One important area to consider is comms. How am I going to communicate with the satellite, and how is it going to communicate with the ground? For now I'll just talk about the space to ground segment hardware. I won't include discussions about licensing as that's for another time. The spacecraft is going to need a radio and an antenna, along with sufficient power to enable the signal to reach the ground.
Out of the 4 PQ "veterans" in space, at least 2 ($50Sat and T-LogoQube) are known to have used the HopeRF RFM22B radio module. This is a tiny 16x16mm RF tranceiver module, that is capable of +20dBm output (That's 100mW for those that haven't googled it yet), in the UHF frequency range. It is based on the Silicon Labs Si4432 Radio IC. Despite the seemingly limited power output, some of the data types can be detected almost horizon to horizon - approx 2900km range! Data packets can also be decoded at up to 900km with just a 10db gain Yagi and a LNA. Not bad for a sub $25 radio module!!!!!
Compare this to Cubesat Radio systems that cost well over $5000. GomSpace sell one for 8000 EUR with a 3W transmitter. Clyde-Space have one for $8600 , with 500mW to 2W RF output.
There's a thread on the DIYsats forum here about this and other HopeRF radio modules.
So where to from here?
So the RFM22B is a proven performer for PocketQubes. Its low power requirements make it a good choice, but it does have some limitations. The data rate used by $50Sat is 1kbps. This is fine if your payload isn't very data intensive. But what are the alternatives? Here's a few areas I am looking into.
HopeRF make another module - RFM23BP. This is similar to the RFM22B, except the power output is 500mW. The higher power should allow higher data rates. The issue with this module is the higher electrical power requirements, which can affect how the PQ power subsystem is designed.
Still on the HopeRF parts list is the RFM69HCW. This module appears similar in specification to the RFM22B, but is based on the Silicon Labs Si4463 module. Now Silicon Labs state that the Si4432 shouldn't be used for new designs, although I don't think they intended their parts to be used in space! The Si4463/RFM69HCW uses an new API style approach to programming. While touted as being simpler and more efficient, testing from the $50Sat team has so far proved otherwise.
Silicon Labs also state that a low cost external FET can be used to boost the output to 27dBm/500mW.
Testing would need to be conducted to validate this approach.
Another HopeRF product is the RFM26W . While similar again, it is described as a device that "operates as a time division duplexing (TDD) transceiver where the device alternately transmits and receives data packets."
On to the RFM96W and the RFM98W. These are based on Semtech IC's, probably the SX1276 . These have the capability of using a patented LoRa modem, which can potentially increase the devices sensitivity, and therefore link budget. Further investigation is required as this ic has duty cycle limits, and whether the LoRa function works when the radio isn't configured for spread spectrum usage.
Then there's the HM-TRP. These are possibly most commonly used in the 3D Robotics Wireless Data Module. These also have 100mW/20dBm output, but operate as a simple transparent FSK transceiver - meaning that you just feed it UART data and it converts it to RF.
Another option could be to use a FEM ( Front End Module) in conjunction with the RFM22B. RFMD make a module - RFFM6403 . This is a 1W Power amplifier, but it can also operate in Bypass mode if required. It can also be a LNA for the receiver. This part is brand new and is available now in a reel, or later in March if you need single items. Definitely worth evaluating.
Other alternatives are pre-made modules from Radiometrix. They have some compact VHF / UHF transceivers. ( maybe not as cheap as $25 though).
Then there's 2.4ghz modules........ I haven't looked that far into this, but given my intention to use a USB SDR Dongle as a groundstation, that's out of their frequency range ( although I did find an interesting article recently using some cheap hardware as a down-converter. See here for info)
Conclusion
The only conclusion so far is that I really need to get my Amateur radio license and start doing some testing! I've purchased some of the HopeRF modules for testing, and have made the rookie mistake of buying the bare modules, sans breakout board. Half pitch pins don't fit in normal breadboards!
Fortunately bare breakout boards for the RFM22B are available from Modtronics in Australia. Others should take my advice and get populated boards such as those from Modtronics or from Sparkfun in USA.
As for the other modules, if you have any skills in KiCad, you can design a basic breakout and get the boards manufactured relatively cheaply from places such as Seeedstudio or Itead Studio.
I'll let you know how I get along!
