Since UL coil actuation is lost, we modified the output matrix of ITMY to use only UR, LR and LL face coils for POS, PIT and YAW actuation. The output matrix was changed to following:
| |
POS |
PIT |
YAW |
SIDE |
| UL |
0 |
0 |
0 |
0 |
| UR |
1 |
1 |
0 |
0 |
| LL |
1 |
0 |
1 |
0 |
| LR |
0 |
-1 |
-1 |
0 |
| SIDE |
0 |
0 |
0 |
1 |
After this change, the damping was still working as good as before. I took PIT to POS/PIT/YAW and YAW to POS/PIT/YAW coupling measurements by exciting C1:SUS-ITMY_ASCPIT[YAW]_EXC and seeing effect at C1:SUS-ITMY_SUS[POS/PIT/YAW]_IN1 when the damping loops were off. Attached are the results. We were able to reduce PIT to YAW and YAW to PIT coupling by 10 dB by this simple change in output matrix. More coil balancing or off-diagonal termsmight help more and should be attempted if required. The coupling to POS did not change much.
Note that attachment 1 shows transfer functions from excitation point to the DOF sensing inputs while attachment two looks at ratio of C1:SUS-ITMY_SUS[POS/PIT]_IN1 to C1:SUS-ITMY_SUSYAW_IN1 which is the actual quantity of interest. I didn't repeat the PIT measurement due to lack of time.
Also note that all such measurements are being recorded in our new measurements git repo. We'll populate this repo with diaggui template+data files as we do measurements. |