Skip to main content

Module sillygoose

Module sillygoose 

Source
Expand description

AerospaceNU’s SillyGoose flight computer’s logs, as the board writes them: text under a timestampMs header, one row per sample.

The SillyGoose is a flight computer of Northeastern University’s rocketry club, AerospaceNU, that logs its barometer, its accelerometer and gyroscope, its own altitude and vertical-speed estimates and its pyro channels’ state. The club publishes its flights’ logs with a README that names each file after its flight and tags it, among others, pre reboot and post reboot where the board restarted in flight and its log is in two files, and doctored where the file was altered after it was offloaded (AerospaceNU, rocket-flight-data, README.md, read 2026-10-12). No published document describes the columns, so each unit here comes from the column’s name or from a check the reader runs on the file’s values, as below. Three layouts are among the project’s sample logs, all read alike:

  • The board’s log: tab-separated, 20 columns, about 100 samples a second in flight and one every 5 s on the pad (here with the tabs shown as spaces):

    timestampMs  pressurePa  tempK  accelX  accelY  accelZ  gyroX  gyroY  gyroZ  imuTemp  …
    1000  97716.500000  290.15  0.012000  -0.020000  9.810000  0.001064  …
  • With its settings: a CONFIG line of KEY=VALUE settings, then the same header with 8 more columns, tiltMagnitudeDeg, angularVelRadS_x … _z and quaternion_a … _d. The one such sample file looks like replayed output, not a recording: after a few samples on the pad it jumps into the boost, it ends still climbing, and its temperature holds one value throughout (a note says so; below).

  • Space-separated: the 20 columns’ header with single spaces between the names, at 50 samples a second, whose values don’t all agree with the units the names state; the checks below find which do.

Recognition. The first line that isn’t blank, or the one after it if that first line opens with CONFIG, is a header whose first name is timestampMs and which names pressurePa and altitudeM, compared without regard to case. A header with a tab has its names and its rows’ cells split at tabs; one without, at runs of spaces. The club’s ground station’s conversion of a downlink, a CONFIG line and then comma-separated rows with no header, is not this layout and isn’t read.

Columns. Each name the reader knows, what it is read as, and where its unit comes from; the column keeps the file’s name. A column the reader doesn’t know is kept as written, Quantity::Unrecognized, with a note (the first 20 such notes kept, the rest counted); a header naming a column twice, with an empty name, or with more than MAX_COLUMNS names is refused:

ColumnQuantityUnit, and from where
timestampMsthe clockmilliseconds, from its name; the board’s own zero, before its first sample
pressurePaQuantity::PressurePa, from its name, if its first value is a pressure near the ground in pascals
tempKQuantity::Temperature: FlightLog::temperature_kK, from its name, if every value is one an electronics bay can have in kelvin
accelX, accelY, accelZQuantity::Accelerationm/s² or standard gravities, from the pad’s gravity (below)
gyroX, gyroY, gyroZQuantity::AngularRateradians or degrees per second, from the board’s turning (below)
imuTempQuantity::TemperatureK, from the values (below)
battVQuantity::Voltage: FlightLog::battery_vV, from its name, if every value lies within 0 to 20 V
altitudeMQuantity::AltitudeAboveGround: the primary trackm, from its name, checked against the pressure (below)
velocityMSQuantity::Speed, up positive: FlightLog::vertical_speed_m_sm/s, from its name, checked against the altitude (below)
accelerationMSSQuantity::Accelerationm/s², from its name: the board’s own estimate, along an axis the file doesn’t state
unfiltAltQuantity::AltitudeAboveGroundm, from the pressure (below); the zero is the board’s, and not stated
flightStateQuantity::Numberthe board’s state code, as written: no published source gives the codes’ meanings
drogueCont, mainContQuantity::Number0 or 1, as written: the pyro channels’ continuity, by their names
drogueFired, mainFiredQuantity::Number0 or 1, as written; each rise from 0 to 1 is an event (below)
tiltMagnitudeDegQuantity::Angledegrees, from its name, if every value lies within 0 to 180°
angularVelRadS_x, _y, _zQuantity::AngularRateradians per second, from their names, if no value is beyond what a gyroscope measures (below)
quaternion_a … _dQuantity::Numberthe parts of an attitude quaternion, which have no unit

A column whose values refute the unit its name states is kept as written, Quantity::Unrecognized, with a note saying why; a column whose unit the values decide gets a note saying on what evidence. The checks:

