Legacy POS AV 510 / IMU6 recording questions

Prepared for a future request to the instrument provider or acquisition team. Not sent. No purchase, authenticated document request or external action has occurred.

We are decoding NASA/NSIDC ILVIS0 .013 airborne recordings. The six representative files identify AV-510, VER5, IMU6, at approximately 200 Hz. Their logged POS system identifiers and firmware/ICD combinations are:

• S/N2720: HW3.1-7, SW04.50-Feb05/08, ICD14.00, OS425B14.
• S/N3861: HW4.0-7, SW05.00-Dec16/09, ICD17.00, OS425B142.
• S/N4126: HW4.0-7, SW05.20-Nov04/11, ICD18.00, OS425B142.

The inventory supplies exact filenames, embedded dates and uncompressed source hashes. These identifiers belong to the logged POS system; please identify the actual IMU part/variant and serials separately where records permit.

1. For these exact firmware/ICD revisions, is Group 4 recorded before or after the navigation filter's estimated gyro/accelerometer bias and gain corrections? Is Earth rotation or transport rate removed from either its angle or velocity increments? Is the logging path different in Navigate and Standby modes?
2. What does legacy IMU type 6 identify? Is its sensor an LN-200/LN200ROM or another variant? What gyro/accelerometer calibration, noise, drift and scale-error limits apply to these installations, and over what temperature/time conditions?
3. Is its 24-byte Group-4 data six signed little-endian int32 increments? What are the exact velocity and angle scales, axis/sign definitions, clipping and overflow conventions? Our empirical hypotheses are 0.4 arcsecond/count and 3.38e-5 m/s/count; these are not assumed manufacturer definitions.
4. Does each packet represent a fixed 5 ms sensor integration interval or the difference between adjacent packet timestamps? Are tags at interval beginning, midpoint or end? How is synchronization performed, and what sensor-to-GNSS latency, jitter or oscillator-error bounds are supported?
5. Which factory or runtime corrections precede the recorded Group-4 tap: temperature, bias, gain, axis nonorthogonality, coning, sculling, filtering, automatic zeroing and lever-arm compensation? Are coefficients or processing flags logged, and can these functions be enabled independently?
6. Can you supply the legacy Message-1 installation/settings layout and rotation conventions, including reference/vehicle/IMU frames and lever-arm origins? Are there surveyed mounting/lever-arm records for the listed epochs? Logged settings reproduce mappings, but an authoritative layout remains needed.
7. Which receiver-only GNSS processing modes and uncertainty conventions applied during acquisition? Were common-mode position errors or IMU feedback present in the primary receiver output? Are independent position/covariance validation or calibration records available?

We use fused Group 1 only for decoder validation and motion context. We seek the processing history of Group 4 so that navigation-derived corrections are not mistaken for independent inertial evidence. Firmware-matched manuals, acquisition logs and calibration records would be more useful than a current integrated-navigation accuracy specification.