|
|
[<< Home](/home#3-front-end-design-panel-charge-3)
|
|
|
|
|
|
[<< Section 3.1](/3-front-end-design/3.1)
|
|
|
|
|
|
## 3.2 Mechanical Design of the ALPACA Cryostat [^a]
|
|
|
| **System Characteristic** | **Specification** |
|
|
|
|:--------------------------------------|:--------------------------------------------------------------|
|
|
|
| Frequency Coverage | 1300-1720 MHz, (420 MHz Total Bandwidth) |
|
|
|
| Number of Dual-pol Elements | 69 |
|
|
|
| Array Geometry & Element Spacing | Hexagonal in 3 sections of 23, 135mm spacing |
|
|
|
| Dul-pol ELement Type | Dipole with Pie shaped Outer Arms |
|
|
|
| Tilt (α) and Pie shape Angle (β) | 20° from horizontal & 40° |
|
|
|
| LNA Bias Power & Gain | 18 mW & >35 dB across band |
|
|
|
| First Cryogenic Stage Temperature | <= 99K |
|
|
|
| Second Cryogenic Stage Temperature | 20K |
|
|
|
| Number & Type of Cold head | 3x & CTI 1020 dual-stage |
|
|
|
| Number & Type of Compressor | 1x & Trillium M700 (+1 spare budget permitting) |
|
|
|
| T<sub>sys</sub>, Spec (Goal) | 35K (27K goal) |
|
|
|
| Number of Dipole assemblies | 69 (+11 spares) with accompanying cryo-LNAs and 20K interface parts |
|
|
|
| Internal Wiring | Semi-rigid coax (at 20K) and multi-channel Flexible stripline |
|
|
|
| Cryostat Rotation | using GBT Sterling Mount [^1] |
|
|
|
| | |
|
|
|
| **Monitor and Control** | |
|
|
|
| Number & Type of Temperature Sensors | 18 (6 per section); Lakeshore Diode sensors (DT-670) |
|
|
|
| Vacuum Gauge | 1x Full Range Gauge - 10<sup>3</sup> hPa to 5x10<sup>-9</sup> hPa (Pirani/Cold Cathode) |
|
|
|
| LNA Bias | 18 M&C Cards (8-channels/card) |
|
|
|
| Compressor | Built-in monitor for Supply and Return line pressure and Reservoir temperature |
|
|
|
|
|
|
|
|
|
The cryogenic vacuum vessel forms the frontend portion of the ALPACA system, and is commonly referred to as the receiver, as this is the unit capturing the incoming radio signal. The shape of the vacuum vessel is a simple cylinder. The cylinder height (~0.6 m) is determined by the internal cryogenic components and structures. Unique aspects of this receiver design include a large radio frequency (RF) transparent vacuum window, modular antenna/LNA units with cyro-clamp receptacles, single-piece flexure-based compliant thermal links between the cryo-cooler cold heads and 1st and 2nd stage base plates, and use of both blade flexure and sliding column thermally insulating standoffs.
|
|
|
|
|
|
|
|
|
<div align="center">
|
|
|
<img src="../uploads/8e23d4656c337a0237cb5903040781ad/dipoles_in_cryostat.jpg" width="800">
|
|
|
</div>
|
|
|
<kbd>
|
|
|
Figure 1: Looking at the ALPACA cryostat with the vacuum window and one section of the RF clear foam removed. We can see the dipole array along with the various components of the cryostat at the first cryogenic stage -- The ground plane (in blue), radiation shield (red) and base plate (green). The 20 K radiation shield is hidden in this view to expose the cold finger assembly into which the dipoles plug in. Nylon rings that have a higher coefficient of thermal expansion than Aluminum are used to clamp the dipole assemblies at low temperature. The 20K base plate (grey) and the flexible striplines (light green) carry the RF signal from the array elements to outside the cryostat.
|
|
|
</kbd>
|
|
|
|
|
|
<br>
|
|
|
|
|
|
The vacuum vessel can be divided into two main parts: the solid top half, and the RF transparent window. The top half consists of a light-weighted 6061-T6 aluminum top plate and a welded aluminum shell. The top plate has all the feedthroughs for vacuum, cryogenic, metrology, and signal readout components, and provides the interface for mounting instrumentation.
|
|
|
|
|
|
<div align="center">
|
|
|
<img src="../uploads/6e8543536e23ec103020d95ea9fd031a/cryostat_cross-section.jpg" width="700"></div>
|
|
|
|
|
|
<kbd>Figure 2: Section view of the ALPACA cryostat along a row of antennas. The radio frequency transparent window consists of HDPE (blue) over 4 layers of Rohacell 71 HF foam (pink). The 20 K antennas pass through the 1st stage (red highlight) and into the foam with no thermal contact. The 2nd stage is completely shielded from room temperature heat loads. The vacuum vessel solid top half is welded aluminum.
