Often a GPS receiver will perform just fine in the laboratory or on the ground, but then start exhibiting tracking issues when airborne. There are many possible factors to consider here, including dynamic RF environments frequently encountered during live flights. Frequently, however, the GPS tracking performance is closely related to real-time power supply quality. The GPS receiver requires (demands) a "consistently clean" source of power. Current and/or voltage transients (spikes) will destabilize the crystal oscillator (clock) onboard the GPS receiver, which in turn will usually result in an unwanted lockbreak condition.
Many power supplies look fine when monitored on the ground, but the real question is: How well does the aircraft GPS power supply suppress transient current/voltage spikes during a live flight when:
a.) various RF transmitters are turned on, then off.
b.) the aircraft engine RPMs are adjusted (rev up, rev down)
c.) the aircraft heaters are turned on, then off.
"Stable" aircraft power supplies can sometimes fail (i.e., pass along transients to the GPS receiver) under these three scenarios, resulting in poor GPS performance under real aircraft operating conditions. If your particular application incorporates other dynamic load devices, it would be worth testing their effects (on/off) on the GPS power supply as well.
The solution is to develop a GPS power supply which effectively suppresses all voltage *and* current spikes/transients even under heavy (and/or rapid) load switching conditions. Some DC-DC converters (the cheaper ones) will provide good voltage transient suppression, but poor current transient suppression. A good quality DC-DC converter will have large capacitors *and* inductors to filter both sides of the power transient spectrum. If your aircraft GPS power supply can provide consistently clean DC power to the GPS receiver even under the load switching scenarios listed above, then odds are good that your GPS tracking performance will be favorable during actual flight operations.