GPS-based time and frequency solutions

More-precise low-cost devices are becoming available as more specification-sensitive app requirements gain market momentum

BY DAVE JAHR, KEN HARTMAN, and KEITH LOISELLE
Connor-Winfield, Aurora, IL
www.conwin.com

The global positioning system (GPS) constellation comprises 24 or more satellites in six orbital planes at an altitude of approximately 12,550 miles. Synchronized with on-board atomic clocks, the satellites continuously transmit ranging signals, their current time and orbital dynamics on two frequencies in the L-band labeled L1 (1575.42 MHz) and L2 (1227.6 MHz). Civilian GPS commonly uses only the L1 band, while surveying and military uses require both the L1 and L2 bands.

The GPS constellation was designed to provide a view of at least four satellites from any location on Earth, at any given time. Using the ranging measurements from four GPS satellites and the transmitted satellite orbit information, a GPS receiver can accurately solve for position and time. This is accomplished mathematically by solving for the intersection between four or more spheres centered on four or more satellites.

The GPS position solution allows for the computation of the current latitude, longitude and elevation to within 5 m (using the L1 band only), along with precise timing to within 10 ns. If the GPS receiver’s position is fixed, then the position solution can be fixed and the ranging measurements can be applied to further increase the accuracy of the time solution. A fixed-position receiver also allows a time solution with only one satellite in view, which can occur if the receiver’s view of the sky is partially blocked.

The GPS system is synchronized to its own time scale — GPS time as provided by the United States Naval Observatory (USNO). The USNO operates two Master Clock facilities that provide precise time to the GPS system, one in Washington, DC (USNO) and one in Colorado Springs, CO (AMC). GPS time is also synchronized with Coordinated Universal Time (UTC), the international time standard. But unlike UTC, which is corrected by leap seconds to compensate for changes in the Earth’s rotation, GPS time is a continuous time scale with no leap corrections.

The lack of jumps in the GPS time scale simplifies the continuous time and position solution of a GPS receiver. In 1980, UTC and GPS time were aligned. But since then, UTC has been corrected periodically by adding leap seconds, so that GPS time is now 15 s ahead of UTC. A UTC correction parameter and a prediction of future leap seconds are included in the data broadcast by GPS satellites. This allows a GPS receiver to calculate UTC from GPS time.

GPS receivers output the satellite information and time/position solution in a standard format. The most common formats are defined by the NMEA0183 or NMEA2000 specifications. These specifications define sentences that include the visible satellites, satellite signal strengths, GPS or UTC time solutions, and position solution. Manufacturers may also define sentences that output data in a custom format or that provide access to special features of their GPS receivers.

GPS-based timing and sync solutions are used as both a time source or time reference and as an easy means of time transfer. In the past, higher-end rack-mount systems, often based on rubidium technology, capable of maintaining extremely tight stability characteristics over time with primary reference clock precision holdover for many weeks, were a standard product offering in this area. Of course, with higher-end technology come higher-end prices. One could and can expect to pay from the thousands of dollars to tens of thousands of dollars for single rack-mount units of these types.

Today, however, with advances in component design, manufacturing technologies and changing specification requirements, GPS based time and synchronization solutions have become very common across many different applications in a variety of different technological areas. From simple GPS timing receivers, available in low volumes under $75, to more sophisticated GPS-based devices, with excellent phase noise and holdover characteristics priced under $500 per device, GPS-based timing and synchronization has become accessible for any application.

The process starts with a GPS receiver optimized for timing requirements. The GPS timing output signal, or 1 pps (pulse per second), depending on the implementation, is generally accurate from nanoseconds to milliseconds and is a direct reflection of GPS time provided by the satellites. Accurate frequency outputs are also available from certain GPS timing receivers, generated from the PPS.

Phase

Precise time is an integral component of the GPS solution. A GPS timing receiver can solve location and time triangulation equations, thus determining its location to submeter and time to nanosecond levels. Continuous solution of the GPS equations allows for a local (at the receiver) reconstruction of the GPS time scale. Process noise events – deviations from the GPS time scale are expressed as phase (time) errors. Synchronizing the local clock of the receiver to the GPS time scale enables the receiver to produce an accurate one-pulse-per-second 1-pps output that follows the GPS time scale to as few as a couple of nanoseconds.

Holdover

During periods of GPS signal loss, the receiver enters a holdover mode and bases the timing outputs on the last GPS time and clock drift solutions and the stability of the local clock. By storing the last known-good time and clock drift solutions, the stability of the 1 pps and frequency outputs becomes dependent only upon the stability of the local oscillator.

For strict holdover requirements, the receiver’s temperature-controlled oscillator (TCXO) can be replaced with an oven-controlled oscillator (OCXO) with the required stability. Upon recovery from holdover, the 1 pps and frequency output return to the GPS-calculated solution.

Advanced GPSDO solutions

More precise, cost-effective GPS based timing and frequency devices are becoming increasingly available as more specification sensitive application requirements gain momentum in the marketplace. Building upon standard GPS timing receivers, more sophisticated GPSDOs (GPS disciplined oscillators) add phase-locked-loop circuitry and tighter-stability oscillators to increase phase and holdover performance of the devices.

These devices are available in many different sizes and form factors, from units with through-hole pins for PCB mounting to enclosed units with standard connectors for easy integration. Prices for GPSDOs are generally found in the $200 to $1,000 range, depending on form and features.

The FTS375 module is a GPS-driven mixed-signal phased-lock loop providing a 1-pps CMOS output from a Connor-Winfield GPS timing receiver and generating a 10-MHz CMOS and a 10-MHz sine output from an intrinsically low-jitter voltage-controlled crystal oscillator.

Highly precise time and frequency information is now available at a fraction of the previous cost. Users are no longer required to buy expensive and complex rack-mount systems when a well-designed GPS receiver with PPS and frequency output would better suit their needs. Here are some simple criteria to follow as you choose your GPS timing/frequency solution:

1. Make sure your GPSDO solution provider fully supports the GPS receiver portion of the device. If the provider is sourcing the GPS receiver portion of its solution, chances are it doesn’t have the requisite GPS experience to appropriately support its customers.

2. Make sure your GPSDO solution provider designs and builds the oscillators used in its devices. Again, if it doesn’t have specific oscillator experience, it will not be able to directly support its customers’ technical requirements.

3. Make sure your GPSDO provider has a long history of providing these types of GPS-based timing and frequency solutions. Combining the elements of these devices into a single design is not a simple endeavor. Complex filtering and GPS firmware designs are integral to the product. There are always less-than-reputable suppliers who have just thrown their solutions together with any available components without a full understanding of either their products or the applications into which they will be integrated.

4. Make sure your GPSDO supplier is familiar with the attributes of your application. Discuss your application with the supplier prior to purchasing any GPSDO device. A simple technical conversation should tell you whether or not the supplier can appropriately support your needs.

5. Make sure you know the exact technical requirements for timing and/or frequency control of your application. The more precise you are in conveying your specifications, the more likely you are to get the precise GPS solution for your application. Often organizations will over-spec their requirements and end up buying expensive systems that exceed their actual needs. ■