Jamie, Dave, Fabrice,

We have implemented the modifications in the isi2stage master model that were necessary to:

1)  Allow  good functioning of the guardian

2)  Investigate the subtraction scheme we have been talking about to improve teh RZ (Yaw) isolation.

We are testing these changes so we can decide which ones to propagate (ECR, svn commit, install on other units, other sites...)

Details of the changes:

- Added a block in the ST1 block, to filter the T240RZsignal induced by the Z drive. This block is called T240 subtract

- Changed the way the WD "ignore" the T240 when they are not in loop. This is now done in the T240 block instead of being done in the c code. It permits to keep the c code generic for all other types of sensors. The c code is back to 9 inputs. I removed the 10th input in Stage 1 and Stage 2 "Muxes" entering in the Simulink. We must check that everything i s compatible with the HAM-ISI and BSC-ISI models

- Changed the logic to calculate the state of Stage 1 ISO. It is now done with boolean operators. The output returns a "0" if all gains are 0, and a "1"  if any of the gains in non-zero. Added an Epics output to monitor this variable.

- Added some logic to reset the T240 counter when the isolation is tuned on.

The model has been compiled, installed and started on ITMY.   The T240 counter does reset when the isolation is turned on. However, on the last test I did with the guradian before to hand the unit back to commissioning, it tripped on the T240s. To be investigated tomorrow morning.

Other problems to be fixed:

- the GS13s trip when the boosts of the HEPI horizontal loops are engaged. We must check if it's a matter of adding phase margin or adding some ramping time.

- all the safe.snaps must be updated to record the T240 saturation counter parameters

- for some reasons, the cont2b filters in ITMY Stage 1 X,Y, Z had disappeared. I assumed they were just constant values of "1" and added such filters with foton. To be checked.

- some HEPI ETMX boost filters were not loaded. We loaded constant values of "1" so we could keep working on the guardian. To be fixed.