Use PTP protocol to generate PPS signal (one pulse per second)

Hi,
Is the Ubiswitch module (or any other of your boards) capable of generating a PPS signal (ie a 1ms pulse starting at the start of each second) based on PTP (precision time protocol) / IEEE-1588 ?

Currently, we get the PPS signal from a GNSS receiver, and transmit it wired, but using PTP would simplify cabling/reduce interference, excepted that some devices still need a physical PPS signal, so we need to reconstruct it.

Thanks a lot in advance
Best regards,
Felix

Hi Felix,

UbiSwitch doesn’t support PTP clock synchronization or generating a PTP-synchronized PPS output. For this application, the platforms to look at are RouterCore and RouterCore Nano.

RouterCore samples should be available in about a week, initially running OpenWRT without PTP. The PTP release is scheduled for later in November 2026. RouterCore Nano hardware is still in development, with availability estimated for February 2027.

Our initial PTP development is aimed at two modes:

  • Transparent clock: forwards PTP messages and accounts for the time they spend passing through the switch
  • Boundary clock: synchronizes a local clock to an upstream grandmaster, then distributes that time to downstream devices

Generating your physical 1 PPS signal with a 1 ms pulse width would require additional development beyond those network-timing functions. There are two approaches:

  1. Software-timed GPIO
    A synchronized clock could schedule a GPIO pulse every second. This is simpler to prototype, but processor load introduces timing jitter. For the Linux-based implementation, roughly 100 µs–1 ms is an initial planning estimate, with potentially larger excursions under load.

  2. Hardware-timed output
    A synchronized hardware clock or timer would generate the pulse directly, avoiding software scheduling jitter. This is the approach we’d investigate for sub-microsecond timing, but it requires additional firmware/driver work and validation of the output connection and clock synchronization.

These are development options rather than validated PPS specifications; PPS output isn’t included in the initial PTP scope described above.

What timing accuracy and jitter can your devices tolerate on the pulse’s rising edge relative to the PTP grandmaster? That will help us determine which approach is appropriate.