Cryostat for HWU LIDAR field test with JPL SNSPD

June 2019 - present

Gregor Taylor

Dmitry Morozov

Collaborators: Boris Korzh, Matt Shaw (JPL), Aongus McCarthy, Gerald Buller (HWU)

DESIGN:

Cryostat designed by GT. All solidworks models and drawings which parts were manufactured from are available on the groupdrive in QSG\Cryostat Designs\UNIFIED DESIGNS\Sumitomo RDK-101 base\1K additions

These were built on to the existing RDK-101D models in QSG\Cryostat Designs\UNIFIED DESIGNS\Sumitomo RDK-101 base\Standard (R Heath, I believe)

We got parts machined externally for speed by Metallo in Blantyre. All copper parts are oxy-free then gold plated by the Twickenham group as they were able to offer 1um gold with no nickel adhesion layer.

The cryostat is based on a standard Sumitomo RDK-101D cold head and Zephy air cooled compressor. This is topped with a Chase Cryogenics GL4 to allow sub 1k operation. The entire fridge is mounted in a custom rack on wheels and has the ability to rotate from pointing upwards for sample loading to pointing downwards during cooldown (in order for Chase stage to work correctly). This is accomplished with an aluminium clamp that holds the outer vaccum can and is mounted into two bearings that allow the rotation. A locking pin holds the fridge either pointing up or down.

For signals the fridge has rigid coax cables installed (Cu) from 40kto 1k, anchored thermally at various points. There are also semi-rigid coax from RT to 40k. In the JPL/HW LIDAR configuration a cryogenic bias tee is mounted at 40K stage which is connected to two ends of a differential SNSPD at 1K. The RT electronics include two LNA-1000 amps, a differential amplifier and a Balun to read out the differential single-endedly as normal.

The fridge is controlled by a touchscreen raspberry pi and SIM900 style electronics. A custom python GUI is there to control the cooldown stages. Refer to chase GL4 documentation for explanation of this. The pi also has teamviewer installed and should connect to Eduroam networks (i.e Heriot-Watt) so can be cooled remotely.

NOTES:

Don't lock the cryostat at any point except for straight up and straight down.

The hold time of the 1k Stage is long, >24 hrs.

ISSUES:

Needs a proper locking pin manufactured. Will ask workshop once COVID allows.

Better way to lock clamp into bearing arms. Currently only uses one grub screw and is not that stable. Improve for V2

Possibility to route compressor lines through centre of the rotation axis? In V2