Practical GNSS and RTK resources for real machines.

Carrier phase RTK reaches centimeters by reducing shared errors, resolving integer ambiguity, and checking whether a Fixed solution remains trustworthy.

RTK GPS adds correction data to ordinary GNSS for centimeter positioning, but the receiver, correction source, antenna, and workflow all matter.

GNSS error budgets start in space, pass through the atmosphere, and end at the antenna, marking which residuals a receiver can model and which require RTK.

RTK GPS can improve field mapping, but acreage, boundaries, datum, antenna setup, and export metadata decide whether the map still lines up later.

Phone-based RTK is a packaging decision: move the antenna and positioning engine outside the handset while the phone keeps apps, maps, and network.

Field-trial repeatability depends on correction source, datum records, iOS or Android app support, drone GCP reuse, and a clean RTK point archive.

Why host-based RTK moved the software engine off-chip, how that architecture lowered cost, and why the responsibility boundary still matters today.

AG3335 suffixes define GNSS responsibility boundaries, from PVT-first modules to RTK+DR packaging, vehicle dead reckoning, and OPEN compute paths.

Kalmix chose AG3335 as a GNSS platform because real machines need L1/L5 tracking, firmware control, RTK behavior, and deployable module architecture.

A developer reference for RTCM 3.x parsing: frame sync, CRC-24Q, MSM mask matrix, satellite and cell fields, and base-station message types.

A field-level NMEA 0183 sentence dictionary for parser builders: GGA/RMC/GSA/GSV field tables, talker IDs, version diffs, and GNSS code values.