P3K RTC AlgorithmControls

1
P3K RTC Algorithm/Controls

• Mitchell Troy
• With much thanks to Chris Shelton
• System Engineering meeting
• 8/9/2007

2
Modal Versus Zonal Control

• Assume A Actuators, M modes to control, S
  of slope measurements, G M gains (one for each
  mode)
• Zonal control
• 1 matrix multiply (S by A matrix)
• Modal control option 1
• 2 matrix multiplies (S by M and M by A matrices)
• M can be as large as A
• Allows for on-fly adjustment of gains for each
  mode
• Model control option 2
• 1 matrix S by A
• Adjust modal gains (G) offline and generate
  control matrix ( (S by M) X G X (M by A)) offline
• Must be able to hot swap reconstructors
• Suggested requirement
• Single matrix multiple S X A
• Must be able to hot swap reconstructors

3
Woofer/Tweeter Control

• How do we split the Woofer/Tweeter commands?
• Two basic methods
• Split commands based on spatial frequency
• Let tweeter DM do everything and then offload to
  woofer
• If split based on spatial freq.
• Still MAY need to handle saturation on tweeter
• If split based on temporal freq.
• Might get better temporal correction using both
  DMs
• GPI PDR Design
• Current design splits spatially
• Does not appear to handle saturation of tweeter!
• Sheltons suggestion for P3K
• Implement BOTH

4
GPI AO block diagram
Slide curtsey of Jean-Pierre Véran from GPI
5
P3K AO block diagram
6
Discussions/Issues

• Is the offload from tweeter to woofer going to
  the correct place?
• We should eventually design the whole P3K control
  design (dont forget)
• design the TT path for woofer and TTM
• Leaky terms