Tag Archives: LOWFS

It’s official: LOWFS will have two POAs

Through an internal design review process and an audit process by ESO for the entire instrument during 2021 and 2022, a high risk of not meeting the performance was identified for the image quality and WCS (World Coordinate System) knowledge, as well as difficulties for maintenance access. As a consequence, major changes to the NGSS LOWFS subsystem were requested.

Since then, the LOWFS team from CAB and University of Durham has been working with the NGSS system team on proposing an updated LOWFS with a two NGSS arm design that would allow for better image quality performance, WCS performance and a different mechanical architecture that allows for easier maintainability and access.

The development work on LOWFS is restarting on a new baseline for many of its content and the main objectives were to present and discuss the updated set of LOWFS requirements, the updated baseline for the architecture and possible changes to NGSS system interfaces. After a meeting at which the new design and architecture proposal for the LOWFS was agreed and subsequently approved by the ESO at the end of last year, the baseline architecture will be further developed during preliminary and critical design work for 2023-2024.

Apart from the addition of the second arm, LOWFS will be placed on a fully extractable bench with rail plus carriage extraction mechanism and also, the NGSS environment temperature increases from -15ºC to +2ºC.

The following image shows the new baseline design for the LOWFS:

New architecture for the LOWFS. Credit: HARMONI Consortium.

How much does this situation affect the current POA design?

The design and architecture of the POA are significantly affected by all these changes. Although the configuration remains “theta-phi”, a “shoulder-elbow” architecture will now be implemented, whereby the primary axis is no longer a rotating mechanism and becomes a fixed device that moves the assembly formed by the secondary axis and the POM (Pick-Off Mirror), which will be steerable.

POA early concept design.

The good news is that all the work, knowledge and studies developed for the secondary axis’ engineering model of the previous architecture will be used for the primary axis of the new design, so it could be said that the mechanism has been renamed, but it remains the same. In addition, both axies will now use the same brushless motor technology and “direct-drive” type configuration.

POA Secondary Axis Engineering Model

The LOWFS Pick-Off Arm (POA) is an assembly of two rotating mechanisms in a “theta-phi” configuration inside the NGSS system. Its purpose is to patrol the telescope’s technical field accurately, to ensure image stabilisation and to transmit the light captured by the Pick-Off Mirror from the technical field to the subsystem detectors.

Due to the high precision to be achieved and all the restrictions imposed by the requirements (especially in volume, as the POA has to be very small in height) it has been necessary to study and compare different motion technologies in depth.

One of the first proposals for the secondary axis was to use stepper motors with a 5000:1 three-phase reduction gearbox. However, since the backlash of the mechanism has to be almost non-existent and the mechanism took up a lot of space, this solution was not satisfactory. Therefore, it was decided to implement a brushless motor in direct drive configuration.

To validate the electromechanical design of the POA’s secondary axis, the CAB HARMONI team has developed an engineering model of the secondary mechanism. The purpose of this prototype is to verify through numerous tests that the chosen technological solution is valid and meets the strict requirements for positioning and repeatability, which are 6.1 μm RMS and 0.8 μm RMS over distances of 3.3 mm respectively at a speed of ±2000 arcsec/s.

It is driven by a 6-pole frameless brushless motor, of 400 VAC power supply and inrunner architecture (the rotor is on the inside and the stator is on the outside). The motor is in ‘direct drive’ configuration, what means that the rotor shaft is directly coupled to the arm, without any gearbox or other mechanism. In addition, an incremental encoder with more than 30,000 lines is used to determine the angular position of the arm.

The following image shows an exploded view of the secondary axis and the structure used to support it. Also, see that cubic mirrors are placed at both ends of the arm to allow the measurement of the position of the arm with external instrumentation, instead of the bending mirrors of the actual design.

Exploded view of the secondary axis and the support structure.

The instrument used to perform the external measurement of the arm’s position is an autocollimator, which is an optical instrument that uses the reflection of a monochromatic laser to measure small angular differences with respect to two axes with a high accuracy (below 1 μrad), but in a very small range (2000 μrad).

Assembly of the engineering model in the laboratory.

After many tests, the results have pleasantly surprised the team and the Consortium alike, as the response of the mechanism is very good and almost within the requirements even without using the final electronics and runnig a fully tuned control loop.