Design: NAVPAC contained three UHF and VHF receivers to receive the 400MHz and 150MHz signals from up to three simultaneously in-view Transit satellites. Each receiver automatically searched for, acquired (without a prediction of when a satellite would be in view), tracked a separate signal set and recorded the uncorrected Doppler frequency of the 400MHz signal with the corresponding ionospheric refraction information. Control logic prevented more than one receiver from tracking a single satellite and other logic prevented a receiver from acquiring and tracking any 400 and 150 MHz signal sets that might be similar to those from a Transit satellite. Acquisition and tracking of signals occurred at elevations down to -10° relative to the horizontal plane of the host vehicle with 400 MHz acquisition occurring at -138 dBm and tracking to levels of -145 dBm. Each receiver channel included a phase lock loop that was required to operate with a maximum Doppler rate-of-change that would be as high as 400 Hz/s for the 400 MHz channel. This greatly exceeded the rate term requirements prior ground based Transit satellite receiver sets had been designed for, and implied if the phase locked loop was only a second order system (two integrators) the bandwidth would have to be extremely wide to limit the dynamic phase error. An alternative to a wide bandwidth second order system was a third order loop (three integrators). However, a third order loop system could introduce instability problems during acquisition. The NAVPAC solution was to use a “two and a half order” loop in which a lossy integrator (a passive lag network) was added to a second order loop, reducing the dynamic phase error but not introducing instability during acquisition. (The two and a half order phase lock loop was previously analyzed and developed by L. Katz of the Electrac Corporation, Anaheim CA.)
The Data Processing Unit (DPU) controlled all NAVPAC data collection and movement. At the heart of the DPU was the central processing unit (CPU) which accepted the necessary data and created special data files stored in a temporary buffer. When the buffer was filled, the contents were automatically transferred to a host vehicle tape recorder. Much later, the tape recorder contents were transmitted to a ground station. A tape recorder bypass mode allowed the buffer contents to be transmitted directly to the ground in the event of a tape recorder failure. The CPU also accepted data from a time code generator driven by an ultra-stable 5 MHz oscillator to create precision time annotation for all recorded data and the event occurrence.
The Power Unit was a DC/DC converter to convert host vehicle unregulated power to the various voltage levels required by NAVPAC. A switching regulator was chosen for its inherent high efficiency, a characteristic of prime importance, and was also sized for maximum efficiency at the normal expected load.
Packaging Techniques: Weight and volume were not limiting factors during the design phase but schedule was - therefore NAVPAC was packaged in the most expedient (not necessarily the smallest or minimum weight) configuration. Several fabrication and packaging techniques that were used represented a major departure from ‘typical’ spaceflight hardware including: a) dual inline integrated circuits mounted on plug-in boards with point-to-point backplane wire wrap interconnects and b) extruded aluminum boxes as enclosures for most of the electronics. Standard printed circuit and flight construction techniques were used for items such as all RF assemblies, the frequency synthesizer, the ultra-stable oscillator and the Power Unit.
NAVPAC, less the MESA, was mounted on a 28 inch x 46 inch baseplate with a maximum height of 10 inches (see photo from APL/JHU TG 1317 referenced above) and weighed approximately 134 pounds. The antenna envelope was a cylinder, 8.5 inches in diameter and 64 inches long and weighed 4.5 pounds.
Results: Six units were fabricated, tested and delivered to the host vehicle integration contractor. There were no plans for any additional systems. System 01, launched during the second quarter of CY 1977, functioned flawlessly. In-orbit results processed by NSWC compared well with the data obtained by prior ground testing using a fixed site antenna. The following results were consistently obtained:
Clock calibration error: <35u seconds
Average filtered range noise 10 to 15 cm
Navigation consistency 1.7 to 3.5 m
Slant range average error 1.8 m
Along track average error 0.3 m
These results were a significant improvement over the prior to NAVPAC accuracy and knowledge of the host vehicle orbit reconstruction problem.