Today we tested our X-band radios and they are working fine. That is a huge relief.
We have two separate radios on each of four spacecraft: an S-band transmitter (2-4 GHz, roughly the same band as a microwave oven) and an X-band transmitter (7-11 GHz, i.e. about two octaves higher). We use the S band for commanding and for basic telemetry, but we're limited to about 5 Mbps over that radio. The X band gives us a higher data rate, about 25 Mbps. That's important, because it means we can downlink a whole day's worth of data in a single ground station pass.
During commissioning, we've been sighting in exactly how to track the spacecraft. NORAD provides tracking information (via space-track.org; you can also see those data as visualizations of every orbiting object, at n2yo.com), but we produce even better tracking/prediction information ourselves, from on-board GPS receivers on each spacecraft. NORAD is good to a few hundred meters. GPS is accurate to within a few meters, which is pretty spiffy in itself -- but also gives the best possible prediction for where each PUNCH spacecraft will be, in the sky.
Tracking information is important, because ground stations have to know pretty much exactly where to look for each satellite as it goes overhead. They have to "paint" the spacecraft with a tiny pencil beam. We use roughly 7 meter ground station dishes, so the diffraction limit (1.22 λ/d) for the S band is about 15 arcmin -- or something like 5km wide at PUNCH. The X band is four times narrower -- just over 1km wide at orbit!
For a while our GPS tracking data were wildly inaccurate: instead of a few meters, it was a few kilometers off. That boiled down a to timestamp issue: we were using local computer time instead of the reported "GPS time". Sometimes those were off by up to a half-second. At an orbital speed of about 7 km/sec, that's a big deal. It meant that we could easily lock on with S band, but often missed the spacecraft entirely with X band, and those ground passes would be "missed".
So tracking was a really big deal. Last week, we sorted that out and started using better GPS timestamps, and our tracking errors dropped by a factor of 1,000. Today, we tried several downlink passes with the X-band radios, and they worked like a champ. Phew.
We're doing a little more shakeout Thursday and Friday, and we're planning on opening our first doors on Monday the 14th.
