Lecture on Aerospace Soldering

Randall Pearl gave a great lecture on his experience soldering satellites and ground control equipment at RCA Astro Electronics in the '80s.

Here's the PowerPoint slides.
Chris M. took a video recording of the talk - ask him if you want a copy.

FC/GFE USB Communication Requirements

Today we had the flight computer (FC) enumerate the node5 generic front end (GFE) for the first time! Yay! Now it's time to decide how we're going to send data. Dave, Jamey, Keith, Dan and Andrew discussed ideas for the USB communication protocol between the two.

Each node has it's own VID/PID.
- Use 0xFFE as VID (common open source VID)
- PID defines what kind of node.
- Communication protocol version is in the USB vendor specific version number. Official name of this field?
Use endpoint 0 (EP0) transfers for all node state information
- Any commands for setting or querrying the state of a node uses EP0 because there's less code to write (we don't have to write the protocol).
1 kHz is obviously our maximum sample transfer rate.
- Control system only needs one sample per KHz, but we'd like all the raw data for logging.
- We could have all nodes report in at 1 KHz, and then just repeat slower data.
The node presents all of its data for all of its sensors in one isoc packet frame.
- 1024 bytes max per 1 KHz frame, which is plenty of data per frame.
- Have to worry about data rates for the 12 (actually 5) Mbps data limit.
- Might need to tune the data rates for this flight with the PPC v1.1 USB host controller.
Each data packet from the node is sent as a pile of data
- Need some metadata for each data transfer: number of samples, errors, failures of sensors, etc.
- Possibly encode meta data in the ends of sensor ranges (magic in-band values for validity).