  • The pressure is read in pascals if its first value lies within 50,000 to 110,000 Pa (the bounds crate::rrc3 argues for a pad’s pressure). Every sample file’s does.
  • The temperatures. An electronics bay’s temperature lies within −40 to 60 °C, 233.15 to 333.15 K. tempK is read in kelvin, as its name says, if every value lies within that range; imuTemp’s name states no unit, and it is read in kelvin on the same evidence: in degrees Celsius or Fahrenheit the same values would be hotter than any bay survives. Otherwise each is kept as written. In the space-separated sample tempK reads below zero, which no temperature in kelvin can, and nothing in the file says in which unit it is; there and in the file with its settings imuTemp is 0 at every sample, a placeholder rather than a reading.
  • The pad’s gravity. A board at rest, or hanging under a parachute, reads 1 g: 9.806 65 in m/s², 1 in standard gravities. The reader adds up the time the three axes’ magnitude lies within 10% of each of the two, every sample counting until the next, but for at most 5 s, the board’s longest regular step: past it the logging paused, and a sample says nothing of the time after, so a coast sample before a jump of the clock can’t count for minutes. The unit whose 1 g lasts at least 1 s and at least twice as long as the other’s is taken; otherwise the axes are kept as written. In the board’s sample logs the magnitude reads 9.8 for minutes on the pad and under the parachutes, against at most 1 s near 1 m/s² in the coast; in the space-separated one it reads 1 for 46 s and never 9.8, so its axes are read in g, × 9.806 65. One step in the sample logs passes 5.04 s, an 8 s one, and the limit changes neither unit’s time in any of them by as much as 0.05 s.
  • The board’s turning. With the board at about 1 g (within 15%), the direction the accelerometer reads is gravity’s, so as the board turns, that direction turns by as much as the gyroscope’s rates across it say. Over stretches of 10 steps at most 0.05 s apart, the reader adds up the angle between the accelerometer’s direction at each stretch’s ends and the gyroscope’s turn across that direction, |ω × â| dt, with ω read in radians per second. The unit, radians or degrees per second, in which the gyroscope’s turn lies within a quarter to 4 times the accelerometer’s is taken, given at least 3 stretches and 0.2 rad of turning; as the units are 57.3 times apart, at most one can be. Otherwise the gyroscope is kept as written. Swinging under a parachute turns the accelerometer’s direction without turning the board, so the ratio often runs below 1: in radians per second it is 0.40 to 1.6 in the 11 of the board’s own logs with stretches to check, and in degrees per second 1.3 in the other two, whose rates reach about 1,000; the 42-sample log, all on the pad, has none, and its gyroscope is kept as written.
  • angularVelRadS_* are read in radians per second, as their names say, if no value’s magnitude passes 349 rad/s, the 20,000 °/s full scale of Analog Devices’ ADXRS649, a gyroscope made for very high rates (ADXRS649 data sheet). In the file with its settings they pass it, so they are kept as written.
  • altitudeM is checked against the pressure: from the first sample to the one of lowest pressure, the climb in the U.S. Standard Atmosphere, 1976 (hpr_atmos::ussa76), against the column’s. Within 25%, meters are borne out; beyond, the file is refused (LogError::Unit): the primary track has no unit to be read in. Under 10 m of climb, or without a pressure in pascals, the check can’t be made and a note says so. unfiltAlt, whose name states no unit, is read in meters on the same check, and otherwise kept as written.
  • velocityMS is checked against altitudeM: over the steps at most 0.05 s long, the least squares slope of the altitude’s change on the speed times the step must lie within 25% of 1; a slope that can’t be computed, as when the sums overflow, isn’t within it. This ties the clock, the speed and the altitude together: a speed in feet per second, or a clock in seconds, would be 3.28 or 1,000 times off. Under 10 m of movement the check can’t be made and a note says so; a slope beyond is kept as written. In the sample files it is 0.86 to 1.08.

The primary track. The record’s altitude is altitudeM, the board’s own altitude above its zero, the pad: in the sample files that start on the pad it reads within 2.1 m of zero at the first sample, but for the replayed file’s 5 m. It is a filtered estimate, smoother than the pressure; unfiltAlt, the board’s altitude before filtering, is on a zero the file doesn’t state (in the board’s logs within 150 m of the standard atmosphere’s altitude of the pressure, in the space-separated file the same as altitudeM’s), so it isn’t the track. The vertical speed is velocityMS, the temperature tempK and the battery battV, where they are read in their units. The track takes the samples in time order, the first at each time, leaving out any sample without a time; a note counts those put in order or left out, and the channels keep every one in the file’s order. Where the altitude reads more than 100 m off zero at the track’s first sample, as in the file after a restart in flight, a note says so: each of the two files is its own record.

Events. Each time drogueFired or mainFired goes from 0 to 1 is an Event named by its column, at the sample’s time, in time order: the names say the board fired that channel. A flag already 1 at its first sample is no event. flightState changes are no events: the codes’ meanings aren’t published.

The settings. In a CONFIG line, BOARD_NAME names the logger and FIRMWARE_VERSION the firmware; the other settings are left out, counted in a note, as no source gives their units. Without a BOARD_NAME the logger is AerospaceNU SillyGoose.

Cells. An empty cell, or one that isn’t a number, is a gap (NaN), counted in a note. A value too large to hold once converted to its unit, such as an accelerometer’s 10³⁰⁸ g in m/s² or a gyroscope’s 10³⁰⁸ rad/s in °/s, is refused, naming its line and column (ADR-258 §3). A row with fewer cells than the header names has the rest as gaps, counted in a note; a row with more is refused, naming its line, as is a file whose rows hold under a quarter of the cells the header names (past the first 65,536), so that a short file can’t make the reader store far more gaps than it writes cells. Lines may end in CR, LF or both, blank lines are skipped, and a byte-order mark is dropped.

Constants§

MAX_COLUMNS
The most names a header may have: the board writes 20, or 28 with its settings.

Functions§

read
Reads a SillyGoose log from its text.
recognizes
Whether the text is this layout: its first line that isn’t blank is a header (module docs), or a CONFIG line with a header after it.