L-Band SETI Dish Build
What did I just buy?

October 19, 2016

September 6, 2016 was only the beginning of my quest to build a private back-yard Search for Extra-Terrestrial Intelligence (SETI) system. I had just received an email about a ten-foot satellite dish on a nearby farm, and for a small sum it could be mine. Fifty dollars seemed a bit steep for an old dish that most people would give away for free. That being said, this was the first and only person who responded to my classified ad. Since I was too timid to knock on doors asking people for their lawn ornaments, I just had to go look at the dish!

When I pulled up the long driveway, I was met by an older gentleman. “It’s all there!” he exclaimed. That should have been my first clue! I did notice that the feed horn support struts were twisted like pretzels, but they could be replaced. By now it was getting dark, but everything did seem to be there, and it was the black mesh type that I had hoped for. I paid the man his fifty dollars and the dish was mine. I couldn’t wait to get it home!

My daughter loves it when I run the amber lights on the truck. To be honest, I think it is pretty cool too. I don’t often have a reason to run them, but they do come in pretty handy sometimes. The trip home was slow and uneventful. We only had one vehicle pass us while we were stopped at the side of the road, and shortly thereafter, we arrived home with my new toy.

Disassembly Begins

October 19, 2016

Shortly after getting the dish home, I began disassembling it. There were 320 screws, 20 bolts, 40 washers and 20 rusty nuts holding it together. The majority of the bolts broke as I tried to remove them, but they all came out. Replacing the hardware ended up costing me more than what I paid for the whole dish! However, it needed to be done for the dish to last any mount of time. There was simply too much corrosion.

Frame Repair and Paint

October 19, 2016

Purchasing a black satellite dish without seeing it in the daylight probably wasn’t the best decision I ever made. It appeared that the folks who unmounted the dish from it’s pole did not do so gently. I am optimistic though, and after some straightening and paint, it started to look like it may be usable. Twenty years or so had taken its toll on the frame, as there was some swelling and cracking from ice. This was mostly on the outside edge, so I am hopeful that it won’t affect the parabolic curve of the dish.

Dish Mount Modification

October 20, 2016

I spent many hours trying to figure out how to steer the dish on a budget. I wanted more than a fixed “bird bath” style of installation, but without the cost of an azimuth/elevation rotator. I finally settled on a “meridian transit” installation. The idea is to modify the dish bracket to allow adjustments in elevation, and let the azimuth position be taken care of by the earth’s rotation! This was a reasonable compromise, but a fully steerable dish would have been my preference. Creating a fully steerable Yagi array may end up being a future project.

Light Years from Completion

October 20, 2016

This is a big project, and the time-lines are tight due to winter coming. There is so much that I want to accomplish with my dish installation!

Completed Dish Mount

October 20, 2016

After spending some time at the welder, the modified dish mount was painted. The actuator arm was removed for testing and repair or replacement.

Mounting the Mesh

October 20, 2016

I cleaned the corrosion from the aluminum mesh, and the first four petals were painted on an old tarp. The paint I used was flat black, as using a glossy surface could reflect heat radiation and melt feed horn components. I started painting with a foam brush, which I quickly determined was not the correct tool for the job. Using a roller worked much better, but care needed to be taken so that the paint did not clog the holes in the mesh. I was able to tolerate a few imperfections, as it was inevitable that some would get clogged during the painting process. My worst fears came true the next day though, when I lifted the stuck petals from the tarp. Bits of green tarp were stuck all over the back! It took some effort, but I was able to wire brush the worst of the mess off of the mesh. The next set of petals would be painted while on the frame. This worked much better than I thought, and there was very little clogging. The main draw-back is that rolling the back of the petals will be more difficult. However, since the backs had maintained most of their black colour, this isn’t a problem for me. Slowly, the petals were screwed back in place with new screws, and it started looking like a dish again!

The BIG Dig

October 29, 2016

October 29, 2016 – Canada – Finally I had a relatively nice day to dig the post hole for my dish. I dug it by hand, just over four feet deep and wide enough for a twelve-inch forming tube. I used 5 bags of quick-set post concrete mix and seventeen and a half litres of water.

The post level was the best $10 I ever spent, and made leveling the post a snap.

I borrowed the idea of running a rod through the post to prevent it from turning. I saw this on the SETI League’s article on the Very Small Array (VSA) setup where they performed a similar procedure using rebar. I used a small length of stainless steel rod for my application.

