Saturday, February 15, 2020

Lemonade!

We had a great harvest of Eureka lemons, plus a few Meyer lemons that we had grafted on to the tree.


Lemonade is great, right? Here's the recipe I used. It's not too sweet, and it doesn't require too much measuring.

Ingredients
1.5 cups lemon juice
1.5 cups sugar

Start with a two quart pitcher.
Add the sugar first.  # This way you can reuse the measuring cup.
Add the lemon juice.
Fill the pitcher with cold water.
Refrigerate.

Aren't recipes a little like a programming language? Ingredients are declared, then steps are executed. You can even have comments or loops like "check every five minutes until done".

I just hope I don't find a bug in my recipe!


Saturday, January 18, 2020

First Pass at a Backyard Loaded Dipole

40 meters is a great amateur band because it's open all night. That is, it's not like the 20 meter band that stops skipping radio signals around the globe as soon as the sun goes down. Unfortunately a 40 meter dipole is 20 meters long end-to end and my backyard can only fit about 15 meters if it's run diagonally. The trick is to add load coils about 1/3 of the way from the end to lower the resonant frequency of the shortened elements.


It looks something like this in the XNEC2 simulator:



The yellow boxes represent the loads. The red box is the location of the balun that connects the antenna to the feed line. In simulation, the trick is to alternately shorten the elements then increase the inductance of the loads to prevent the resonant frequency of the antenna from rising. Resonance is identified by the frequency at which VSWR is at a minimum.

Here's the NEC2 card stack I finally arrived at:


CM --- NEC2 Input File created or edited by xnec2c 3.6.1-beta ---
CM 40 meter loaded dipole
CM (null)
CE --- End Comments ---
GW     1    15  -6.00000E+00  0.00000E+00  7.62000E+00  6.00000E+00  0.00000E+00  7.62000E+00  1.50000E-03
GE     0     0   0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00
EX     0     1     8      0  1.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00
FR     0   201     0      0  6.00000E+00  1.00000E-02  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00
NH     0     0     0      0  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00
NE     0    10     1     10 -1.35000E+00  0.00000E+00 -1.35000E+00  3.00000E-01  0.00000E+00  3.00000E-01
LD     0     1     3      3  1.00000E+00  2.45000E-05  1.00000E-06  0.00000E+00  0.00000E+00  0.00000E+00
LD     0     1    13     13  1.00000E+00  2.45000E-05  1.00000E-06  0.00000E+00  0.00000E+00  0.00000E+00
RP     0    19    37   1000  0.00000E+00  0.00000E+00  1.00000E+01  1.00000E+01  0.00000E+00  0.00000E+00
EN     0     0     0      0  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00  0.00000E+00

After fiddling with the GW and LD cards I came up with a loaded dipole optimized for 14.070 MHz (the PSK frequency).

Antenna Length: 12 Meters
Number of Segments: 15
Load Segment Locations: 3, 13
Load Inductance: 24.5 uH

Next step is to see if a 24.5 uH inductor, 12/15 meters long is even practical. The next iteration will probably be to move the inductor closer to the feed point, and increase the number of segments.

Next I'll compare my design to published designs as a sanity check on the simulation. If that check out, I'll build it and measure the performance with an antenna analyzer.

Anyway, here's what the VSWR and radiation pattern look like.




Tuesday, January 7, 2020

Processing Tracks from the Columbus V1000 GPS

The Columbus V1000 is a really cool device for logging coordinates with which to geotag photos. Because it can store data in the GPX format, its files can be imported right into geotagging apps like Prune. The V1000 can also log temperature but unfortunately there's no place to in the GPX format for that. Instead temperature can be stored in a CSV file or Columbus' propriatory GPS format. However those formats can't be easily imported into the geotagging apps that I prefer.

