2.1

Pump fiber receptacle

The SMA pump fiber connector used for early prototype models is replaced by a custom made connector based on Diamond’s MINI-AVIM fiber connector design (brown in Figure 2). This custom made MINI-AVIM pump fiber connector includes the pump optic collimation lens.

2.2

Oscillator

The oscillator sub-assembly contains all optical components beginning with the focus lens up to the output coupler of the resonator and further parts and components needed for temperature stabilization and mounting. The focus lens is mounted identically to the collimation lens, but is kept under ambient pressure, since it is within the sealed part of the housing. The following resonator consists of three sub-assemblies. The first sub-assembly contains the laser crystal and the passive q-switch mounted together in spring loaded slotted grips. Basically this approach was used since the prototype model, improved only by clamping the optical components via their barrel surfaces thus avoiding a mechanical contact to the optical surfaces.

The second sub-assembly within the oscillator is the Brewster window, which is axially clamped, with its mounting mechanics. This sub-assembly needs to be aligned rotationally in respect to the first oscillator sub-assembly to optimize losses within the resonator.

Both previous sub-assemblies are located within an additional tube. This is mounted axially spring loaded and fixed by the third sub-assembly, the output coupler mount, which contains the resonator output coupler and is adjustable with three rotational degrees of freedom prior final fixation. All three axes need to be alignable to optimize resonator performance.

2.3

Frequency conversion stage

Within the frequency conversion stage, two non-linear crystals – KTP and BBO – are mounted for frequency quadrupling in the same way as the output coupler. Both crystals can be aligned rotationally in three axes – as the output coupler – to achieve optimum phase matching and conversion efficiency of these crystals.

Due to risk mitigation reasons and simplification of coatings used on following optics, a beam cleaner mirror together with two optical absorbers are placed between both non-linear crystals to filter out unconverted IR light. Also the very small portion of UV light that is back reflected from the second conversion crystal is dumped into an absorber by this mirror. Main design driver here is the available space for the three ∅4 mm optical elements. They are integrated in an enveloping cylinder with ∅13 mm diameter and a length of 5 mm, limited by the thermal hardware which surrounds the optical elements of the frequency conversion stage. The BBO crystal is integrated as close as possible to the oscillator and KTP to achieve high conversion efficiency.

2.4

Monitoring stage

The monitoring stage has multiple purposes. Its main function is to measure the laser pulse energy with two integrated photodiodes. Portions of both the out-coupled 266 nm laser beam and the internally dumped 532 nm laser beam are measured with separated photodiodes. The monitoring stage acts also as a beam dump for the frequency doubled 532 nm laser light. Beam shaping lenses are integrated in the monitoring stage to increase the beam size on the coated exit window and to focus the UV laser beam onto the sample beneath the mass spectrometer ion inlet within the ultra-clean zone where Martian soil samples are presented to the MOMA instrument.

The separation of the frequency doubled and frequency quadrupled wavelengths is realized by two mirrors, highly reflective for 266 nm wavelength and highly transmissive for 532 nm wavelength. Both mirrors are mounted individually within a mount with 2 degrees of freedom for each mirror (one lateral, one rotational). With these alignment possibilities the beam can be alignment towards the nominal center of the second integrated lens. This lens has two lateral DoFs to pre-align the out-coupled laser beam towards the design target position.

2.5

Sealed housing

The design of the sealed housing combines vacuum brazing of sapphire windows to titanium frames and laser welding of the titanium structural elements. Only due to welding process needs, a metal coated sealing ring was needed for the last welding step of the final laser, to keep particles produced by welding out of the inside of the laser head, and to reduce a possible impact from process gases to the dew point within the laser head. All brazed and welded joints were X-ray scanned and rated in means of inclusions (bubbles) and grooving. Additionally, both housing parts were leak tested individually and had a sum leak rate below 10^-9 mbar·l/s.

The finally assembled laser head encloses three volumes (compare Figure 3), with individual requirements regarding leak rate and time of final closeout (see Table 1).

