The New ING/RGO Infra Red Test Camera

The Perpetrators - Derek Ives (DJI), Simon Tulloch (SMT)
and Ian Baker (ibaker)

1st image taken with the silicon Multiplexer at 150K

Introduction


Figure 1 : The camera on its test stand.

The purpose of the test camera is to gain experience in operating the HAWAII device in advance of the manufacture of the science camera. It will provide a test bed for the SDSU controller software and allow us to electrically, thermally and to a certain extent optically characterise the detector chip. It will also serve as an introduction to closed cycle cooler technology and cryogenic mechanisms.

The test camera consists of a modified liquid nitrogen CCD cryostat. The nitrogen can has been removed and replaced with a CTI model 350 closed cycle cooler. Coupled to this cooler are the detector itself and a 5 position filter wheel. The filter wheel was driven manually via a rotary feedthrough in the cryostat wall. The temperature of the detector can be servoed at any temperature above 65K. The filter wheel contains a blanked off position for dark current evaluation, 950nm and 2 micron bandpass filters for QE measurement, and a single element lens for crude imaging in the laboratory. External to the cryostat is mounted a stable IR light source that can be used for flat field illumination and QE determination.

An IR labs detector PCB and external preamplifier are incorporated into the camera.


Figure 2 : Camera electronics including SDSU controller.

Closed Cycle Cooler

The cooler is conveniently sized in that it can be mounted in one of our Oxford Instruments LN2 cryostats once the LN2 can has been removed. The end plate of the cryostat needed to be re-machined to mate up against the vacuum flange on the cooler head. An L-shaped polished aluminium plate was screwed to the end of the cold finger, an intervening disc of Indium foil ensuring good thermal contact. This plate was instrumented with a diode temperature sensor, a 100 Ohm heater resistor and a charcoal sorb with integral heater. Electrical connections were made via a 15 way micro-D connector. Two copper springs extending from this plate make thermal contact with the detector mount and the filter wheel.


Figure 3 : Camera cold finger with charcoal getter attached.

To check operation of the cooler, the cold finger was wrapped in foil to exclude radiation and the cool down curve measured using the diode sensor. Once the cooler had bottomed out, the heater power was steadily incremented and the corresponding cold finger temperature at each level noted. A power-temperature graph was then obtained. At 77K the cooler is capable of pumping 16 Watts.


Figure 4 : Cold head cool down rate.


Figure 5 : Closed Cycle Cooler Performance Graph.

Temperature measurements were done using a UG8AT power diode that had been calibrated against a PT100 sensor. The diode came in a TO220 package that was easily mounted to the camera with an M3 screw. Sensing was done using by connecting the diodes to a 50 micro-amp constant current source and measuring the bandgap using an op-amp circuit. The circuit used had a top-panel switch allowing up to four temperature channels to be measured. This was sufficient for the test camera which had a sensor on the cold finger, the detector mount and the filter wheel.


Figure 6: Diode sensor calibration graph.

Cryostat Body Details

All of the inner surfaces of the cryostat vacuum vessel were given a mirror polish to reduce emmissivity. This allowed us to dispense with aluminised mylar and thus reduce outgassing problems. The filter wheel/detector assembly was mounted, using three Delrin spacers, from a ring welded to the inside of the cryostat walls. Delrin was chosen for its good vacuum properties and very low thermal conductivity. The vacuum vessel contained 6 side ports, all but one of which were used. The functions were : pump port, rotary feedthrough, sorb heater/calibration photodiode connector, temperature servo connector (two heaters, 3 diode sensors) and 26 way detector connector. The cryostat lid, which was also polished on its inner face had a central 60mm diameter Spectrosil window. Although not an ideal material given its slight transmission dip between 2 and 2.5 microns , it was adequate for the purposes of the test camera. The pump port contained a T-piece that mounted a vacuum gauge and an MDC vacuum valve. The gauge verified that when cold, the pressure inside the cryostat was approximately 1e-6 mBar, well below the level at which gaseous thermal conduction is significant.


Figure 7: View into top of the open cryostat.
Filter wheel/ detector assembly removed
to show top of cold finger.


Figure 8: Pressure drop in cryostat as
the camera is cooled.

Detector/Filter Wheel Assembly

This comprised the following elements : IR Labs detector socket/PCB, two part cooling block that encloses and cools the detector, a cylindrical box that encloses and cools the filter wheel, a bevel gearbox and spur gear to drive the filter wheel and the filter wheel itself. Once bolted together, this assembly could then be mounted inside the cryostat. Figure 9 shows the assembly. The two copper springs are used to cool the detector and filters. The filter wheel box is lower most, the detector cooling block uppermost. The temperature sensor and heater resistor are visible on this block. The back of the IR labs PCB is clearly visible stood off from the back of the filter wheel box.


Figure 9: Detector/Filter Wheel Assembly.

Figure 10 shows the assembly from a different angle. One half of the rotary coupling is more clearly visible. The right angle gearbox on the back side of the filter box can also be seen as well as the pillars which support the PCB.


Figure 10: Detector/Filter Wheel Assembly.

Figure 11 shows this assembly mounted in the open cryostat. The rotary feedthrough can be seen protruding in the six 'o' clock position. One turn of this rotates the filter wheel by one position via a white Delrin spur gear. A spring loaded detect mechanism ensures that the filter wheel latches at the correct place. Two bandpass filters are shown loaded, one with a centre wavelength of 950nm, the other 2 microns. The third filter location is kept spare. One further position is used for a single element lens. The motion of the filter wheel inside its box is lubricated by a PTFE washer under the lid of the box.


Figure 11: Detector/Filter Wheel assembly
shown mounted in cryostat.

The fully assembled camera has a polished aluminium radiation shield fitted over the filter wheel box. This is best shown in figure 12, here the camera innards are shown minus the cryostat vessel.


Figure 12: Camera Innards.

Calibrated Light Sources For Q.E. Measurement.

Measuring the QE of HAWAII is fairly straightforward at shorter wavelengths since it has a sensitivity overlap with Silicon detectors in the Z band. We chose to make a QE measurement at 950nm since we had a stable narrow band light source at this wavelength. Prior to mounting HAWAII we placed a Silicon photodiode of known QE at the focal plane of the camera and used it to calibrate the light source. When measuring the source brightness the filter wheel was cooled to operating temperature and the 950nm bandpass filter selected.


Figure 13: Calibration photodiode mounted into
the detector cooling block.


Figure 14: The 950nm light source attached
to the front of the camera cryostat.

The light source consisted of a temperature compensated IR LED that was driven from a constant current circuit. A thermistor mounted close to the LED was used to vary the current thus overcoming the high temperature coefficient of the LED. RGO has a number of similar monochromatic sources spanning the range 380 to 950nm.


Figure 15: Calibration Light Source.

QE measurement at 2 microns is a possibility also since the filter wheel contains a narrow band 2 micron filter. We would first need to obtain a calibrated IR sensitive InGaAs photodiode. It is not clear whether such a measurement is worthwhile given the considerable expense of such a device.


I am... apo