Recently I found a solution. Adam Snodgrass wrote a slick little command line app that converts the Columbus GPS format to GPX for geotagging, and CSV for temperatures.

https://github.com/asnodgrass/columbus-v1000

To install it first install Git:

$ sudo apt-get install git-core

Then install the conversion program:

$ go get github.com/asnodgrass/columbus-v1000
$ go install github.com/asnodgrass/columbus-v1000

You can run the program by with the following command:

$ ~/go/bin/columbus-v1000 gpx -i abc.gps -o abc.gpx
$ ~/go/bin/columbus-v1000 csv -i abc.gps -o abc.csv


Sunday, October 13, 2019

Setting Linux USB for ic-7300 and fldigi

It took me a while to figure out how to set this up, so here are a few notes.

The first issue was that the user needs dialout privileges.

$ sudo usermod -a -G dialout $USER

The user may also need to get audio privileges. Check settings like this:

$ id
uid=1000(tester) gid=1000(tester) groups=1000(username),4(adm),20(dialout),24(cdrom),27(sudo),29(audio),30(dip),46(plugdev),112(lpadmin),129(sambashare)


The next issue was that the old laptop I was sending a corrupted signal. Looking at the audio FFT on the 7300, I could see that the image had wide "skirts". Maybe old laptop wasn't up to the job. I found out later that fldigi has a checkbox for "slow cpu" under "misc", "cpu". That might have fixed it, but I just switched to a desktop computer and now the audio FFT now looks something this:



Sunday, July 8, 2018

Grafting Fruit Trees

I've always thought grafting was an amazing bio-hack. It seemed like an incredibly difficult task to get a branch from one tree to grow on another. It turns out it's not that hard, especially because there are a lot of resources available. If you search the web for examples, though, you'll be overwhelmed by all the different techniques. But first, where do you get cuttings to graft onto you trees. What we did was to go to a scion exchange. This is where you pay $5 for admission and can take any scions, or fruit tree cuttings, that you want to try grafting. There are also demonstrations of techniques that work for the types of trees in your area that you want to graft. After you've selected a technique, you can then search for more information and videos.

We tried both bark grafting and cleft grafting. I think the cleft graft is much easier, but it requires cutting toward yourself, so you have to be extremely careful. I used a very sharp grafting knife and put a spool around the root stock to protect my hand, but even that was very scary.

We did grafts on apples, pit fruits, pears, and citrus in two grafting sessions. A little more than half of the grafts were successful. Not bad for a first try. Apples and pit fruits were easier. Citrus trees were more difficult. Some of the fastest growing grafts were onto a volunteer nectarine tree - it must have grown from a pit I tossed into the back yard. The fruits on these tree have never been good, so it was a good candidate for grafting.

Here's a photo of the nectarine a couple of months after the graft:


And here's a photo of a Meyer lemon on Eureka lemon root stock.


Unfortunately, in many cases I can't tell you what the scion is until after we have a harvest because I tagged it with Sharpie on plastic and the weather faded it to unreadability! Scratching the name of the scion type onto an aluminum tag works much better.


Tuesday, April 17, 2018

Groundplane Antenna

I got my ham radio license and a Baofeng handy-talky last year. I've found that most of the conversations go on over 2 meter repeaters. Unfortunately the nearest repeaters to me are 10 and 15 miles away. With the 5 Watt Baofeng, to reach the repeater at 10 miles I have to stand out in front of my house and I can't reach 15 mile repeater at all.

The solution is a better antenna. As an easy first project, I selected a ground plane antenna from the my public library's copy of the The ARRL Handbook for Radio Amateurs.

Here are the sources for the parts and additional tools I needed.

From Amazon

BCP 5pcs UHF Female Jack Solder SO-239 4-Hole Chassis Mount Coax RF Connector by BCP
$6.89

MPD Digital smaF/pl259f SMA Female to UHF SO239 PL259 Female RG58 Adapter Cable, 3-feet
$19.59

DHT Electronics Handheld Antenna Cable, Adapter, SMA Female to UHF SO-239 Female Connectors, 3-Feet 
$7.00

From McMaster-Carr

1 89015K176 6061 Aluminum Sheet, 0.05" Thick, 8" x 8",  
8.19 each

1 8859K164 Ultra-Formable 260 Brass, 3 Feet Long Rod, 1/8" Diameter, packs of 5 
17.69 per pack

2 6750K11 Hard Anodized 6061 Aluminum Rod, 1/8" Diameter, 3 Feet Long
4.63 each


From the local big-box hardware store.