I considered using rope secured by hose clamps to keep the post level, but using a two-by-four frame to secure the pole worked quite well.

Great Progress!

November 6, 2016

The only thing better than building a SETI station on a perfect autumn day, is doing so with friends. Rasika was on-hand to share in the triumphs as major station milestones were acheived. The dish now stands proudly on it’s mast, and thirty feet of two-inch electrical conduit provides a channel for cosmic signals to reach the receiver in the lab. What a great day!

The Angle

November 11, 2016

Luckily, my little dish project has been progressing without any major issues. Of course there has been a few minor glitches, and another revealed itself when I was pondering ideas for dish positioning. I thought through the bracket modification while it was on the ground, arranged for the welding to be done, and then put a big check mark beside the task. However, after getting the dish mounted, it quickly became clear that the actuator arm didn’t have the motion of freedom required to move the dish more than fourty-five degrees. Ninety degrees of motion is the least amount that I am willing to accept. Today, I spent my lunch break over at the welding shop discussing options for a corrective bracket. If they can put corrective lenses on a space telescope, surely I can put a corrective bracket on my dish! Next week I should have a new bracket in place, and with any luck I will have the dish moving electronically as well.

A couple bits of good news presented themselves this week as well. The first is that my cylindrical feedhorn with choke, and low noise amplifier (LNA) are expected to ship on Monday! Yippee! The second, is that I have sourced a second C-Band dish! I am told that this new-to-me dish is mine if I haul it away. I won’t be able to put up another in my back yard AND remain married, so I will need to dismantle it and store it away for SETI II!

Stay tuned, more to come!

Activating the Actuator

November 23, 2016

Another small step towards my goal of having the SETI system operational by December 1, 2016. The bracket was painted and mounted, and I temporarily put an old actuator in place to try it out. It seems that a 24 inch actuator will give the correct amount of arc, which is about 90 degrees. I often wish I had a full azimuth/elevation control, but this is what I have to work with. One only needs to glance up at the heavens to see how many billions of stars there are in every direction, so it won’t be a show-stopper. It may however, hamper my ability to confirm candidate signals detected by other stations. As you can see, the L-bracket works nicely, allowing the actuator to lift the dish. When extended, a 24 inch actuator will lift the dish beyond zenith allowing the maximum north-south view at the site. The actuator seen here was the one that came with the dish. I can pull the arm in and out manually, so it seems that the worm drive is messed up. The heavy-duty actuator I have is missing the limit switches, so I am going to purchase a new 24-inch heavy-duty arm so that I don’t run into problems with snow load.

More Parts

December 3, 2016

I thought that you might be wondering how things went with my second dish. Although the dish was mesh, unfortunately it was made of steel. Twenty-plus years of service had taken its toll on the steel frame, and much of the supporting structure was rusted beyond repair.

I salvaged the steel post, feed horn, LNA and downconverter. This dish also had a cool position system. It used a motor with a pulley and cable that moved the dish. I kept the motor, but I am not sure if I can figure out what voltage it uses. All-in-all, it was a good day spent with my son and friend Rasika, who both helped me take it down. The reflector itself ended up in the scrap heap.

‘Tis the Season for Miracles!

December 18, 2016

Finally, the feed horn is on the dish! It took some effort, but with some help from my son, we managed to get it installed. I calculated the focal point with a nifty calculator from SETI League as 3' 9 1/2? for a 10' dish 19? deep.

One disappointing discovery was that the new conduit we laid was not waterproof. I was worried about that, as the conduit was moved a bit after gluing to position it in the trench. We also have a high water table in the area. Today when I tried to pull through the cable, I discovered it was blocked, presumably by a block of ice. (-7C and -14 with wind chill today). I will need to revisit the conduit situation at a later date. For now, I don’t have much choice other than that o run the cables over the ground.

Note the amplifier has Power Pole connectors for quick disconnection while working on the feed. I promptly removed the LNA prior to mounting the horn so it didn’t end up in the snow!

I have also started work on the SETI Elevation Control Module which will use a PIC16F690 microprocessor and will talk back to SETI Net (www.seti.net) using RS-485. I am using a prototype/demo board to experiment. Although I have had the PIC environment for some time, I haven’t played with it much. I look forward to learning how to make the most out of it!

Gimme a hand, err, I mean an Arm!

December 21, 2016

The actuaor arm is now in place on the SETI dish. It was windy, dark and cold. The dish was heavy. However, I was able to push through and get the job done!

