empty .nc files in production scripts

@mschleiss (cc: @msourzac, @rmackenzie )

There is a bug in the script which is running in production disdrometers, which I dont fully understand, but after a whole day of data being written to the .nc, then it the files is empty.

This issue is present in commit 3baed463 which is the "version" or

This however does not happen with PAR003,PAR008, or PAR009 that are running more recent versions of the software.

I apologize for not having noticed this issues before.

I will:

• list instruments and dates have been affected.
• merge !29 (merged)
• update git head in all disdrometers
  • update git head in parsivel004pi
  • update git head in parsivel006pi
• parsivel006pi investigation
• check incoming netCDF files
• remove empty data files from PD