A GPS-disciplined clock combines a local crystal or oven-controlled oscillator with a GPS receiver, using the extremely accurate time broadcast by the satellites to continuously steer the local oscillator so it tracks a global standard. On its own an oscillator drifts; anchored to satellite time it becomes an accurate, self-correcting reference that a site can use as its master clock. This guide explains how the discipline loop works, what happens during holdover when the signal is lost, why antenna placement matters, and why a stable on-site time source keeps SCADA timestamps aligned across many sites.
GPS-Disciplined Clock in one line: A GPS-disciplined clock, sometimes called a GPSDO, is a device that uses the precise time signal from GPS or other satellite systems to continuously correct a local oscillator, producing an accurate reference clock. It is commonly used as the top-level time source for a site, feeding NTP or PTP so that every device downstream traces its time back to a common satellite-derived standard.
Every free-running oscillator drifts. A quartz crystal changes frequency slightly with temperature and age, and even a high-quality oven-controlled oscillator wanders over hours and days if nothing corrects it. A GPS-disciplined clock solves this by pairing that oscillator with a GPS receiver and continuously comparing the local clock against the time recovered from the satellites, then nudging the oscillator's frequency to keep the two in step. The satellites carry atomic references, so the time they broadcast is far more accurate than any modest local oscillator, giving the site an anchor to a global standard it could never hold on its own.
The correction is deliberately gentle. Rather than yanking the clock whenever a momentary measurement looks off, the discipline loop applies small, smooth adjustments and averages the satellite signal over time to filter out noise. This blends the best of both sources: the long-term accuracy of the satellite reference and the short-term stability of a good local oscillator, which is steady from one second to the next even though it slowly drifts over longer spans. The combined output is both accurate and smooth, which is exactly what a time reference needs to be.
The reason to build this into a site is that it turns the location into its own authoritative time source. Instead of every device reaching across the internet to some distant server, the site has a local master that is accurate to a small fraction of a second, or far better, and that can drive NTP for general synchronization or act as a PTP grandmaster for the tightest applications. Everything downstream then traces its time back to the same satellite-derived reference, which is what makes timestamps across the whole site mean the same instant.
The satellite signal is not always available. An antenna can be obstructed, the receiver can lose lock in bad conditions, or the feed can be jammed or degraded, and when that happens the disciplined clock enters what is called holdover. In holdover the device no longer has a reference to correct against, so it falls back on the local oscillator alone and coasts on the last known good frequency correction. From the moment lock is lost, the clock begins to drift away from true time, and the whole value of a good disciplined clock is how slowly that drift accumulates.
This is why the holdover specification matters so much when choosing one. A device built around a cheap crystal may drift noticeably within minutes of losing the signal, while one built around a high-quality oven-controlled or rubidium oscillator can hold accurate time for many hours, sometimes far longer, before the error grows past a useful threshold. The holdover spec, usually stated as how much error accumulates over a given period without a reference, tells you how long the site can ride out a signal outage before its time can no longer be trusted, which is a direct measure of resilience.
A well-designed time source treats holdover as a normal condition to be managed, not an emergency. It flags that it has lost the satellite reference so downstream systems know the clock is now coasting, and it keeps disciplining the oscillator's environment, such as holding the oven temperature steady, so drift stays minimal until the signal returns. When lock is reacquired, a good clock steers back to true time gradually rather than jumping, so it does not create sudden gaps or duplicate timestamps in the data that depends on it.
The disciplined clock is only as good as the signal it receives, and that signal starts at the antenna. A GPS antenna needs a clear view of a wide swath of sky to track enough satellites for a reliable time fix, so it is typically mounted high and in the open, away from obstructions and from sources of electrical interference. A poorly placed antenna that only sees a narrow slice of sky loses lock intermittently, forcing the clock in and out of holdover and undermining the very stability the device was installed to provide. Getting the antenna right is often the difference between a rock-solid reference and a flaky one.
The payoff for getting all of this right is that every site becomes an island of accurate, self-sufficient time. A remote wellpad, pump station, or substation with its own GPS-disciplined clock keeps its clocks correct even when the wide-area link to the outside world is slow or intermittent, because its reference comes down from the satellites rather than across the network. That local accuracy is what lets buffered data recorded during an outage carry correct timestamps, so it lands in the right place on the trend once the link recovers.
For a cloud SCADA platform such as Merobix, on-site time discipline is what makes a multi-site timeline coherent. When each location traces its clock back to the same satellite standard, an event at one site can be lined up against a related event at another with confidence, because both were stamped against the same global reference rather than against independently drifting local clocks. The GPS-disciplined clock is the quiet foundation under all of it: the accurate, holdover-resilient source that keeps timestamps meaningful across an entire fleet of remote assets.
A plain GPS receiver reports the time and position it recovers from the satellites, but its output is only as steady as the moment-to-moment signal, and it has nothing to fall back on when the signal drops. A GPS-disciplined clock adds a stable local oscillator that the satellite time continuously steers, so the output is smooth from second to second and keeps running accurately during signal outages. That local oscillator is what provides holdover and makes the device usable as a site reference.
It depends entirely on the quality of the local oscillator, which is what the holdover specification describes. A device built around a basic crystal may drift noticeably within minutes, while one using a high-grade oven-controlled or rubidium oscillator can hold useful accuracy for many hours or longer. Always check the holdover spec against how long your site might realistically lose the satellite signal.
Yes, that is one of its most common uses. Because it produces accurate, satellite-anchored time locally, it can act as the top-level reference that NTP distributes across a network, or as a PTP grandmaster for applications needing sub-microsecond accuracy. Every device downstream then traces its time back to the same satellite standard through the disciplined clock.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.