Retrospective note
Written from project notes after the fact.
The spacecraft could become part of the measurement
The VLF receiver had to coexist with switching and digital electronics that could radiate or couple energy into the same band I wanted to measure. That made spacecraft-generated interference part of the instrument design rather than an integration detail.
Problem
The measurement is a 10 microvolt signal at the coil connector. A CubeSat contains a processor, a 2.4 GHz radio, an OLED, LED drivers, a switching regulator and a power harness, all of them switching at rates whose harmonics land inside 3 to 30 kHz. High-current wiring near the coil couples magnetically into exactly the quantity the coil exists to measure. Once a clock harmonic has been through the same amplifier and converter as a lightning impulse, the receiver cannot tell them apart.
Decision
I made interference a science requirement rather than an EMC checkbox: spacecraft-generated interference has to stay below the payload's detection threshold during science observations. The responses that follow are operational and physical — Wi-Fi, the OLED, the LEDs, switching circuits and unnecessary radios disabled during sensitive measurements rather than trusted to be quiet, repeated spacecraft noise frequencies identified so they can be recognised in the data later, and the coil to be mounted away from high-current wiring.
Test / evidence
The characterisation plan is what makes attribution possible. Record a baseline with all subsystems off. Turn on each subsystem individually and record the change in the VLF spectrum. Build a spacecraft interference database. Then repeat the whole exercise in the fully assembled satellite, because subsystems that are individually quiet can be noisy together.
Result
Most of the noise this instrument would have to reject would come from a few centimetres away, from the vehicle carrying it. That is an unusual position for an instrument to be in, and it is not something that could be fixed at integration.
Next step
Measure it. None of this has been carried out — there is no hardware, and the noise floor is a target.