Next: Contact people   Up: Analysis mini-topics   Previous: Crosstalk

Last modified: 7-OCT-1999


Numerous parameters are monitored through Slow Controls (and recorded to tape through MiniDAQ), while other information appears in a data header from the six readout boards which is usually recorded to tape. Most of this information has not been adequatedly checked.

  1. Temperature stability
  2. Readout board voltage stability
  3. Table compressibility
  4. Readout board integrity checks
  5. Geometry event integrity

Temperature stability

The sector temperature is controlled by the FEE and readout-board cooling manifolds; using temperature information from thermistors mounted on the sector and read out through Slow Controls, we should determine the stability of this system.

Readout board voltage stability

The readout board voltages are read out through Slow Controls. We should determine the stability of this system.

Table compressibility

Noise tables appear to be compressible by a factor of three with gzip; one could check the compressibility of other data types, and by other algorithms. DAQ (Marcus Schulz of BNL) has been studying data compression using Huffmann encoding and other methods, but has not looked at the problem using our Cosmic-Ray Test data formats.

Readout board integrity checks

The readout board data header contains 8 bytes of data to verify accurate data transmission (the string ``deadface''). There is also a readout board synchronization byte that is supposed to be reset at the beginning of a run and retain synchronization. Finally, there is the trigger token, which should contain a fixed value at present. We should verify that all of this information is reliably present.

Geometry event integrity

Small anomalies have been seen in geometry events through spring, 1999 (at least). Usually this involves corruption of the first time bin. This should be rechecked for the new electronics.