tstana / proto-cubes-obc Goto Github PK

After daq0773.dat, the next file name was daq7740.dat. The error on the serial monitor shows:

Could not open daq7742.dat, file already exists!
response: TRANSACTION SUCCESSFUL
CMD_REQ_PAYLOAD received
  Obtaining latest data from SD card...
Reading data from daq7741.dat
CMD_DAQ_STOP received
response: TRANSACTION SUCCESSFUL
------- Invoking REQ_PAYLOAD ---------
response: TRANSACTION SUCCESSFUL
--------------------------------------
Could not open daq7742.dat, file already exists!
CMD_REQ_PAYLOAD received
  Obtaining latest data from SD card...
Reading data from daq7741.dat

There were more files written on SD card after daq7742.dat. It went up to daq7762.dat but they contained same data inside i.e all HK information was same. SD card was cleared and reinserted.

On startup, the Proto-CUBES OBC tries to find the last written DAQ file on the SD card by sequencing through the currently existing files on the SD card, starting from file DAQ0000.txt. With a large number of files, this process can be very time-consuming.

Fix by storing the last written DAQ file in EEPROM, or (probably easier) in a dedicated file on the SD card, e.g., _lastfile.txt?

When attempting to operate Proto-CUBES with a long cable (preparing for the 2020 radiation test), one of the issues that appears is that only one histogram (of six) is written to data files.

Setup:

• Arduino > (Long I2C cable, ~2.1 m) > CUBES
• Proto-CUBES OBC code from commit a37aed0
  • 400 kHz
  • potential for DAQ timing mismatch between Arduino and CUBES exists...

Four example data files attached (*_Citi is data file equivalent from CitirocUI):

• DAQ0004.txt
• DAQ0004_Citi.txt
• DAQ0005.txt
• DAQ0005_Citi.txt

38 bytes is also the number of bytes sent during HK readout. It could be that the Arduino keeps this number as the recv_len when sending the REQ_PAYLOAD to CUBES...

This issue on a separate repo contains more details on replicating the issue.