gps-details

Garmin Niggles

Introduction

There are a few minor niggles in the Garmin data, which I have listed on this page.

Temporal Offsets

Temporal offsets of 1 or 2 seconds seem to affect Garmin watches. I don’t think that I’ve ever seen the data being early, only ever delayed.

This is not a major issue per-se, but something to be aware of when testing or comparing devices. The delay can sometimes change by a second or two in the middle of a session.

Frozen Positions

If you look at position-derived speeds in GPSResults you will sometimes see some artefacts like the one shown below. This happens because the latitude + longitude is occasionally not updated, resulting in a speed of zero, then double, then normal.

Some notes about this behavior:

• This only affects the modes which do not implement major filtering; e.g. windsurf, kitesurf, other, etc.
• It may happen once or twice during a session, or it may happen frequently.
• These artefacts have also been evident on older watches, and I’ve seen it on the vivoactive 3 or 4.
• This behavior only ever affects the latitude + longitude positional data, not speed data.
• The behavior isn’t ideal, but it is not hugely problematic, even for alphas.
  • The latitude + longitude only dictates where the alpha can start and end.
  • The actual calculation of alpha results uses the Doppler-derived speeds, which are unaffected.

Aliasing

The Airoha chipset has a fixed update rate, which could well be 10 Hz in Garmin watches, but with an NMEA output rate of 1 Hz.

This results in some degree of aliasing, which I wrote up when studying my Garmin and COROS devices.

• The early COROS firmware(s) for Airoha-based watches also exhibit aliasing, just like Garmin and 1 Hz Motions.
• In May 2024, COROS broke their Airoha-based watches with a similar issue to Garmin watches in early 2022.
• In Sep 2024, COROS resolved the May issue (APEX 2 + APEX 2 Pro) and also got rid of the aliasing.

It is worth noting that Garmin watches using the MT3333 do not exhibit any aliasing, despite the chipset also being 10 Hz. The ideal would be for a low-pass filter being applied in the Airoha, avoiding the aliasing as per my earlier article.

Fortunately the potential for aliasing does not appear to have a significant impact when speed sailing, softened somewhat by the use of 2s averages.

Spikes

This isn’t a niggle per-se, just a quick note to say that spikes do occasionally occur during a crash, or coming ashore.

This is pretty rare and I had to sail for 100s of hours before seeing any spikes from my Forerunner 255.

However, the spikes are suppressed by the standard acceleration filters of GPSResults and GPS Speedreader.