DIY: Build a 70cm Band Yagi for Amateur Satellite Tracking

It’s been a long time since I’ve chased any amateur satellites. It’s been a long time since I’ve had the time to chase them, and sadly, things have changed. In the seven years or so, satellites I once “knew” as friends were no longer operating as they used to – some had partial failures and unpredictable schedules due to battery failures, others were completely failed and never to be heard from again. Even though they were designed using commodity hardware on shoestring budgets, and most of them well outlived their design lives, those damn inanimate objects certainly worm their way into your heart if you’ve ever spent some time chasing them and heard some QSOs, or even just the reassuring telemetry beacon.

As it turned out, there was a new batch of cube-sat launches in late April (AAUSAT4, [email protected] and OUFTI-1) where some prizes were awarded for reception reports. That gave me the inspiration to get back into it. My old cardboard and clothes-hanger antenna was damaged in a house move, so it was time to build another more robust version.


As per before, I was only going to focus on 70cm band because of the proliferation of paging transmitters adjacent to the 2m band in Australia which makes reception of satellites a little difficult (due to the intermodulation and front end desensitization). Besides, 2m band yagis are quite big and long, and I know my neighbours are a bit freaked out by seeing someone swinging around a large piece of metal and wood … so I’ll just stick to focusing on 70cm downlinks only.

To make it more durable, I decided this time, I would build the antenna from a piece of wood for the backbone, and with copper pipe for the elements as it is easy to cut. A trip to my local Bunnings, and I had left with:

I already had a drill, some assorted screws and left-over BNC RG-59 75 ohm TV coax which I could use. I figured that the difference in impedance wasn’t particularly important for reception purposes.

I could have followed the simple Yagi instructions from before, but this time, I instead opted to use K7MEM’s Yagi calculator to design one.



Construction is a pretty straightforward affair. With the measurements, I first started by cutting all my pipe to length and cutting the driven element in half.


I formed an insulator for the halves of the driven element by cutting a short length of dowel and using a hobby knife to shave down a little of each side so the pipe can “slide” on. It’s not a solid fit although a bit of hammering might have helped.


I was intending to drill out holes to pass the elements through the main piece of pine, but then I realized that I didn’t have a 12.7mm drill bit, nor did I have a ream that would open it up to 12.7mm. I tried various files, but that just made for a hilarious mess.

Instead, I decided I would use the wide side of the spine, and drill holes through each of the elements (hard) and use wood-screws to screw the elements directly to the spine.


And there it is, as if by magic. The conductors were stripped and “hose clamped” to a cleaned section of the element to ensure good contact. Electric tape was placed around the wood to avoid any potential for splinters, and to provide some strain relief for the cable.

Initial Testing

Initial testing involved using my IC-R20 handheld receiver “manually tuned”, to try and chase the satellites by ear and by hand with limited computer assistance. I checked for the upcoming passes with Orbitron first and got the frequencies from various places including AMSAT, PE0SAT, EOPortal, JE9PEL Satellites List and N2YO. Then I ran outside with my receiver, put the frequency up by 12-15khz for the start of pass and swung the antenna around like a bit of a madman looking for the signal. Of course, being an IC-R20, there was onboard recording so I could examine the audio later.

Unfortunately, I had lost some of my skill in aiming, so it took me a few satellites before I could redevelop the patience and visualize where the satellite is based on the time elapsed since the pass started. I saw the same signal fading that you regularly experience with a linear polarization antenna receiving a circular polarization signal from a tumbling satellite, but the performance was marginally better than the clotheshanger and cardboard hack I used to use. I was pleased.

Sadly, I didn’t actually receive any of the newly launched satellites – probably because their output power is a bit too low. But I did receive something.

29th April 2016

HORYU-IV at approx 1630 (UTC+10)

Morse code beacon identification successfully received, but not sure if any errors in the frame data.

JG6YBW HORYU4 FAB60E0F9691BB679808

SEEDS-II at approx 0800 and 2030 (UTC+10)

Formerly heard with its digitalker as its primary mode of operation, sadly, no digitalker operations were heard, but partial morse beacon received with difficulty and errors due to tumbling.


XI-IV at approx 1840 (UTC+10)

Telemetry morse beacon at ~80mW was actually very well received for the first time. A full set of frame data was received.


The University of Tokyo has a decoder application available to decode the morse beacon.


By shoving the data in, we get the full human-readable form of the data. First time for a full decode … I’m fairly happy.

