Hello again from the thermal testing team at the ESA CubeSat Support Facility, Belgium.
After suiting up for the cleanroom again we began the day with our first functional test in vacuum. This was a Very Reduced Functional Test (VRFT) that was designed to simulate the sequence of operations PROVE will perform while in orbit. This includes the initialisation of cameras, imaging with both cameras in parallel, shutdown of cameras and transfer of image data to storage. After this successfully finished it was onto the pump down to high vacuum (~10-7 mbar, essentially the cold cold vacuum of space!).
Once high vacuum was achieved, another VRFT was performed, but this time with less success. Our visual camera had failed to collect images, and from this point forward the BeagleBone was only intermittently able to ping the camera at all. After completing further functional tests on the Basler, it appeared there were three possible culprits:
- Vacuum: The visual camera does not behave reliably in a vacuum
- Temperature: The camera does not behave reliably at the temperature we were testing
- A combination of both affecting the cable and connector: The release of gases from spaces inside the connector and changes in temperature had loosened the mechanical connection of the connector.
Once we had considered these options it was decided to vary the chamber temperature from ambient and see if a pattern emerged that might shed some light on the situation. Sadly, when varying the temperature the camera behaviour seemed almost random, with some tests passing and others failing under the same conditions. A new failure mode was also noticed, not seen outside vacuum, where the camera would ping and even grab a single image but then loose connection and fail the test. In light of all this, and given that the Basler had performed fine in a low vacuum, the most likely cause appeared to be the third option, the cable and connector.
The best course of action was now to vent the chamber, inspect the cable and try to repair any mechanical problems with it. The spare visual camera was first tested to see if the problem persisted, which it did, indicating that the camera itself had not suffered some internal failure. After adding a number of washers to constrain the bolts holding the D-sub connector that carries the data interface to the camera, the camera behaviour was much better, consistently passing tests. The chamber was resealed and pump down was initiated. Finally, we were ready to start the thermal cycling of the TVAC test!