned. The open end i placed over the exposed Tuna Topper circuit board. I thought there would be issues with the RF in and out wires being sandwiched between the two cans, but there are not, (at least for this installation.) I made a contact with AE7CG with the shielding in place. Here's the QSO map: You can see the signal was reaching out fairly far, but no further than Utah. This is an afternoon thing here in San Francisco that I haven't entirely figure

Just before heading to a POTA yesterday, I taped an empty can of Progresso Soup over the top of the Tuna Topper. In my typically lazy fashion, I left the radio's new nose on when I arrived at the POTA site. I could hear a few AM stations in the background, but nothing that was screaming loud. I also hear a few CW stations. Out of curiosity I removed the soup can and... So much AM screaming!!! Curious if the can was also causing signal to not make it out, (was it simply forming a capacitor between the RF in and out leads?) I checked the RBN results. The radio was five feet off the ground next to a highway, and reached two stations:  nearby W6BB and KO7SS, each with 10 dB.

The 12 year-old and I got a chance this last weekend to test out the Rockmite complete with Tuna Topper amplifier on a POTA activation. The results were outstanding! We were making a QSO per minute for most of the activation. Something we've never done before. On top of  that, the prototyped packaging of the combo of devices solved a few problems we've always had when we suspended the Rockmite from the antenna. Here's a look at the how the pair were packaged together: The first, biggest advantage, is that we moved the BNC coax connector out of the equation. As you can see, it is now the input to the Tuna Topper. That means that it is not supporting the antenna. That means  that there are no longer any torques on the BNC connector in the case. This fixes an issue that inspired so many washi tape fixes . The second advantage is that there is now access to the power leads outside of the Rockmite. This frees us up to send power in on a different path than the Ethernet cable ins

The twelve year-old, Daize, and I flew the Flying Rockmite with the Tuna Topper II+ amplifier installed for the first two times ever this weekend! The results were outstanding! And! The kid made her first QSO ever on Sunday! She and I wrapped up amplifier construction late in the week. From there, we were able to do some basic testing on the back balcony of the house where we learned that the LED was soldered in upside down, and that our multimeter was reading rather wonkily because it's battery was (already!?) going dead, and maybe also a heavy dose of RFI (radio frequency interference.) I'll be writing up way more about the initial tests, the ideas they spawned for the radio/amp packaging you'll see below, and other things soonish. Anyway, after our one brief day to test the amplifier, the kid and I were scheduled to hop on a plane to New Mexico where we would camp just below Manzano Peak and eat  breakfast each morning in Mountainair, (Yum!). We took the Rockmite and

The Reverse Beacon Network website is down today (at this moment, 23/09/18 17:29 Z). And! The reverse beacon network telnet port is still up and running! It can be reached by typing  telnet telnet.reversebeacon.net 7000 Assuming you have telnet installed. Orrrrr a spiffy script that will look for the frequencies you're interested in. (Along with Python)

The gang and I have been working on a Tuna Topper amplifier for the rockmite radio. Today, we got what seemed like positive results. The reverse beacon network reported that we had 17dB snr into K2PO/7. Seems good, but was it better than before the amp? That's where it seemed like it would be a good idea to use Datasette to review vs earlier results from the rockmite deployed at the house. That's where we started to run into issues. I'll talk more about the amplifier in another post. Here's what we learned (we is the 12, 10, and 8 year old gang and I), when we tried to convert our comma-separated-variable data to sqlite using csvs-to-sqlite . The short version pandas 2.0 is not the same as previous versions! We encountered all the problems reported in this issue for the package. We got to learn a lot, so I guess it's kinda good we didn't see the issue earlier. In short, as I wrote on the issue , you can get around the whole thing for now using: pip install pand

The 12 year-old unschooling kid here has been soldering since she was eight. She'll be taking her ham exam soon. (She's also contributing to the development of the ham practice exam app she's using. That's how we've been working on programming.) So, with all this ham radio work going on, it only made sense to ask if she'd like to help build an amp that will hopefully help the Rockmite get a little further out. So far things are working out great! She soldered the first third or so including most (all?) of the resistors last night. I got up this morning and soldered in most of the capacitors and the Zener diode. We'll be up and running in no time! Anyways: There are also ideas for suspending the Rockmite and the Tuna Topper in the antenna: There might even bee a second can of tuna for the batteries!

