First on-sky validation of the 3-sided near-infrared unmodulated pyramid wavefront sensor #
We carried out the first on-sky validation of our 3-sided near-infrared unmodulated pyramid wavefront sensor (3s-NIR-uPWFS) at the Observatoire de Haute-Provence (OHP), integrating into the PAPYRUS adaptive optics bench.
- Why pyramid?
It has very high sensitivity.
- Why unmodulated?
Because modulation reduces the sensitivity and limits the speed at which the adaptive optics can operate, and we do not want that.
- Why infrared?
Using the PWFS in the infrared increases its linearity range.
- Why 3 sided?
A 3-sided unmodulated infrared pyramid is significantly easier to manufacture with high tip precision compared to a 4-sided one.
Installation #
We installed the 3s-NIR-uPWFS on the infrared arm of the PAPYRUS bench.
The existing setup already included a 4-sided visible pyramid wavefront sensor (4s-VIS-PWFS), and the system was designed so that:
- both pyramids could operate in parallel
- a simple software switch allowed selecting which pyramid controlled the loop
This configuration enabled a direct comparison between the 3s-NIR-uPWFS and the 4s-VIS-PWFS.
First camera images #
Calibration #
We calibrated the system using standard modal push–pull techniques. A sequence of Karhunen–Loève (KL) modes was projected onto the deformable mirror (DM). These modes provide a optimal basis for atmospheric aberrations. The calibration video shows these KL modes being beautifully applied to the DM and seen by the 3s-NIR-uPWFS.
Open loop and closed loop observations #
Once calibration was complete, we observed a star in both open loop and closed loop, comparing the 3s-NIR-uPWFS and the 4s-VIS-PWFS. In the video we can see the open loop case with strong PSF motion and distortion. Closing the loop with each pyramid dramatically improves the PSF. We get a PSF with trefoil when we close the loop with 3s-NIR-uPWFS, and a PSF with astigmatism when we close the loop with 4s-VIS-PWFS. The improvement is greater with the 3s-NIR-uPWFS.
More observations #
We took several observations over the course of 3 nights. During a long 5‑minute observation, the integrated PSF peak obtained with the 3s-NIR-uPWFS is twice as high as that obtained with the 4s-VIS-PWFS, meaning more than twice as much light is delivered by the AO system for science. On an average across all observations the the performance in terms of Strehl was 1.8 times higher when closing the loop with 3s-NIR-uPWFS compared to performance when closing the loop with 4s-VIS-PWFS. We also did tests with an advanced controller DD4AO which further improved the performance.