Here comes Trouble!

dishwactuator.jpg

December 22, 2016

Well, it wouldn’t be a challenge if everything went right. I have ran into a problem with the hinged part of the dish mount not quite lining up properly. The dish needs to move exactly in a vertical plane in order to accurately tell where it is pointing in the sky. Being just a fraction of an inch out on earth can translate into thousands of miles (or more) in space. I think I can cut the notch in the steel a bit longer, and get things fixed up. I will try that tomorrow. Otherwise the dish is taking shape!

Here is the feed horn and the original C-Band headend cone that I was able to use to protect the 1420 MHz gear.

Smooth and Steady

December 23, 2016

Yesterday was a good day at the Trillium Radio Observatory! The dish alignment issue that plagued the dish was corrected, to my surprise, with-in a couple of hours. The issue which prevented the dish elevation from moving in perfectly vertical position was corrected by extending the adjustment slot by an inch and a half with a cut saw.

Once the slot was extended, I needed to run to the local welder here in Dutton, to extend the adjustment rod. He was shutting down for the holiday, but I caught him in the shop and he was kind enough to help me out! The new piece went in nicely, although I had to put a few washers in as spacers. Unfortunately, my ratchet slipped while tightening it, and I whacked myself in the forehead with it!

Once the new piece was installed, I gave the piece a squirt of black paint, and it was done. I looked up into the night sky, and realized that the sky was very clear. It was a mild evening, close to the winter solstice and a good night to try and align the dish with Polaris.

There was quite a bit of artificial light to the North, so I had a hard time finding the little dipper, or any dipper for that matter. With the help of my SkyView program from the Apple App Store, I found the general location and spotted Polaris. (Right where it was last time!) While searching the sky, I saw two shooting stars!

My mentor and newly found friend explained to me the best way to align the dish to Polaris is to use a green laser. The only lasers I have are red, which are pretty much useless for this purpose. I hoped to point the red laser at the house to get the azimuth correct, but Polaris is to the right of the house and well above the trees. I am a bit leery about pointing lasers in the sky, as I am located on some heavily travelled air routes. The aircraft are quite high, but any attempt to use a laser must be done with caution.

What I did to align it was to start with the lights out. I pointed my flashlight upwards to illuminate the feed. The top strut reflected the light, so from behind the dish, I was able to vertically line up the strut with Polaris above. Although a laser would be a much more accurate solution, I am quite happy that it is pointed true north so I can begin monitoring.

I purchased LMR400 coax, but I ordered 40 feet, which wasn’t enough. I hope it will to through the conduit at least. I was able to dig out a run of RG8U which is at least 50 ohms. I used it with some adapters for now. I ran outdoor extension cords for the dish actuator and LNB power. This also provided a third line for a future bandpass filter switch. Telephone wire, although indoor type, was used for the actuator reed sensor. I need to find some better wire for this purpose but it will work for now. All lines run into the control room.

I used power pole connectors for the LNA and banana connectors for the dish actuator. I almost used power poles for both, but I wanted to prevent sending 21 volts into the LNA! I searched for a power supply. The first gave me 30 volts; too much. Then I took a 3 amp 12 volt regulated supply apart and found it had 21 volts before the regulator. To spare the regulator circuit, I installed a fuse holder and two new binding posts to the rear of the supply. I tested the actuator, and it worked first try! No crunching, squeeking, thumping or other unwanted sounds. I took it right up to zenith and back down. Awesome!

So the plan is to set everything up and have first light today at 1420h local time. I thought 1420h would be a cool time to light it up.

The Search Begins!

December 24, 2016

My L-Band SETI station is now operational. I will be working on automating the actuator so that the SETI Net software will steer it automatically. The system is posting results to SETI Net (www.seti.net) for those wishing to take a look!

The system was turned on at 1420h local, but some tweaking was needed to get it receiving properly. Updates to the SETI Net software were also made to bring the system components to the latest versions.

PIC or Pi?

December 28, 2016

I originally planned on using a PIC microprocessor for the actuator arm, but my good friend James Brown highly recommended the popular Arduino for controlling the dish elevation instead. The Arduino has a large number of modules available to interface with, so I can understand why James would recommend it.