|
|
|
</kbd>
|
|
|
|
|
|
The functional requirements for the window are that it must be vacuum tight, RF transparent, structurally sound, and transmit a manageable heat load to the internal cryogenic stages. Our solution is to functionally separate the vacuum from the structural and thermal requirements. As demonstrated with our PAF prototype[^2], a durable vacuum tight structure is achieved with a welded HDPE top hat that is only ~3 mm thick. An RF transparent foam, Rohacell 71 HF, fills the vacuum side of the HDPE, supporting it from within and keeping it from collapsing. The foam transfers the atmospheric load to the 1st cold stage, which in turn transmits it to the top plate of the vacuum vessel where it is balanced by the atmospheric load on the other side. This design also solves the black body heat load problem because the foam is opaque in the infrared. This changes the window heat transfer method from radiation to conduction, and since the thermal conduction of the foam is rather poor (0.025 W/m/K), the heat load on the 1st stage is limited to a manageable 50 W.
|
|
|
|
|
|
Since it is more efficient to intercept environmental heat loads at higher temperatures, an intermediate first cold stage “surrounds” the 20 K second cold stage, sinking room temperature thermal radiation before it reaches the 2nd
|
|
|
stage, and providing a temperature guard for wiring and structural components. This two stage configuration is readily aligned to commercial two-stage Gifford-McMahon cryocoolers. In our case, three CTI-1020 cold heads from Trillium are sufficient to handle the load. However, the distance between the antennas and LNAs must be small to minimize signal loss and added noise. This results in a unique design where the antennas at 20 K pass through the ground plane at the first cryogenic stage temperature with no thermal contact (see Figure 2 below).
|
|
|
|
|
|
The cryostat was originally designed for use on the 305 m Arecibo telescope. After extensive computer modeling by German Cortes (Cornell) and Karl Warnick (BYU) an array of 69 dual polarization dipoles with pie shaped arms in a hexagonal configuration was settled on. The dipoles are spaced 135 mm (0.68λ at 20 cm wavelength). Thermal modeling showed that three CTI-1020 cold heads driven by a single Trillium M700 compressor would provide adequate cooling capacity to maintain the dipole array at 20K and provide an intermediate guard stage in the cryostat. The 69 element array has a three-fold symmetry allowing all the internals to be manufactured as three mechanically independent sections. This is required as three cold heads attached to a single, large metal plate would develop stresses due to thermal contraction.
|
|
|
|
|
|
<div align="center">
|
|
|
<img src="../uploads/54950ac23532229f6e982fc61416fc51/view_from_top.jpg" width="700">
|
|
|
</div>
|
|
|
<kbd>Figure 3: ALPACA Cryostat, showing the turbo-pump assembly for vacuum and the three CTI-1020 cold heads that provide the cooling to maintain the operating conditions for the dipole array and LNAs. The room temperature base plate is shown in red. The cold heads are installed on risers to align the cold head interface with the internal stages. </kbd>
|
|
|
|
|
|
### Easily Replaceable Dipole Assemblies
|
|
|
|
|
|
A major feature of the design is the ability to easily remove a dipole/Low Noise Amplifier (LNA) assembly in case of failure of one of the 138 LNAs without disassembling the entire cryostat. The major advantage of the system was that, when installed on the rotary floor of the Arecibo telescope’s Gregorian dome, after warming up, an LNA could be replaced by removing the HDPE top hat and foam layers, pulling out the relevant dipole assembly, replacing it, the foam layers and the top hat, all without removing the cryostat. Given the cool-down / warm-up time of ~2 days (there are warm up resisters in the cryostat) it is not clear that replacing failed LNAs at the prime focus of the GBT is practical. However, the ability to so easily replace dipole assemblies will still avoid the need to completely open the cryostat.
|
|
|
|
|
|
[Section 3.3 >>](/3-front-end-design/3.3)
|
|
|
|
|
|
[^a]: Please refer to [Parshley et al. 2020](../uploads/7c6759a47aba883325e3d4aabe9049e8/ASME-PVP2020-21818-Parshley-vFinal.pdf) for more details of the cryo-mechanical design of the ALPACA Cryostat
|
|
|
[^1]: Limits defined as ±204° [GBT Control System Manual, Pg 12](https://safe.nrao.edu/wiki/pub/GB/Operate/Training/GBTControlSystem.pdf)
|
|
|
[^2]: Please see [Cortes-Medellin et al. 2015](https://doi-org/10.1109/TAP.2015.2415527) for more details. |