XI CW Converter  -  ISSL, University of Tokyo
Conversion [email protected]/04/29 19:06:07

UT2    7B7523
UT3    D0B8003B
UT4    4324318
UT6    7777787741

ISSL URL        (Does this show ISSL's web site correctly?)
XI-IV internal time        8090915   (0x7B7523)
Uplink count            16/31
Camera count            6/7
SEL count (communication subsystem)        0/7
Antenna deployment        Finished
CW duty ratio            More than 0.3
OBC reset                Normal (There is no reset, or unexpected reset by the radiation.)
Battery charging        Trickle Charge
OBC existence            Alive
FM Opeartion            Waiting
Battery voltage            5.0V
Battery temperature        -42.4Ž
Solar array voltage        2.7V
Solar array current
  +X panel                471.3mA
  -X panel                319.1mA
  +Y panel                229.9mA
  -Y panel                438.8mA
  +Z panel                303.2mA
  -Z panel                337.4mA
  +X panel                1.5deg
  -X panel                2.0deg
  +Y panel                0.3deg
  -Y panel                2.4deg
  +Z panel                1.5deg
  -Z panel                13.7deg
  Battery                2.3deg
  Transmitter            2.7deg
RSSI                    4.8dBu

* Results are not shown for the blank data.
* If there are some decoding mistakes, this software may offer unexpected results.

Thank you for your receiving the signal from XI-IV.
Please refer the ISSL's web site to know the details of the data format.

Intelligent Space Systems Laboratory  CubeSat Project

Doppler Tracking with an SDR receiver, Orbitron and SDRSharp

I wasn’t really happy enough though. Hand-tracking is really old fashioned, since we’re now in an age of SDR receivers. Just prior to the explosion of RTL-SDR, I spent a decent amount of money (AU$187) on a FunCube Dongle Pro.

2016043009424101 2016043009424102

It’s a nifty device based around a microcontroller abusing the USB HID class to receive frequency and gain commands, a USB sound card as the IF sampling device (96khz maximum sample rate, 80khz of real bandwidth, which is small by RTL-SDR standards but 16-bit depth resolution), with an Elonics E4000 tuner and preamps. I felt a bit disappointed after the RTL-SDR announcement, as that had 2+Mhz of bandwidth … and this couldn’t even listen to a WFM station completely. Oh well, bad timing.

But at least I could use this … just for the sake of it. I’ve tried an RTL-SDR dongle as well, and that worked pretty much just as well as long as there wasn’t too many out of band signals screwing with you.

Anyhow, if you’ve already got SDRSharp and Orbitron set-up, then getting automated doppler tracking means installing a plugin, and then enabling the MyDDE driver inside Orbitron and “magically” things will work provided you’ve plugged the right centre frequency into Orbitron.

This works well if you have a laptop battery powered and a long USB cable as I did, as it lets you see the signals even if they’re below the SNR required to actually hear them. This makes aiming the antenna a snap as well, as you’ve removed one variable (namely frequency tuning) out of the equation, so you don’t have to focus on twiddling knobs.

Unfortunately, the frequency updates are about once per second resulting in a “stepping” pitch on the output which can be a little confusing when decoding CW in the “frequency” domain. But looking in the time domain results in a clear signal.

I managed to receive PRISM’s CW beacon (partially) on 29/04/2016 at 2224 (UTC):


I also managed another fragment of SEEDS II telemetry paying attention to spaces this time, but the signal was fading too much:

001 002 004 000 000 002 945 94R 9B7 1F

Rainy? Just Leave It Out

Unfortunately, around this time, Sydney had a long rain spell which meant that going out and actively chasing satellites wasn’t an option.


So I decided to grab my bike servicing stand (a cheap Chinese one) and clamp the antenna facing north at about a 45 degree angle. This, combined with the beam-width of the antenna, meant that it could see a fixed patch of the sky. I set up Orbitron and SDRSharp along with an RTL-SDR to capture the whole amaetur satellite band looking for spectrogram bursts which would indicate satellites were detected. I didn’t have time to analyze them all, but some packet data satellites were indeed “seen”, but reception requires much better signals and specialized equipment. It’s comforting to know they’re working though, and the doppler shift shows and proves to you that you’re receiving a satellite.


GOMX3 (although GOMX1 also shows the same sort of wide-data-bursts)


HORYU-IV (left) and PRISM (right)


KKS-1 and CHUBUSAT3 (most likely, left) and CHUBUSAT2 (right)


LILACSAT-2 (left) and XI-V (right)


SPROUT (left, SSTV?) and ITUPSAT-1 (right)


UNISAT-6 (left), XI-IV (right)


Firebird-4 (wide data?)


Want to chase an amateur satellite? Well as it turns out, it’s not so much work to build a serviceable antenna. Sure, it doesn’t have the best gain or cross-polarized elements, but it’s simple to build and doesn’t cost the world. I’m sure I wouldn’t appreciate having cross-polarized elements because the antenna is heavy enough as it is that it’s a bit of a workout tracking the pass from horizon to horizon.

With a decently quiet laptop, RTL-SDR or equivalent dongle, and a long USB extension lead, having doppler compensated tracking is possible and inexpensive. If only that was available to me back in 2009!

While I didn’t end up receiving the satellites that I wanted to get, at least I had a chance to hear some new satellites I’ve never heard before.

I’m sure there’s going to be quite a few more cubesat launches to come, so maybe you can try this yourself as well :).

About lui_gough

I'm a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!
This entry was posted in DIY, Electronics, Radio, Satellite and tagged , , , . Bookmark the permalink.

Error: Comment is Missing!