Jeff, Oli
Today we tested out the SR3 estimators for both Pitch and Yaw (86491). The ASC loops where we see a difference are DHARD, MICH, SRC1, and SRC2, since these all go through the SRC to their output ports. As expected, we don't see any effect on CHARD, PRC, or INP.
I've plotted comparisons of the ASC control and error signals during those times along with a reference time of earlier that morning.
The ultimate conclusion is that the SR3 P and Y estimators do work to make some of the ASC noise better, and the good thing is that in the places where we see that the estimator makes the noise worse, they all seem like easy improvements/fixes! The amount of improvement that we see doesn't look like much, but that's also expected since it's just SR3 that has the estimator out of the several suspensions in each ASC path. We expect that as we add more estimators along these paths, we will keep picking away at the suspension noise. We already see really impressive broadband improvement between 1-5 Hz on SRC1 and SRC2!
Main areas of concern:
- For all ASC loops and both P and Y, we see heightened SR3 resonances on around 0.7 Hz when the SR3 P estimator is on. This is because the OSEM P blend filter was made without including the 0.7-0.8 Hz mode. That is being added in and that should solve that issue
- Between 0.1 and 0.2 Hz for some loops, turning on the estimators puts in some extra noise from the GS13s, which aren't as sensitive at those lower frequencies. To solve this, we will be adding a high pass filter above those frequencies
- A few peaks aren't being damped as well with the estimator as they were with normal damping; however, now that we've shown that the estimator decreases our noise(86455, 86503), we are hoping to be able to increase gain somewhere in the estimator system (possibly in SR3_M1_EST_Y_FUSION_MEAS_BP) to better damp our resonances while still seeing improvements in noise as compared to normal damping
Here are the times used to grab the traces:
Estimator ON for SR3 P and Y (RED) 2025-08-21 16:35:00 UTC
Estimator ON for SR3 Y (BLUE) 2025-08-21 15:34:06 UTC
Estimator ON for SR3 P (GREEN) 2025-08-21 16:05:00 UTC
Estimator OFF (PINK) 2025-08-21 10:27:00 UTC
| ASC | Pitch | Pitch Notes/Comments (besides 0.7 Hz SR3 P estimator resonances) |
Yaw | Yaw Notes/Comments (besides 0.7 Hz SR3 P estimator resonances) |
| DHARD | Error | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON | Error | 0.2-0.4 Hz - Transient noise with only Y Est ON? 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON |
| Control | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON | Control | 0.2-0.4 Hz - Transient noise with only Y Est ON? 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON |
|
| MICH | Error | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.1 Hz - Every estimator ON configuration lowers noise |
Error | 0.1-0.2 Hz - Estimators ON elevates noise; need to high pass GS13 0.3-0.4 Hz - Transient noise with only Y Est ON? |
| Control | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.0 Hz - Y estimator ON alone adds some noise, but P estimator fixes it 1.1-1.5 Hz - Every estimator ON configuration lowers noise |
Control | 0.1-0.2 Hz - Estimators ON elevates noise; need to high pass GS13 0.3-0.4 Hz - Transient noise with only Y Est ON? 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON |
|
| SRC1 | Error | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.1 Hz - Every estimator ON configuration lowers noise |
Error | 0.15-0.4 Hz - Transient noise with only Y Est on? 0.6 Hz - Every estimator configuration lowers noise 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.0 Hz - Estimators ON elevate noise 1.2-3.1 Hz - Every estimator ON configuration lowers noise |
| Control | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.1 Hz - Every estimator ON configuration lowers noise 3.0-10.0 Hz - P and Y estimator damping lowers noise; this is good but we don't know why it's doing that there |
Control | 0.25-0.4 Hz - Transient noise with only Y Est on? 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.0 Hz - Estimators ON elevate noise 1.2-3.0 Hz - Every estimator ON configuration lowers noise 3.0-10.0 Hz - Just P estimator ON gives us the best noise performance, and P and Y estimators ON gives us the worst noise performance. This is NOT good, and also doesn't match with what we see with the SRC1 P Control signals, so we don't know why it looks like this at all |
|
| SRC2 | Error | 0.1-0.2 Hz - Estimators ON elevates noise; need to high pass GS13 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.1-3.5 Hz - Every estimator ON configuration lowers noise |
Error | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.0 Hz - Estimator ON elevates noise 1.1-2.1 Hz - Every estimator ON configuration lowers noise 2.1 Hz - Estimator ON elevates noise 2.5-5.0 Hz - Every estimator ON configuration lowers noise |
| Control | 0.1-0.2 Hz - Estimators ON elevates noise; need to high pass GS13 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.1-3.5 Hz - Every estimator ON configuration lowers noise |
Control | 0.6-0.8 Hz - SR3 P resonance elevated whenever P estimator is ON 1.0 Hz - Estimator ON elevates noise 1.1-2.1 Hz - Every estimator ON configuration lowers noise 2.1 Hz - Estimator ON elevates noise 2.5-5.0 Hz - Every estimator ON configuration lowers noise |
