Difference between revisions of "Mar 24, 2011 Calibration/Monitoring"

From GlueXWiki
Jump to: navigation, search
(Minutes)
(Minutes)
Line 30: Line 30:
 
##* Large number of nodes, low speed, modest data volume
 
##* Large number of nodes, low speed, modest data volume
 
##* CAN needs special controller. One ethernet to CAN bridge supporting a few segments, where each segment has up to 100 nodes.
 
##* CAN needs special controller. One ethernet to CAN bridge supporting a few segments, where each segment has up to 100 nodes.
##* Questioned the need for CAN complexity and suggested ethernet, Note: Later in discussion when 100 controllers were being considered, this suggestion was less attractive
+
##* Questioned the need for CAN bus complexity for the monitoring application and suggested ethernet, Note: Later in discussion when 100 controllers were being considered, this suggestion appeared less attractive
 
## Kappos reminded everyone of capability of present controllers
 
## Kappos reminded everyone of capability of present controllers
 
##* setting triggering modes
 
##* setting triggering modes
 
##* setting voltage  
 
##* setting voltage  
 
##* flashing control
 
##* flashing control
## Discussion focused on that drove the needs for a control system
+
## Discussion focused on the requirements that drove the needs for a control system
 
##* Elton mentioned that selective triggering was one reason for control. But after consideration it was decided that a passive system with four trigger lines would meet the requirement
 
##* Elton mentioned that selective triggering was one reason for control. But after consideration it was decided that a passive system with four trigger lines would meet the requirement
 
##* It was also agreed that a single voltage could be supplied to all LEDs (no specific settings per control board)
 
##* It was also agreed that a single voltage could be supplied to all LEDs (no specific settings per control board)

Revision as of 12:56, 24 March 2011

Teleconference Time 11:00 EDT

  • Connection: ESNET

Background information

Action Items

  1. Can George calculate the % deviation from linearity on his plot?
  2. Christina will probe Kapos for cabling and other board issues.

Tentative Agenda

  1. Communications
    1. Summary of CAN protocol for Fcal bases (Paul)
    2. Comments on Hall D slow controls (Elliott/Hovanes)
  2. Points for Discussion (Elton)
  3. Updates from Athens (Christina)

Minutes

Attending: Christina, George, George Lolos, Kappos, Stigaris (Athens); Matt, Paul (IU); Elton, Beni, Eugene, Hovanes, Elliott, Dave (JLab); Zisis, Andrei, student? (UofR).

  1. Communications
    1. Summary of CAN protocol for Fcal bases (Paul)
      • Fcal bases require voltage setting and read back, temperatures, 7 ADCs on board.
      • Large number of nodes, low speed, modest data volume
      • CAN needs special controller. One ethernet to CAN bridge supporting a few segments, where each segment has up to 100 nodes.
      • Questioned the need for CAN bus complexity for the monitoring application and suggested ethernet, Note: Later in discussion when 100 controllers were being considered, this suggestion appeared less attractive
    2. Kappos reminded everyone of capability of present controllers
      • setting triggering modes
      • setting voltage
      • flashing control
    3. Discussion focused on the requirements that drove the needs for a control system
      • Elton mentioned that selective triggering was one reason for control. But after consideration it was decided that a passive system with four trigger lines would meet the requirement
      • It was also agreed that a single voltage could be supplied to all LEDs (no specific settings per control board)
      • Triggering would be achieved via four external signals provided to all controllers that would pass these on to each of the four chains per control board.
  2. Points for Discussion (Elton)
    1. Elton went over the slides and space constraints for the monitoring boards
    2. With the simplification of the control board, the system is likely to fit into the available space.
  3. Updates from Athens (Christina)
    1. Kapos explained how the circuit would be rearranged from star to daisy-chain configuration:
    2. The new daisy-chain board would add a buffer and capacitor to accept high impedance as a driver for the LED trigger.
    3. Also, the new control board simplified would require considerably less power. The voltage bias signal would be filtered only.
    4. There is also a new scheme to install the LED at a 15 deg angle to the board (presently it is parallel), to each attachment of fiber as input to the light guide. This should be considered carefully for mass production of LED boards.