The 1 PPS output of a Motorola M12+ GPS receiver gives a rather precise pulse at each UTC second. The <100 ns accuracy of this pulse is more than adequate for hundreds of applications.
But some applications want better than 100 ns accuracy. In this case a careful study of the 1 PPS pulse is in order. The older Oncore VP had a 9.54 MHz oscillator and thus exhibited a sawtooth of approximately ± 52 ns (1 / 9.54 MHz = 104 ns). The newer M12 receiver has a higher frequency local oscillator and a much smaller sawtooth.
Each second the receiver computes a solution and determines when the next 1 PPS signal should occur. However to simplify the hardware it is most convenient to generate the 1 PPS signal using the leading edge of the internal clock. As the clock continues to tick away at ~20 MHz the M12 has more than enough time to determine which clock pulse should be routed to the 1 PPS output. The result is a certain granularity of the 1 PPS output. Given that both the local oscillator and the GPS solution drift in time there is a tendency for a "sawtooth" looking plot of the 1 PPS output.
Below is a collection of traces showing normal and unusual sawtooth patterns.
| M12+ 1 PPS phase - fairly typical sawtooth pattern |
|
| M12+ 1 PPS phase - sawtooth pattern getting smaller |
 |
| M12+ 1 PPS phase - sawtooth period getting longer |
 |
Not sure what causes this effect. There appears to be a larger drift in operation below the usual oscillator drift.
| M12+ 1 PPS phase - very short sawtooth and phase jumps |
 |
The "suspension bridge" is simply what happens when there is a oscillator reversal. Note the slope of the sawtooth on the left and two minutes later on the right. This is the pattern you see when the local oscillator changes sign relative to GPS.
| M12+ 1 PPS phase - suspension bridge effect, slow rate |
 |
Do not apply sawtooth correction here:
Drift in opposite direction.
| M12+ 1 PPS phase - upside down suspension bridge effect |
 |
This one is cool.
| M12+ 1 PPS phase - suspension bridge, quick rate, wiggles |
 |
This was a trace shortly after the GPS antenna was re-installed. This trace is 400 seconds.
| M12+ 1 PPS phase - GPS re-acquisition phase jump correction |
 |