How our simple radio interferometer works:

Imagine a boat wake on a flat lake. The crest of a wave may reach the beach at exactly the same time all along the shore. This is really fun to watch.

Now imagine electromagnetic waves (which in our case is light with an 11 centimeter wavelength) arriving at the antennas, spaced 22 kilometers on an almost East-West line.

Please see the diagram below.

Just like a boat wake, the waves reach the antennas at the same time.

In reality the waves coming in from space are not nice and smooth as shown in the diagram. But we filter (or smooth out) the waves in our receivers. The amplitude or distance between crest and trough of the wave is fairly constant for a duration of about a thousand wiggles. That's why I drew the waves the way I did. By the way, the wiggles I drew in the diagram enter the antennas during an interval of just two billionths of a second!

The green arrows show how the Earth is rotating. Note that the two antennas are not moving the same. When the West antenna is moving exactly to the East, the East antenna is "falling" slightly. This means that the waves don't have the same frequency (due to Doppler shift differences). Imagine the changing sound of a train passing you. It's like that. We have to correct for this shift even though it is small. This correction is all taken care of before we capture the data.

The waves received by the antennas are converted to bits and transmitted to HP using two ham radio links. We send a +1 bit if the wave is up and a -1 bit if the wave is down. We do this two and a half million times every second. And we do it all the time, day and night, except when we are repairing something that broke, or when we are checking things to see if they are broken. And, we're also not sending data while we are telling anyone listening that my ham call sign is KC7KQ.

We know that the bits from the East antenna will arrive at HP sooner than the bits from the West antenna, because the distance is shorter. So we delay those East bits to make up for the path difference. We aren't exactly sure about the delay, so we control the delay with a computer and record with several slightly different delays.

There are four possibilities for the bits when we see them at the input of the box labelled X (which stands for multiplication):



         1. The bits are the same and are both +1, 


         2. they're the same and both -1,  


         3. West is -1 and East is +1,


         4. or East is -1 and West is +1.

If only noise is entering the antennas, these four possibilities will be equally likely and each will happen one quarter of the time. If we multiply the bits together and average the result, the average will go to zero.

However, every now and then, the antennas are staring at a black hole, or what's left after a supernova. When this happens, the bits might be the same more of the time. After multiplying them together and averaging the results (look at the products when bits are the same), we will start to see a positive number.

Or, if the waves arrive out of phase (peak at one antenna and trough at the other), the bits will be different more of the time. So we will start to see a negative number.

So a positive or a negative number that differs a bunch from other numbers nearby is interesting.

In the Mabuhay interferometer, we currently average 2.5 million bits per second, for about 0.2 seconds, and we store the result. We change the delay to seven slightly different values to make sure we don't miss something.

radio interferometer

In the very unlikely event we receive a signal from ETI (Extra Terrestrial Intelligence), we will spot a positive number or a negative number in our data. Sort of like the -20 that popped out of the data shown below, but THIS IS NOT FROM ETI!!!. Well, I really don't know where it came from. My guess it is noise, but it could have been a glitch in the radio links. I SAY IT IS NOT FROM ETI.

The spacing in time between lines in the tables is about 1.6 seconds. We use the Julian Date because it keeps track of time and date with one number. The number changes by one every day.

