The Ham Radio Series 44 - More Intro to Moonbounce

A week ago, I did a post on communicating by bouncing ham radio signals off the moon (and gave the post a title so that after a few months I'll never find it again). Looking back at that, I found a few things that I didn't include that are very important, and another few things that make a good addition to it, so more on the subject tonight. 

The thing I forgot to mention is vitally important to getting started and describes my situation perfectly.  That article talked about sending a signal to the moon and hearing your own echoes, but it's possible to complete contacts via moonbounce without being able to hear your own echoes.  The important thing is there are two stations involved and it all comes down to whether the other operator can hear you, and his station is powerful enough for you to hear his signal bouncing off the moon. 

Let me just lift a little from that prior piece so that you don't have to keep another tab open.

The next big concern is the same as every communications link everywhere else: the amount the signal attenuates - weakens - over that 500,000 miles. The term for this is path loss, and back in the Voyager article, I used a handy form that gets you within less than half a dB of the more theoretically-backed equation.  

Path loss in dB = 37 dB + 20log(f) + 20log(d)  where,f is the frequency in MHz and d is the distance in miles.

So PL = 37 + 20log(50) and 20log(500,000) or 185 dB.  

Let's say we put 1000 W out of our antenna (it could be less power in the transmitter and more antenna gain, or a simpler antenna and more out of the transmitter).  That's +60 dBm (power compared to 1 milliwatt in 50 ohms) or one million milliwatts. 

That means the signal coming back is  +60dBm output -185dB path loss or -125 dBm at our receiver input. In a 50 ohm receiver, that's 0.13 microvolt (130 nV). Is that usable?  I almost hate to say this, but it depends.  It's weak for a 0.13 microvolt SSB (phone or voice) signal, but experienced CW operators won't have much trouble if it's a Morse code (CW) signal.

The thing you're looking for here is the sensitivity of the system - the heart of how we used to buy receivers as hams - and in a room temperature receiver the noise floor can be given by 

 Noise Floor in dBm = -174 dBm + 10log (receiver BW in Hz) + Noise Figure (or NF)

Let's say our 6m receiver has 4 dB NF, the CW noise floor could work out to  -170 + 10Log(500) or -143 dBm.  That means the -125 dBm signal from the moon has an 18 dB SNR - piece o' cake.  In SSB mode, I'll make that -170 + 10Log(2000) or -137 dBm and the resultant SNR is 12 dB, or 6 dB worse.  A 12 dB SNR for phone is not as easy to understand as an 18 dB SNR for CW, but it's not bad. Some speech compression to raise the volume of the quieter parts of speech would help.

Here I have to add there's a lot more potential places for this to break down in reality.  I alluded to how much signal is lost on the reflection from the moon:

Wait.  There's a nasty assumption hidden in there, that the reflection from the moon is perfect. No signal loss, it just changes direction. That implies the signal reflected back has an angular diameter less than the moon - or some would be lost  around the edges.  The diameter of the moon is just over 0.5 degree, which is very tight for an antenna beam.

Let's say the transmitting antenna's beam is twice the diameter of the moon - 1.0 degree.  That figures to be saying half the signal doesn't reflect back - the return is 3 dB less than the calculated 185 dB.  It also implies that the reflection off the moon's surface is the radio equivalent of  a perfect mirror. There are no losses. While I don't know that the losses are I'd be pretty sure there are some.  The transmit and receive numbers aren't including the actual power at the antenna, and there are always losses in the cables connecting the power amp to the antenna.  There shouldn't be much, tenths of a dB rather than whole numbers, but don't forget it's something to keep track of. 

The place where the most improvement seems to have come is in the Weak Signal digital modes that are available now, especially the WSJT-X software that has taken the amateur radio world by storm.  JT is Joe Taylor, a Princeton University physicist and ham who has developed algorithms that make these digital signal processing tools easy to get into your station. There's more than one mode that is specifically intended for moonbounce.  

JT4, JT9, and JT65 use nearly identical message structure and source encoding (the efficient compression of standard messages used for minimal QSOs). They use timed 60-second T/R sequences synchronized with UTC.  JT4 and JT65 were designed for EME ("moonbounce") on the VHF/UHF/microwave bands.  JT9 is optimized for the MF and HF bands.  It  is about 2 dB more sensitive than JT65 while using less than 10% of the bandwidth.  Q65 offers submodes with a wide range of T/R sequence lengths and tone spacings; it is highly recommended for EME, ionospheric scatter, and other weak signal work on VHF, UHF, and microwave bands.

Unfortunately, I've never seen numbers for things like the SNR or input signal required for given small error rates in the received signal. 

While every example I've worked and included was aimed at 6m, it might be that the easiest band to get started with moonbounce is 2m.  The same gain antenna is much smaller because everything scales by wavelength, so while those names (6m and 2m) aren't really the electrical lengths, 1/2 wave (approx. a yagi element) is 9'4" on 6m, it's 3'3" on 2. On 6m, I have a 5 element yagi that's 12' long.  That scales to 4' 2.5" long on 2m. Or more antenna gain in the same length.  Gain and low noise figure are relatively cheap on 2m compared to microwaves and there are many all mode radios for 2m. 

A screen capture from this video.

There are several videos where guys build a station that they only put in place for EME, like this.