Figure 3:

Cross-section through the MOMA laser head, highlighting the enclosed gas volumes; in green the receptacle volume; in magenta the main LH volume and in turquois the beam deflection unit volume

Table 1:

Enclosed gas volumes within the MOMA laser head

*1)

for 4 of 6 seals (below detection limit for other 2, tested in sniffing mode)

*2)

UCZ window

2.6

Prism

Most challenging in the design was the need to have an optical element that can align the outgoing beam with respect to the target beneath the ion inlet of the mass spectrometer and have this optics in a hermetically sealed volume in the end. The final design uses springs for clamping a prism, and also uses this spring load against three screws that can be accessed from outside for prism aligning through small bore holes, which can be closed after alignment. The prism itself is an uncoated fused silica substrate, cut in a way, that both optical surfaces are used close to Brewster angle, and are deflecting the beam by 44° towards the target.

2.7

UCZ window

The so called UCZ window closes the ultra-clean zone hermetically towards the laser. It is designed as an uncoated fused silica substrate mounted under Brewster angle in respect to the laser beam. The UCZ window integration is realized by gluing it with very low outgassing glue in a frame structure made of Invar 32.5.

2.8

Thermal

The MOMA LH temperature stabilization concept is realized as heat-only temperature stabilization. A heat-and-cool stabilization concept would result in a significantly more complex design. In early prototype models only one heater was integrated inside the LH within the frequency conversion stage. With the highly temperature sensitive non-linear crystals, it is mandatory to keep the temperature stable. During the prototype model phase, it turned out that the influence of the FCS temperature stabilization cannot be neglected for the stable operation of the oscillator. The influence is quite slow depends on environmental temperature. To improve the unwanted coupling between both subassemblies, later models are realized with two additional independent heaters on the oscillator to reach the set temperature fast and to reduce the dependency from the frequency conversion stage and environmental conditions [3].

For the FM LH, each of the three heaters consists of a custom-made Kapton foil heater (manufacturer: Minco), and a AD590 temperature transducer from Intersil, both space-qualified. These are powered and controlled by the LPU. The heater foils are mechanically clamped to the barrel surface of the oscillator and the frequency conversion sub-mounts; the temperature transducers are glued close to the corresponding heater.

Thermal insulations are implemented between frequency conversion stage and oscillator and within the sealed housing and mounting towards the mass spectrometer. Between oscillator and frequency conversion a thin-walled titanium cone was implemented, in order to enable the independent control of these items. The mounting, thermally low-conductive spacers are used on one side (left mounting I/F in Figure 4), which are also used to shim height deviations between LH and MS. Towards the right I/F, the titanium housing itself acts as thermal insulation (compare Figure 4).

Figure 4:

Thermal concept of the laser head with the three heaters (H1 to H3) clamped to the oscillator and the frequency conversion stage (FCS). The laser crystal (LC) is shown as additional heat source and the monitoring stage (MS)

2.9

Metrology

The mass requirement given for the MOMA laser head was not to exceed 270 g including 20% margin for the whole laser head, which is without margin 225 g. From design the predicted laser head mass were 221 g, and the measured value during metrology testing was 217 g net weight. In dimension, the LH does not exceed 44.5 mm by 56.2 mm by 218.8 mm including a MPS-provided bend-protection for the optical fiber. Metrology testing on the flight model showed that all interfaces are within required tolerances for integration (compare Figure 5).

Figure 5:

3D view on MOMA LH with metrology values shown

3.1

Functional tests

Since all other wavelengths are separated from the 266 nm radiation and dumped inside the LH, the UV output energy is the main characteristic of the performance of the oscillator in combination with the frequency conversion stage.

The detection of the pulse release time (PRT) as an intrinsic property of the oscillator helps to relate any changes of the UV output energy to a misalignment of the oscillator (leads also to a change of the PRT) or the frequency conversion stage components (PRT remains unchanged).