Before purchasing an Arduino board, I decided to look at the Raspberry Pi as an option for a number of reasons. First of all, I have a Pi already. I am also quite familiar with Linux, so the learning curve isn’t as steep as the PIC or Arduino. Also, the Pi has an Ethernet port built-in (I use a Raspberry Pi II), unlike the PIC board I use, and the Pi pretty much has all of the features of a Linux PC. The GPIO ports are pretty easy to interface with using Python, which I have done successfully using a relay board.

Originally, I planned to use three GPIO output pins on the Pi, and three relays out of my eight-relay board. I was going to use two of them to act like a DPDT relay to switch polarity of the actuator power. These would act as the up and down controls. The third relay was to be used for switching power to the actuator on or off. I was a bit concerned about the delay between turning on the two pins for polarity. If the delay on one relay was too long it could short out, so I opted to drive both polarity relays from the same GPIO pin which seems to work fine. The on/off relay would use a second GPIO pin for a total of three relays and two GPIO pins.

Another nice feature about the Raspberry Pi is that I can run Digitemp software by Brian Lane to read my Dallas Semiconductor 1-Wire temperature sensor. I understand that this can be done with a GPIO pin, or read directly with a PIC, and probably an Arduino as well. However, I got it working with a USB to 9-pin serial adapter and a parasitic interface.

I plan on daisy-chaining several sensors to obtain temperature measurements indoors, outdoors and inside the cone covering the headend components.

Building the SECU

January 7, 2017

The SETI Elevation Control Unit (SECU) is the part of this project that I look forward to most! Being able to steer the dish remotely from my control room, or half way around the globe fascinates me. Having the dish move automatically is even better!

The nineteen-inch rack case that I will be using was once a telephone patch. Then it was a battery backup, which I finished the electronics for, but for which the battery never got installed. I have had this box in my hoard for about 20 years! Wait to see what it looks like when I am done with it!

SECU Work Continues

March 3, 2017

I have been programming the interface between SETI Net and the SETI Elevation Control Unit (SECU). The DLL is written in C++ and I lovingly call it Back End Control Application (BECA). This software has been consuming my down-time for some weeks now.

Last night I finally got some work done on the SECU. My goal is to have the dish moving via my test application by Monday. I have taken tomorrow off of work with hopes of spending a bit more time towards reaching my goal.

Pictured above is the relay board, Raspberry Pi 3 and DC to DC convertor. The red and black jacks for the actuator can be seen on the back, and terminal strips will connect the magnetic reed sensor to determine the dish elevation. I have a new gyroscope that will eventually be used for position measurements, but for now counting arm rotations should work fine.

More exciting posts to come as the project inches closer to completion!

Reed Sensor Repairs

March 11, 2017

Looks like I solved the sensor issue today. The reed switch in my actuator arm was intermittent, causing all kinds of weird readings. I dug out an old alarm magnet switch and carefully disected the glass reed switch from it’s plastic tomb. Once it was delicately soldered in place, I mounted it on the actuator arm. I ran the software, and BINGO! It worked!

ECU Packaging

Update 2018

My Elevation Control Unit (ECU) took on a new look with its larger rack case, fresh red paint, Arduino-controlled push buttons and LEDs. The original case wasn't large enough to hold the actuator power supply, and I wanted all dish elevation components to be in a single integrated unit. The fuse was originally supposed to be a breaker, but I "broke" it when mounting. As long as I can keep water out of the actuator motor, a fuse works just fine. The ECU Arduino controller reads inputs through the 9-pin Multi-Purpose Interface (MPI) port from the main Raspberry Pi controller. This allows software to automatically move the dish as required. It also reads the front panel push buttons for manual control.

The red LED goes solid when the dish is moving, and flashes "ERROR" in morse code (8 dits) if more than one input is turned on at the same time. The Arduino will also stop the dish if more than one input is detected. The green LED is the power LED and flashes several times when powered up. The home button latches the dish in the down position until the dish is at its lowest elevation, which allows for head end maintenance. The dish does not respond to inputs until the home button is pressed again to unlatch it. This is done to prevent the Raspberry Pi from moving the dish while it is being worked on.

Gyroscope Installation

Update 2018

Originally I mounted a water-proof see-through case on the side of my conduit junction box. I mounted the gyroscope Arduino inside, which looked pretty cool! However, it seems that the gyroscope did not want to operate with such a long wire between the sensor and the Arduino. I was forced to move it. The large red light is a 12 volt trailer marker light, which I use as an LNA voltage indicator. I can see the light from the kitchen window so I can easily tell that the voltage is present. Below is an image of the gyroscope sensor, which I had removed for troubleshooting.