Julian Date the seven delays a number indicating how far off the PC clock is from the Earth's standard atomic clocks 2451405.68811 -6 -2 -6 -5 -3 -3 -1 -9 2451405.68813 -10 -5 -2 0 7 -7 -9 -9 2451405.68815 -2 -8 -2 -6 0 -1 -3 -9 2451405.68817 0 -2 -6 3 -13 1 7 -9 2451405.68819 -2 -2 -3 -3 -2 -9 -5 -9 2451405.68821 0 -1 -3 0 -12 4 -4 -9 2451405.68823 0 0 -1 -5 -5 0 -8 -9 2451405.68825 4 -4 0 -5 -8 -7 -1 -9 2451405.68826 -4 4 -4 0 -11 0 -1 -9 2451405.68828 -5 3 -6 -1 -7 -4 -8 -9 2451405.68830 -5 -9 -1 -4 -4 -11 8 -9 2451405.68832 0 0 -4 -3 -6 -7 -5 -9 2451405.68834 -11 -1 -6 0 -20 -6 -7 -9 <<<<<<<< THIS IS _NOT_ ETI !!!! 2451405.68836 -3 0 -1 0 -1 -1 -9 -9 2451405.68838 3 0 0 0 0 -4 -4 -9 2451405.68840 -6 0 -8 -2 5 -2 0 -9 2451405.68842 2 -8 -4 -9 -5 0 0 -9 2451405.68844 -4 -5 3 -6 -13 2 -2 -9 2451405.68845 3 -5 -7 -12 0 -12 5 -9 2451405.68847 -2 -1 0 4 -1 -4 -4 -9 2451405.68849 -6 -4 -6 2 1 0 -1 -9 2451405.68851 1 -1 -1 6 4 0 0 -9 2451405.68853 5 7 -3 0 2 -5 -2 -9 2451405.68855 4 -3 0 -11 -1 1 1 -9 2451405.68857 -2 0 -8 6 -3 -3 -2 -9 2451405.68859 -1 -3 -3 -2 -5 -5 -2 -9 2451405.68861 0 0 2 -1 -1 -5 -4 -9

An object in the sky, such as a black hole, might look like this:

2451405.68891 -2 0 -1 3 -6 0 0 -9 2451405.68893 1 -6 -2 0 -3 -8 -6 -9 2451405.68895 -6 3 4 -2 -2 2 -6 -9 2451405.68897 5 -2 -4 0 1 -10 -7 -9 2451405.68898 1 -1 -2 -4 3 -3 -5 -9 2451405.68900 -1 0 1 4 0 3 1 -9 2451405.68902 0 -5 -3 3 -1 1 -3 -9 2451405.68904 3 -8 -7 -1 6 -4 -5 -9 2451405.68906 -1 0 0 0 0 -4 -6 -9 2451405.68908 -1 0 1 0 0 -1 -6 -9 2451405.68910 4 1 -1 7 -7 -8 -3 -9 < Note that the fifth data column 2451405.68912 6 -7 2 -3 -1 -6 -1 -9 < numbers are mostly 2451405.68914 5 -3 -1 0 -4 -3 -4 -9 < negative (around here). This may 2451405.68915 -5 4 -2 0 -8 -3 -2 -9 < be from something in the sky, 2451405.68917 -2 -13 0 -9 -5 -6 -2 -9 < or maybe due to some drift 2451405.68919 -7 -3 -7 -7 -4 4 -3 -9 < in the receiver. The receiver 2451405.68921 0 -5 1 -4 -1 -7 1 -9 < is being improved, and more 2451405.68923 -1 -4 -1 -6 1 3 -9 -9 < data is being taken. 2451405.68925 -4 0 -5 4 -6 -10 -2 -9 < By looking at the same point 2451405.68927 0 -6 -2 -1 -8 -3 0 -9 < in the sky two days in a row, 2451405.68929 -2 0 -1 7 -1 -2 -3 -9 < and compensating (about 4 min. per day) 2451405.68931 0 -7 0 -3 -3 1 -10 -9 < for the Earth's orbit around the Sun, 2451405.68933 -2 -3 -1 0 -1 1 0 -9 < we see this plot . 2451405.68934 0 -10 0 1 2 0 -6 -9 (It's not at the same time as this 2451405.68936 -6 3 0 -4 0 0 0 -9 tabular data. Note the interesting 2451405.68938 2 2 -3 1 0 2 -3 -9 daily repetition in the plot.) 2451405.68940 -1 0 -3 3 0 -4 0 -9 2451405.68942 -3 0 -1 -3 0 -6 4 -9 2451405.68944 0 -2 -3 -1 1 0 0 -9