Ionospheric Plasma Irregularity Payload — Rev A
I designed the first electronics revision of a CubeSat payload intended to measure small-scale irregularities in ionospheric plasma. Rev A brings together a low-current probe front end, programmable probe bias, high-resolution digitisation, GNSS time and position, onboard storage, magnetic and thermal context sensing, and an RP2040 payload computer. The design review then identified the physical-probe, analog, guarding, bias-range and interference-control work still required before fabrication.
Summary
- Electronics & PCB Design
- Project
- In development
Evidence on file
My contribution
I drew the Rev A electronics schematic in EasyEDA and reviewed it against what the measurement actually demands. Alongside the drawing I wrote the payload architecture, the measurement chain, the top-level requirements, the subsystem breakdown, the measurement modes, the four levels of mission success, and the verification method for each mission function.
Rev A is a design-stage schematic. Fabrication and hardware testing are not part of this revision. The review at the end is my own assessment of where the drawing is sound and where it is not yet good enough for a picoampere measurement.
Project overview
Why I built this
The mission is to measure small-scale variations in ionospheric electron density and map where they are, how strong they are, and how they relate to geomagnetic conditions — density enhancements and depletions, equatorial plasma bubbles, auroral structures, steep gradients, and storm-related change.
These structures matter because they change the amplitude and phase of radio signals passing through them. Severe scintillation can degrade GNSS accuracy or stop a receiver holding lock, so a map of the irregularities is a map of where navigation becomes unreliable.
Mission objective
The intended payload would hold a probe at a known voltage outside the spacecraft, measure the current the plasma delivers to it, and geolocate every measurement. Current rises and falls with electron density, so a continuous record along the orbit becomes a density profile, and rapid changes in that profile are the irregularities.
Ground processing converts calibrated current into density, identifies enhancements, depletions and steep gradients, and produces geolocated event records for comparison against geomagnetic indices.
That is intended mission behaviour, not a Rev A result. Rev A is a schematic, and the probe that would collect the current is not in it.
What Rev A contains
By subsystem, with designators for anyone following on the drawing.
- U1 RP2040. Controls bias, sampling, timing, storage and operating modes.
- JP1 coax connector with D1 and D2 low-leakage clamps. The probe is not on the sheet.
- U3 ultra-low-bias amplifier with the R1 and R2 high-value feedback network. Device still to be specified.
- U5 and its buffer set the probe potential. Its pinout resembles an I2C DAC — a review item.
- U4 and U6 buffer and filter between the front end and the converter.
- U7, differential against an external reference. Device still to be specified.
- U8, the JP2 antenna connector and TIMEPULSE, for UTC and along-track geolocation.
- U9 LIS3MDL magnetometer and U11 temperature sensor, sited beside the front end.
- U2 SPI flash for firmware and configuration; TF-015 microSD for bulk science data.
- Split analog and digital 3.3 V rails with an L1 ferrite between them, F1 fuse, D4 TVS, D5 reverse protection, USB ESD for the development path, and the H1 spacecraft header.
Rev A schematic
EasyEDA Schematic V1.0, created and updated 16 June 2026, titled PAYLOAD13; Ionospheric Plasma Irregularity Payload. Page 1 of the export, rendered as drawn — nothing redrawn, rearranged or tidied up.
The dates are the drawing's, not the project's. The net labels are small at this size: open the enlarged view to follow the layout, and the PDF to read it properly.
The measurement chain
Plasma current in at one end, a geolocated density profile out at the other. Most of this chain is on the Rev A sheet. What is not on it is the part that touches the plasma.
Source
Ionospheric plasma
Electrons and ions, across a target density range of roughly 1e8 to 1e12 per cubic metre.
Plasma-facing probe
External Langmuir-type probe
Conductive tip, insulating support, guarded connection, and a mount clear of the spacecraft wake. The schematic shows JP1 and nothing of the probe.
- Represented in Rev A
Payload electronics
Low-leakage input protection
D1 and D2 clamp transients away from the front end. Their own leakage can rival the signal.
Electrometer / transimpedance front end
U3 with the R1 and R2 network, converting probe current into a voltage.