The gang and I were out camping for five weeks this summer. I'm finally getting a chance to take the time to write, and to look at old pictures, so you'll be hearing more about all of this, but a post on the GQRP group this morning asked about using radios without enclosures, driving me to this particular topical niche of our camping trip first.  The GQRP poster mused about using either a plastic enclosure for a radio, or no enclosure at all. The 12 year-old here, Daize, and I had discussed mounting a radio on its battery a few times at the start of this year. As it happened, we had an extra Rockmite 20 laying around, so right before we went out for our camping trip, we moved the radio out of its case and onto its battery, this was mostly done to reduce voltage drop between the battery and the radio since the Rockmite is typically mounted at the antenna feedpoint when we're out and about. Here's the rockmite about to be launched with the antenna: and here it is up with

I thought publishing the amateur radio general class license exam question pool on Datasette would get the kid and I writing (videoing in this case) help topics for the exam pretty quickly. Sure enough, a little over 36 hours later we released our first video. The kid asked what vdc was while she was studying this morning. I responded that I had no idea. (Turned out she hadn't pronounced dc like they were the sub script characters they are.) Rather than squinting at the practice app on the kid's phone, I clicked over to our Datasette instance and searched for questions with vdc and voila! A quick conversation about the dc power formula later, the kids and I were making the video above. Anyways: This isn't the kid's first appearance on ham radio videos. This This will also lead to using generative AI to create search rerences in the future.

Well, that was easy. In my previous post, I wondered about being able to publish the ham radio general class license exam question pool so that anyone could explore it using Datasette . There's a CLI command for that. Put simply, if you're already running things on Google Cloud (like a ham radio practice exam ), then you can issue a single command: datasette publish cloudrun mydatabase.db --service=my-database But, if you're running on windows... ahem... you'll also need to get rid of the tmp file cleanup routine because it'll hose up the entire process, preventing you from deploying at all. Fortunately, the fix is pretty simple once you've located the file it lives in. So, here's the general class question pool ! Explore away.

The twelve year old here is routinely passing practice exams for the United States amateur radio technician class license. Consequently, she and I started to work on updating our practice exam app to the latest general class question pool . We wrapped up the new deployment of the app earlier this afternoon, and I just wrote a python script to dump the question pool into a sqlite3 database. With the question pool all databased up, my next fun project was to study the questions in Simon Willison's datasette app . This allowed me to find out a few interesting things about the question pool right away. As an example, supposing you were to just go in to the test without studying and guess answers? Is there an answer you'd be better off leaning towards? A facet of the question pool based on the correct answer, (A, B, C, or D), indicates that no guessed answer is particularly better than any other: I suppose you'd get a tiny bias guessing either 'B' or 'C', bu

Got the following data back from my rbn_telnet.py program this afternoon. It's not what the script was created for, but it's wound up being a nice use model: b'DX de VE6JY-#:   14057.5  KD0FNR         CW     5 dB  20 WPM  CQ      2254Z\r\n' b'DX de VE7CC-#:   14057.5  KD0FNR         CW     3 dB  21 WPM  CQ      2254Z\r\n' b'DX de W6BB-#:    14058.1  KD0FNR         CW     2 dB  21 WPM  CQ      2255Z\r\n' b'DX de VE6JY-#:   14058.1  KD0FNR         CW    10 dB  21 WPM  CQ      2255Z\r\n' Note that the same station, VE6JY has the rockmite coming in first at 5 dB on 14057.5 kHz at 2245Z and then at 10 dB one minute later on 14058.1. I had suspected there might be battery issues with the radio just getting tired. That doesn't seem to be the issue since the two spots were within a minute of each other and, in any event, the later signal was stronger than the earlier one. Looking at the ionograms from Pt. Arguello, there is variance in the F2 layer&#

I really enjoy the reverse beacon network. It's been a tremendous first-blush check of whether or not the little Rockmite is working, and how far out the signal is getting. Having said that, it hasn't been useful in the way I think it might be for big contesting stations. I always imagine someone sitting in their plush chair locked away in a shack watching the RBN . A path pops up on the map to some faraway DX station. Said ham leaps into motion clacking several switches to bring the appropriate antennas, tuners, and amplifiers into play for the band the dx station happens to be on, and then belts out a 100+ Watt signal to make the QSO. But, that's not how it works with the Rockmite when you live in a densely populated urban area like San Francisco. Typically, the radio and I aren't even at home, much less in a shack. We're perched on the side of a mountain somewhere because there's less noise, and there are whole swaths of hams like SOTA and POTA that want to

I'm not sure what to make of this yet, but I'm seeing different signal strengths received at the same receiving stations based on the output frequency of the Rockmite which only works on two different freuqencies: 14057.5 kHz and 14058.1 kHz. Here's the data I collected this morning: I wondered if this was simply variations in the atmostphere, but notice that in the case of reception by ND7K, there's a difference of 10 dB in the two transmitted frequencies with only a two minute difference between the measurement times. I saw even clearer results watching the waterfall display of the Utah SDR : Notice the nice, well-defined line at 14057.5, and then a few seconds later, the fuzzy, weaker line at 14058.1: first second I don't think it's an issu with the radio since sometimes the stronger recpetion was on 14057.5 and others it was on 14058.1. I'm not sure what's causing the differences. I'm looking forward to finding out.