SETI League EME (Moonbounce) Station

Part of being able to copy the beacon signal is knowing what frequency it's on. Your job is to compensate for Doppler shift by knowing where the beacon is (shown on the status page), knowing where you are, and calculating the shift caused by the moon's relative motion. Our job is to keep the beacon on the stated frequency, 1296 MHz exactly, to the best of our ability.

Realistically, due to the physics of the EME path, frequency accuracy of more than ten Hz or so isn't critical.  Even so, it's fun to keep it right, and we have to have some way to know if it's even that close. So, disregarding the two-watches theory, we use two watches. Specifically:

1:  An atomic standard - an Efratom Rubidium standard.  This has a typical accuracy of parts in 10^11th power, or 1 Hz in 100 GHz.

2:  A GPS standard, which uses a disciplined quartz oscillator whose frequency is controlled by reception of GPS satellites. 

This, too, has long-term errors that are even smaller than the Rb oscillator, although its short-term stability is slightly poorer. This standard was contributed to the project by Trimble Navigation.

Both standards derive their frequency accuracy independently, so as long as they agree they can both be assumed to be correct. Periodically we calibrate a high-quality quartz standard oscillator (H-P105B) against the atomic standards, so that if the two should somehow vary the quartz standard can determine which is correct. This hasn't been necessary yet.

The GPS standard (or the Rb standard - they're both the on same frequency) are then used to drive the timebase input of a H-P5334B counter, which is IEEE-488 controllable and has three input channels. One channel can directly measure the 1296 MHz output signal of the H-P 8660A frequency synthesizer.  (Can anyone say eBay again)? The synthesizer is adjustable in 1 Hz steps, and the computer commands it to 1296000000 as well as controlling the output level in 1 dB steps. In addition to using an atomic standard as the timebase for the counter, that same signal, the other standard, along with the 1296 MHz signal are measured in turn during the beacon transmission interval.

Although the raw data usually indicates that the synthesizer is low by 2-4 Hz, we have determined that this is an artifact of the counter, which also measures its own timebase low by, typically .01 Hz (in 10 MHz.). Backup measurement by spectrum analysis and other instruments gives us confidence that the synthesizer is within +/- 2 Hz, and more likely +/- 1 Hz of being correct.