Filter and buffer
U4 and U6 buffer the result and limit switching and aliasing noise.
High-resolution ADC
U7 digitises the conditioned signal against its own reference.
RP2040 payload computer
U1 sets the bias, reads the ADC, timestamps samples, detects density changes, manages storage.
GNSS time and position
U8 supplies UTC, position and velocity; TIMEPULSE disciplines the sampling.
microSD storage
Raw current, bias, time, position, temperature and quality flags.
Beyond the payload
Spacecraft computer
Reached through the H1 header, which is on the schematic. The computer is not.
Ground irregularity mapping
Where calibrated current becomes density, and density becomes a catalogued irregularity.
Supporting context sensors
None of these measures the plasma. Each answers whether a plasma measurement can be trusted, or where and when it happened.
- Field direction, so a structure can be checked against the geometry of Earth's field and spacecraft interference characterised. Contextual unless fully calibrated and mounted away from spacecraft fields — both Rev B items.
- Front-end temperature to a half-degree target. Amplifier, resistor, ADC and bias drift all move with it, which is what makes drift correction possible.
- UTC, position and velocity, and with velocity the along-track spacing between samples. Position and timing only: a navigation module generally cannot supply the raw carrier measurements scientific S4 or phase-scintillation analysis needs.
What the design review found
This schematic has most of a working instrument on it. The review asked whether it can measure what it claims to, and the answer turns on picoamps.
The instrument's sensor is missing. JP1 is a connector; the probe is a physical object with a material, an area, a shape, an orientation and a cleaning method, and every one of those changes the measurement. Defining it is instrument design, not packaging to settle later.
Then bias. The present 3.3 V circuitry cannot perform a negative-to-positive Langmuir sweep without an additional supply or level shifter, and the fixed-bias range is not yet specified. U5 is labelled a bias buffer but its pinout resembles an I2C DAC, so its part number and buffer circuit need settling — as do the exact devices for U3 and U7, where input bias, leakage, noise, resolution and sampling rate are the whole argument for choosing one.
The rest is the signal surviving the board it sits on. One fixed transimpedance gain may not cover both quiet and dense plasma without losing resolution or saturating. A high-impedance input needs a driven guard, a clean surface and separation from digital nets. Clamp diodes can leak more than the science signal. Analog and digital returns need deliberate separation. And the magnetometer sits near current-carrying traces.
None of this makes the design wrong. It is the difference between a schematic that is logically correct and one that can carry a picoampere.
Critical Rev B priorities
The review produced seventeen items. These are the ones that decide whether the instrument can make its measurement at all.
- Define material, geometry, area, mounting, orientation, cable and cleaning method.
- Fix the fixed-bias range, and decide whether bipolar sweep capability is required.
- Select and verify the exact ultra-low-bias front-end device.
- Provide selectable transimpedance gain if one setting cannot cover the range.
- Driven guard ring, clean board surface, coating strategy, separation from digital nets.
- Verify the clamps do not leak more than the picoamp signal they protect.
- Select and verify the converter, its reference, its rate and its noise.
- Separate the sensitive analog return path from the digital one.
- Move U9 away from current-carrying traces and magnetic hardware.
- Hardware-safe recovery that disables probe bias on fault.
- Probe input, bias voltage, electrometer output, ADC input, reference, TIMEPULSE, analog and digital supplies.
Measurement modes
Four modes. The first is the mission; the third is optional and needs circuitry Rev A does not have.
- The main irregularity-detection concept, and the highest spatial resolution. Probe held at constant positive bias, current measured continuously at a target of 2 kSPS or faster, with time, position, temperature and magnetic context on every record.
- Triggered by a steep gradient or strong disturbance. Sampling rises toward a 10 kSPS target, full-rate raw data is kept including the interval before the trigger, and the event is marked high priority.
- Sweeping the bias would give a current-voltage curve, and with it electron temperature and spacecraft potential. A full bipolar sweep needs a negative supply or level shifter that is not in this schematic, so it is an intention rather than a Rev A capability.
