Not logged in
Mon Mar 15 08:55:45 2021, Paco, Anchal, Summary, training,
Mon Mar 15 20:40:01 2021, rana, Configuration, Computers, installed QTgrace on donatella for dataviewer
Mon Mar 15 08:55:45 2021
Reply to this:
Found IMC locked upon arrival.
Since "allegra" was set up as an additional workstation, we tried using it but discovered the monitor
For the sake of debugging, we tested VGA and DVI inputs and even the monitor lying around (also labeled "allegra") with no luck. So <ssh> it is for now.
IMC Input sensing matrix
Rana joined us and asked us to use Rossa for now so that we can sit socially distantly.
Attaching some intermediate results on our analysis as pdf and zip file containing all the codes we used.
We used channels C1:SUS-MC1_USSEN_OUTPUT (16 Hz channels) and so on which might not be the correct way to do it as Rana pointed out today, we should have used channels like C1:SUS-MC1_SENSOR_UL etc.
During the input matrix calculation, we used the method of TF estimate (as mentioned in
) to calculate the sensor matrix and inverted it and normalized all rows with the maximum absolute value element (we tried few other ways of normalization with no better results either).
We found the peak frequencies by fitting lorentzian to the sensor data rotated by the current input matrix in the system. We also tried doing this directly on the sensor data (UL for POS, UR for PIT, LR for YAW and SD for SIDE as this seemed to be the case in the old matlab codes) but with no different results.
The fitted peak frequencies, Q and amplitude values are in fittedPeakFreqs.yml in the attached
Uploaded Mon Mar 15 12:52:10 2021
Uploaded Mon Mar 15 12:54:23 2021
Uploaded Mon Mar 15 12:55:00 2021
Uploaded Mon Mar 15 12:55:25 2021