In our transition from LV2 to LV2.3, we've moved from S to C band for our telemetry because it's Just Better Up Here - we get smaller antennas, and less noise. More importantly, it gets our amateur TV out of the L band and into the now vacant S band, reducing interference with the also L band GPS. In the amateur radio world, running 802.11x in the amateur radio regulations (FCC Part 97 is called High-speed multimedia radio (HSMM).
- Should be USB-based.
- Should support a knon C-band standard, e.g. 802.11a.
- Should have soft MAC or other features which would allow us to customize how the radio behaves.
- Should be small and cheap, with an external antenna connector.
- Should support ad-hoc mode.
- Must be supported in the Linux kernel, or have a communication interface supported by Linux.
We found the D-Link DWA-160 Xtreme N Dual Band USB Adapter, which fit all of these features. Based on the Atheros AR9170 chipset, it's small, not too spendy (~ $70), and seems to be supported by Linux. We confirmed this on 2009/06/30, with a test of three cards chatting in Ad Hoc mode.
It outputs 15 dBm in the C band.
iwlist, the adapter supports the following channels:
Channel 36 : 5.18 GHz Channel 40 : 5.2 GHz Channel 44 : 5.22 GHz Channel 48 : 5.24 GHz Channel 52 : 5.26 GHz Channel 56 : 5.28 GHz Channel 60 : 5.3 GHz Channel 64 : 5.32 GHz Channel 149 : 5.745 GHz Channel 153 : 5.765 GHz Channel 157 : 5.785 GHz Channel 161 : 5.805 GHz Channel 165 : 5.825 GHz
The adapter literature claims it supports 5.15GHz to 5.35GHz and 5.725GHz to 5.825GHz. That's good, except the high end ought to be a bit higher, like 5.83 GHz to really run in channel 165. Huh.
If possible, we'd like to aim dead set in the middle of one of those two bands, so we have room to move up or down in order to match whatever our actual cylindrical patch antenna center point ends up being. And of course, we like higher frequencies, so maybe aim for channel 157 (5.785 GHz)? Although the higher band has a ± 40 MHz range while the lower channel has a ± 60 MHz range.
The connector seems to be a Murata MM8430-2610, which is some sort of bizarre test point. It physically switches out the antenna when you insert the test probe.
- and page 9 of http://www.murata.com/catalog/o30e8.pdf
We're now strongly leaning towards unsoldering the connectors and soldering on coax right to the pads... but we still have two connectors, what do we do with that?
- Must support (1,3,5) watt output (see the link budget page)
- Must be small enough to fit in the rocket
- Should have decent locking connectors, like SMAs.
- Must have a (3.3,5,12) V power input
We found the RF LINX Corporation's Antennafier 5800CAE series power amplifier, and it seemed like a good fit. It 32dBm peak TX amplifier (1.5 W) and a 10 dBm RX amplifer with a not-so-great 2.5 dB noise figure. It takes 12 V in at < 1 A.
It's a very compact and hackable little PA, so it should work nicely for mounting in one of our RF side chambers in the avionics system.