For the pointing measurement a bore in the reference block in front of the fluorescence target with a known distance of its center point to the target position was taken as a reference especially during the TVAC test to compensate for any movement of the reference block relative to the camera. The laser itself was rigidly mounted onto the reference block (as in the final FM configuration). The center point of this bore was visually determined via the adjustment of a circular aperture and compared to the position of the beam centroid (Gaussian beam profile fitted to the real shape). The absolute beam position with respect to the target value was not important during this test campaign because it can be aligned when the laser is finally mounted on the MS.

3.2

Vibration test

For vibrational testing, the laser head was mounted on a structural test unit (STU) of the MS instrument provided by GSFC (see Figure 7 (right)). This was done because the load requirement (shown in Figure 7 (left)) was only given to the instrument and not to its sub-components.

Figure 7:

(left) Vibration load requirements to the MOMA Instrument at its mounting interface towards the ultra-clean-zone (UCZ) for in plane and out of plane acceleration; (right) Flight model of MOMA Laser head mounted on MS STU for vibration test

Notching was implemented to the control (compare Figure 7 (left)) and response limiting to the Wide Range Pump (WRP) mounting interface and the top of the WRP sensors in order to limit WRP responses to known allowable values.

Response channels along with a corresponding max g²/Hz response (in the form of a “limit”) were to be patched into to the control loop so that the controller was able to adjust the input to prevent testing beyond WRP maximum response limits.

All three axes were tested, with resonance search before and after; loads were applied with reduced level and increased in steps (-12 dB; -6 dB; -3 dB and 0 dB). Besides other sensors, the most relevant three axes sensor was mounted on top of the center part of the LH.

3.3

Shock test

The shock test was done on a shock test table at Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR) Bremen.

The shock profile was calibrated with the STU including LH mass mock-up. The calibrated shock level was then applied to the X-axis of the MS STU with the FM laser head. The nominal spectrum to which the STU with the FM LH was exposed is shown in Figure 8.

Figure 8:

Input load; acceleration plotted versus frequency

3.4

Thermal-vacuum test

Thermal vacuum testing of MOMA LH was carried out at MPS under Martian atmosphere. For this, the laser head was mounted onto its reference block as it was for all previous functional tests. Main difference here was that the beam deflection volume was purged and needed to be hermetically sealed, due to the fact that condensation at low temperature had to be avoided to prevent damage to optics when firing the laser head.

In total five thermal cycles were performed with the flight model of the MOMA laser, which includes besides the laser head also the laser pump module built by MPS. The first and last cycle were extended to non-op limits, while the three intermediate cycles were to operational temperatures (compare Figure 9).

In addition to the functional characterization of the FM-LH before and after the thermal-vacuum (TV) test the relevant laser parameters were also checked at the minimum (-45°C) and maximum (+25°C) operational temperatures as well as at two temperatures on the rising edge of an operational temperature ramp (-20°C and 0°C). The behavior at low and high bus voltage (26V and 29.4V instead of 28V as nominal value) was also determined within a full operational cycle. Besides internally logged parameters, either pulse energy (marked as “E”) or pointing (marked as “P”) was also measured outside of the thermal-vacuum chamber (compare Figure 9).

Figure 9:

Temperature profile used during thermal vacuum testing the MOMA laser head; five cycles were applied with two cycles up to non-operation temperature. In red boxed “E” states externally measured energy and green boxed “P” pointing measurements

3.5

Environmental tests results

The UV output energy and the energy tuning behavior did not change significantly change after transport, vibrational and shock load. Only during the TVAC test a slight irreversible drift of the complete energy tuning curve towards higher FCS temperatures was observed.

Besides expected effects on the pointing (e.g. from dn/dT refractive index change within the prism), two further effects were discovered. Firstly the beam position on the target in Y direction changed slightly over almost all tests; this could be identified to be due to a mechanical effect of the prism mount, which could later be avoided by a change of the handling procedure. Secondly, the measurement data indicated that the volume of the beam deflection unit was not leak tight. However this could be solved by a change in the handling procedure. Both effects did not occur again in the thermal-vacuum test campaign for the complete MOMA instrument.