deadpopstar.com

Several times a day, wherever you are on earth, a number of polar orbiting NOAA weather satellites will fly overhead. These satellites orbit at an altitude of 800km, and their lower resolution imagery can be received with relatively simple equipment. It is transmitted at 137-138MHz. If you could tune your FM radio this high, and had an appropriate antenna, you would hear an unusual tick-tock sound each time the satellite passed overhead. Each tick-tock is one line of a satellite picture. Feed these sounds into a PC with a sound card and they can be decoded to obtain a picture showing data from two of the satellites image sensors, as well as calibration and telemetry data.

I had quite a bit of trouble trying to pick up these signals. I couldn’t get access to a cheap receiver that would tune to 137-138MHz. I tried building a downconverter so that the signals could be received with a standard FM radio, without any success. I finally got on track when Silicon Chip magazine published a receiver design, and Dick Smith Electronics started selling a kit.

The kit is specially designed for NOAA satellite reception, with two frequencies settings, typically 137.5 and 137.62MHz. It has an ideal bandwidth for reception of ±30kHz. The kit also includes a preamp.

The kit itself is straightforward to assemble, with a little bit of SMD soldering required. Tuning and alignment is another matter. Not having access to a frequency counter or generator, I tried tuning it by ear. I didn’t have much luck. I could pick up some various FM traffic (such as commercial radio stations and aircraft transmissions) but furiously tuning during 10 minute satellite passes just wasn’t yielding results. I eventually gave up and brought the receiver into the university where I study and borrowed a spectrum analyser and RF signal generator. I followed the tuning instructions that came with the kit and it only took 10 minutes to get the receiver working and tuned to the right frequency. I had to pull apart coil L3, reducing its inductance, to make it work. The local oscillator just couldn’t reach a high enough frequency, which would explain why I was picking up commercial FM radio stations.

I took the receiver home, hooked it up, and on the next satellite pass I picked up the signal. It was very noisy, but it did work. I had some tweaking to do to improve the situation.

The antenna I’m using incidentally is a quadrifilar helix. I found it easy enough to put together, although it is a bit hacked together with all the dodgy soldering and silicone. I had all the parts (PVC pipe and copper tubing) lying about my house. It took about an afternoon to construct.

The noise problem… The first thing I tried was installing the preamp on the antenna. That required taking down the antenna and fiddling with the preamp’s alignment during a satellite pass. That certainly increased the signal strength back at the receiver, but the signal was still noisy. I tried holding the antenna at different positions. The changes were dramatic. If I had the antenna straight up, clear of metallic objects, the signal became clear.

The lesson here is don’t let anyone install a TV antenna next to your satellite antenna. I removed the antenna from the barge board mounting on the edge of the roof, next to the TV antenna, and placed it on a pole attached to a small metal fencepost. This put the antenna well clear of any interfering devices and metal objects. This is not a permanent arrangement. A stiff breeze will likely snap the antenna pole.

Notice the TV antenna and mess of PVC piping to the left of it where the antenna used to be.

The 137.5MHz transmissions are severely disrupted by some sort of local interference. It’s a sort of clipped ‘ding-dong’ sound that occurs every few seconds. If I find out that it is a doorbell in the neighbourhood, I’ll be taking to it with a cricket bat in the middle of the night. 137.62MHz transmissions are less affected.

Here is a sample of my received imagery. I’ve cropped away the areas where I was fiddling with the tuning resulting in severe noise. The imagery is still marginal, but it’s getting better with additional tweaking.

I use WXtoImg to record and decode the images, because it runs on Linux. WXtoImg also predicts satellite passes. The software is pleasant enough, although in the free version I use some features are disabled.

In about 6 months time I plan on having this system running continuously, uploading imagery to a small web server.

Links
The OZ Weather Satellite Site - Also based somewhere around Melbourne, VIC, Australia.
Siding Spring Observatory - NOAA APT Imagery from Coonabarabran, NSW, Australia.
WXtoImg - Satellite Imagery recording, decoding and processing software. Available for Linux, Windows and MacOS X.

FAQ
How do I tune the DSE receiver?
Follow the instructions. Oh, you don’t have thousands of dolllars worth of RF diagnostic equipment. Well, things are going to be much more difficult then. You can look up when the next pass is, and try tuning while the satellite is overhead. If that doesn’t work, there could be a number of things awry. Maybe your antenna/preamp isn’t connected properly, or you’ve made a mistake in assembling the kit? Can you pick up other signals? If so, things should be working. To get a handle on what frequency range you’re in, try and indentify the other signals. Are they FM radio stations? This indicates that your frequency range is too low, try pulling apart L3 which moves the tuning frequency up.

If you’re still having trouble, you may have to seek out the aforementioned RF diagnostic equipment. Ask any electronic engineer or technician friends. Most TV repair shops should have this sort of equipment. You could pay them for their time, or perhaps offer a little liquid incentive.

It may be possible to build a crystal-based 137.5MHz frequency source which will make things easier.

What’s the deal with the antenna/preamp/receiver impedance?
It seems that one end of the preamplifier is 50 ohm, while the other is 75 ohm. I’m unsure why this design decision was made (presumably to match an existing antenna design), but it does mean that losses will be incurred if impedances are not matched. You could just suck it up and accept any losses (if you’ve little noise around, you might be okay). Alternatively, you could integrate an impedance matching transformer, or choose an antenna design with the correct impedance.