Lab Experiments for Space Interferometry

Lab Experiments for Space Interferometry

AEI is the largest laboratory worldwide for LISA interferometry. We cover the widest range of technologies while keeping close contact to all other participating groups.

Main work in this group concerns the development and testing of essential components for space interferometry.


We built an experiment that allows to routinely measure a rarely used but important property of multi-stage lasers, namely the relative phase stability between carrier and sidebands (dispersion) in the amplifier stage. We could show that several combinations of seed laser and fiber amplifier fulfill the requirements for the clock noise transfer. The remaining question for the LISA laser is mainly the verification of space compatibility, reliability and lifetime, which is an industrial task.


This is considered a critical component, and our in house development is well advanced. The noise performance with electrical signals is compliant, and connected to optical signals. We collaborate with APC Paris to independently test the phasemeter.

Photodiode preamplifiers

Equally important for the LISA noise performance are the photodiode preamplifiers, for which we have developed a circuit with discrete transistors which outperforms the op-amp designs used elsewhere. A quadrant version is under development.

Clock-noise transfer

All components for the clock-noise transfer via GHz sidebands have been tested and suitable components have been identified, including RF cables, electro-optical modulators and frequency multipliers/dividers.


An important ancillary function of the laser link is the measurement of absolute spacecraft distances to sub-meter accuracy (ranging) and the transfer of data (roughly 10 kbits/s). In an optical experiment we have implemented a weak phase modulation of the carrier with pseudo-random codes (PRN spread-spectrum). We have shown that the required performance can indeed be achieved both for ranging and data rate. A non-standard feature of this scheme is that one data bit is shorter than one PRN code. Since the usual shift-register codes of length 2n-1 are awkward to subdivide into data bits and moreover have a non-zero average, we developed codes of length 2n by numerical optimization.

Point-ahead angle actuator

Another critical component for LISA is the point-ahead angle actuator mechanism (PAAM) which must be able to move ±0.4 mrad while not disturbing the optical pathlength by more than 1 pm/√Hz. The latter requirement is particularly difficult to verify, and we have built a test setup to do so on behalf of ESA. The PAAM in question is part of a three-mirror cavity, the resonance frequency of which is compared to that of a known stable cavity. We were able to show that at least one, possibly two, industrial actuators fulfill the requirements.

Reciprocity of the back-link fiber

A long-standing open question in the LISA design with huge consequences for the spacecraft design is the so-called reciprocity of the back-link fiber, i.e. the equality to pm level of the two optical pathlengths through a single-mode fiber when used in both directions. We achieved a noise level of 5 pm/√Hz, which now allows to go ahead with the planned LISA design. Many tricks and stabilization loops were required to achieve this performance, which may have some consequences for the LISA baseline design.

Data transfer

First, the data from the three spacecraft with their independent free-running clocks must be synchronized to nanosecond accuracy based on the ranging timestamps, and the interspacecraft distances must be extracted simultaneously. We have begun to develop optimal filters for this purpose, which are similar to Kalman filters but non-standard because the real time of a raw measurement is not known a priori but instead contaminated by the clock offsets which are in turn part of the unknown state vector. Next in the processing chain is the well-established technique of time-delay interferometry (TDI). All existing publications are based on an outdated LISA layout and are oversimplified in some aspects like the absolute ordering of frequencies. We have begun to rewrite the equations for more realistic scenarios and will perform numerical simulations for verification.

LISA breadboarding

The demonstration of offset phase locking with sub-nW power levels and a study of an alternative LTP-like frequency stabilization scheme as fallback option for LISA is also part of our research. The largest project here is the testing of a complete breadboard of the LISA optical bench, for which we have a contract with ESA as part of a team with Astrium, Glasgow and TNO. Our role is to build the test equipment and perform the tests.

Go to Editor View