- Bias set to zero or a protected value, high-rate recording stopped, stored data preserved, temperature and supply still monitored.
Verification plan
Grouped by what each exercises. None has been carried out: there is no hardware, and several need a plasma chamber. Writing the tests beside the requirements is what turned guarding from a note into a priority.
| Area | Planned verification |
|---|---|
| Probe and plasma response | Calibrated plasma chamber at varied density, with wake and placement simulation. |
| Analog chain | Bias under electronic loads, plus transimpedance gain, leakage and noise at every gain setting, against the 50 mV, 1 per cent, 10 pA and 100 pA RMS targets. |
| ADC and sampling | Traceable references across every range, reference and decoupling checked, a long run examined for missing or duplicate samples. |
| GNSS | Timestamps against a disciplined reference toward 1 ms, TIMEPULSE confirmed, position toward 20 m, antenna disconnected mid-run to test the quality flag. |
| Context sensors | Calibration repeated across thermal cycling toward a 5 per cent drift target; magnetometer on a Helmholtz coil, repeated with spacecraft subsystems running. |
| Storage, power and EMI | Card filled and power interrupted toward 99 per cent integrity; every mode against a 3 W allocation; self-generated noise per subsystem and combined. |
| Faults, environment and processing | Forced lockups with recovery inside 60 seconds; vibration, thermal-vacuum, materials, radiation and atomic oxygen with calibration repeated afterwards; blind simulated datasets replayed. |
What I learned
I expected the hard part to be measuring a very small current. It is not, quite. The hard part is stopping the instrument from drowning its own signal.
At picoamp level everything on the board is a candidate source of error. The protection diodes can leak more than the plasma delivers. The high-impedance input picks up whatever the board surface offers it. The RP2040, the microSD card, the USB interface and the GNSS receiver all switch, and their return currents have to be kept off the analog path. A schematic can be entirely correct as a circuit and still fail as an instrument, because being correct is not the same as being quiet.
The other thing I had wrong was treating the probe as mechanical. Its material, area, shape and surface condition appear directly in the relationship between current and density, and its mounting decides whether it samples the ionosphere or the spacecraft's own wake. It is the first component in the measurement chain, and Rev B starts there.
Technical notes
The detailed plasma-instrument requirements, measurement modes, analog design, verification plan, and Rev B review behind this schematic.
Read the technical notesEvidence
Log entries for this project
- The probe is part of the instrument, not just a connectorI drew the electronics for a plasma instrument and left the part that touches the plasma for later, as though it were packaging.InstrumentationDesign Review
- Protection diodes can be noisier than the signal they protectThe clamps guarding the front end sit on the one node in the design where leakage is indistinguishable from science.Analog DesignInstrumentation
Related work
I designed the first electronics revision of a CubeSat receiver intended to detect very-low-frequency electromagnetic signals produced by lightning. Rev A combines a search-coil input, analog amplification and filtering, 12-bit digitisation, event-processing electronics, power and protection, and ground-development interfaces. Reviewing the design made the harder part clear: a sensitive VLF receiver has to distinguish natural signals from aliasing, calibration signals, and electrical noise generated by the spacecraft itself.
I designed the first electronics revision of a CubeSat photometer intended to measure the brightness of a selected star over several hours and look for the small, repeatable dip caused by a known transiting exoplanet. Rev A combines a BPW34 photodiode, a transimpedance amplifier, a 16-bit ADC, a calibration-light interface, onboard storage, power and development electronics. The design review made the harder problem clear: a transit-like dip can also be produced by pointing error, temperature drift, stray light or electronics drift, so those effects have to be measured well enough to rule them out.
I designed the first electronics revision of a CubeSat payload intended to detect gamma-ray bursts and timestamp individual detector pulses precisely enough for comparison with other spacecraft and observatories. Rev A brings together pulse amplification, triggering, pulse-height digitisation, GNSS timing, storage, SiPM bias power, housekeeping, and an STM32 payload computer. Reviewing it exposed the most important remaining gaps: the scintillator and SiPM detector head is not yet present, pulse shaping still has to be added, and the timing path must be calibrated end to end before the timestamps can support scientific comparison.