10-24 screws, nuts, and washers.

4-40 screws and nuts - 2-56 would have been better but they didn't have any.

3/4" PVC Pipe.

Three 3/4" PVC tees.

Four 3/4" PVC caps.

From Harbor Freight

Step Drill Bit

The first order of business was to mount the SO-239 connector on to the aluminum sheet. The logical way to do this would be to use a 5/8" Greenlee chassis punch, but when I saw that it costs almost $60, I looked for an alternative. What I found was called a step drill bit that lets you drill various large holes and costs less that $10. After drilling the mounting holes, I bent the corners of the aluminum sheet to 45 degrees.

Next step was to figure out how to mount the round aluminum rods onto the flat aluminum plate. Pounding the ends flat with a hammer was crude but effective. As I hammered the ends would curl to the left or the right depending on which side was stretching more. If it curled to the left, I would have to pound more on the left side causing that side to stretch more and correct the curl. When the flattened end was wide enough for a 4-40 clearance hole I filed off any remaining black anodizing and drilled the holes. If I do this again, I'll procure 2-56 screws, because drilling such big holes in the end of the rod leaves it weaker than I would have liked. Finally, I used a cut-off wheel on my rotary tool to cut the rod length to spec. This process was repeated four times.




Now for the brass vertical element. It's got to go into the solder cup on the SO-239 jack, however, the end of the brass rod needs to be ground down to fit into the cup. For this I used the rotary tool again, this time with a cylindrical grinding burr. It was a tedious process, but when the end of the rod was ground down sufficiently, I was able to solder it into the cup with a liberal amount of flux. After the rod is connected to the jack, the jack can be connected to the plate with 4-40 screws.

To support the antenna, I made a fixture from PVC pipe. In the tee in the middle, I used the step drill bit to make a hole so I could run the cable through the pipe. I ended up not using any solvent glue because the fit was pretty tight, and the 10-32 screws I used to bolt the fixture to the aluminum sheet hold it together pretty well.

Here's a photo of the finished project mounted on a bar stool (not its permanent home).




Now I'm able to reach the repeaters, but to really tell how much better it works I'll need a field strength meter. Maybe that'll be my next project.

Saturday, August 8, 2015

Calculating Three-Quarter Lighting for Railroad Photography

Lighting is important for any photographic subject. Three-quarter lighting is a term from portrait photography meaning the light shines on both the subject's front and side, yielding a flattering three dimensional appearance. In railroad photography, a three quarter shot is disparagingly called a "wedgie". Rather than artistic, a wedgie is intended to be documentary, highlighting the details of the equipment. Unlike portrait photography, a railroad photographer can't move the subject, but has to plan on the sun being in the right place.

Here's what a back-lighted image looks like:


 And here's a three quarter lighting example:


Given a time and date, which locations are good for photographing a train? Not only does the sun move with respect to the rails throughout the day, its path is different throughout the year. For example, in the winter in the Northern Hemisphere, the sun will never shine on the front of a train headed directly north. Another variable is the direction the train is travelling. I made these pages using JavaScript, Google Maps, and Google Charts to identify those locations. So far there are three pages:

Roseville, California
Carpenter Road near Colfax, California
Hakata Minami Station (博多南駅), Japan
 
Be aware that for these pages to calculate correctly, you may have to adjust your computer's time, date, and time zone so that it matches the selected location. Also, be sure to read the Map Information block for help on how to use the page. You may wonder why a solar elevation of less than thirty degrees is highlighted green. This is because when the sun is low, the details of the undercarriages of locomotives are well illuminated, which is considered desirable in a roster photograph.

To determine where the sun is for any time or location, I used equations, with permission, from a document called  Computing planetary positions - a tutorial with worked examples by Paul Schlyter of Stockholm, Sweden.