In the next photo, you can see that I mounted the Arduino to a firm piece of plastic which fit snugly into the water-proof case. This allowed me to remove it for maintenance, and I did not need to drill holes which could potentially let water in. To the left of the picture you can see my RS-485 converter which is connected to a computer in the receiver room. The computer queries the Arduino using the RS-485 connection every 250ms to get the position. The position is then updated in a database table so it can be read by the software. I plan on having other devices such as a computer Bulletin Board System (BBS) query the position.

Here is it's final resting place on the back of the dish. It still looks pretty cool at night! The red LED is the RS-485 tranceiver, and the green LED is power. The on-board amber LED shows activity, so it blinks away when it gets a position request from the computer.

Calibration

Update 2018

This is a photo of the directional coupler and noise source that will be used for calibrating the graphs on my radio telescope. I plan to connect around 100db of attenuation between the noise source and the coupler. The noise source runs on 12vdc and will be switched remotely using the Raspberry Pi GPIO port. The telescope will use relative measurements, comparing the incoming signal to the signal level of the noise source.

Graphing

Update 2018

I have been using Radio SkyPipe for the output of my radio telescope, as seen in the above image. The software allows the images to be sent via FTP to the web interface. I am also experimenting with rtl-power-fftw by Klemen Blokar and Andrej Lajovic. I have been using the Hackster.io article by Mario Cannistrà as a guide which uses Skyfield and matplotlib to plot the Fast Fourier Transform (FFT) values to a heatmap graph. This will be particularly useful for providing visual representation of signals for my SETI work. Thanks to those who have made this possible!

Once the FFT values have been captured, and the heatmap images are generated, I have an script that sends them up to the web site. This provides a visual representation of signals for review. I plan to eventually process the data to identify non-terrestrial radio signals, record the data and notify me that a candidate signal has been detected. It would appear that I currently have a couple of signals that are interfering at 1420 MHz. I am working on isolating and eliminating this interference.

Control Computer

Update 2018

To operate my hydrogen line radio telescope, I require several components. The head end cone contains the LNA and Band-Pass Filter. It will soon contain the directional coupler, attenuators and noise source as well. The RG8 coaxial cable sends the signal back to the lightening arrestor and splitter where the signal goes to two separate RTL-SDR tuners. One tuner is on a Microsoft Windows system that feeds Radio SkyPipe via RTL-Bridge software. The second tuner is on a Raspberry Pi running rtl-power-fftw. The Raspberry Pi generates the heatmap graphs and runs Python scripts that control relay boards (yeah relays) which send control signals to the Elevation Control Unit (ECU). One-Wire temperature sensors are read by Digitemp software on the Raspberry Pi and values are updated in the databsase. On the Microsoft Windows system I run a Java program that queries the Arduino on the back of the dish reflector and updates the database with the current value. I also start a terminal to run the task management software, and run Radio Skypipe and its FTP management software

The web graphical user interface for the dish allows observation tasks to be created. It also shows the current dish elevation, counter value, task status, log file and temperatures (outdoor, LNA, receiver room).

TODO List

Update 2018

  • Rain Sensor needs to be installed so that remote operators can tell if rain may be producting artifacts on the graphs.

  • Weather Radio needs to be installed so that a weather alert will automatically position the dish at zenith to protect it from high winds.

  • Snow Removal needs to be done during winter months by tilting the dish vertically once an hour or so.

  • Experimenting with plotting graphs using Linux should be further pursued in hopes of reducing the hardware resources required.

  • Additional temperature sensors need to be added.

  • Cooling for LNA and hardware needs to be addressed.

  • Raspberry Pi and relay boards need to be rack mounted.

  • DC Power lines should have surge suppression added.

  • New regulated power supplies should be built for the L-Band LNA, Raspberry Pi, Arduino and the Ku-Band LNBF.

  • Add Ku-Band LNBF to 3-metre dish. (potentially)

  • Add magnetometer (probably not necessary).

  • Integrate various Java software components into a single server application.

  • Automate analysis of FFTW data to provide candidate signal notification.

  • Automate SETI League Project Argus Reporting

  • Dish audio, graph and FFT data logging and archive.

  • Add dish elevation and frequency information to heatmap metadata.

  • Improve uninterruptable power system.

  • Migrate to Linux Raspberry Pi platform.

  • Set up deployment system for quick Raspberry Pi recovery.