ubx-cfg
Static ESP Testing - Phase 7
Overview
The objective of phase 7 was to compare the optimal 5 Hz and 10 Hz configurations. The SYRAC GPS devices were all intended to use GPS + Galileo + BeiDou B1C (no GLONASS), and a maximum of 28 satellites.
The tests were started on 2026-07-14 @ 0031 and the duration was just under 10 hours. The results showed that 5 Hz is far more accurate than 10 Hz, which is consistent with phase 4, 5, and 6. Speed Accuracy (sAcc) can also be quite misleading.
Configurations
The following device configurations were tested.
| ID | Constellations | Rate | Max Sats | ESP | Observed time distribution (ms) | Drops? |
|---|---|---|---|---|---|---|
| D1 | GPS + Galileo + BeiDou B1C | 10 Hz | 28 | 160 MHz | 99: 11, 100: 338258, 101: 11 | - |
| D2 | GPS + Galileo + BeiDou B1C | 10 Hz | 28 | 160 MHz | 99: 5, 100: 352086, 101: 5 | - |
| D3 | GPS + Galileo + BeiDou B1C | 5 Hz | 28 | 160 MHz | 199: 18, 200: 176100, 201: 18 | - |
| D5 | GPS + Galileo + BeiDou B1C | 5 Hz | 28 | 160 MHz | 199: 12, 200: 176461, 201: 12 | - |
| S3 | GPS + Galileo + GLONASS | 5 Hz | 32 | 80 Mhz | 199: 4, 200: 176099, 201: 4, 400: 8 | Y |
| SY1 | GPS + Galileo + BeiDou B1C | 10 Hz | 28 | 160 MHz | 99: 3, 100: 351202, 101: 3 | - |
| SY2 | GPS + Galileo + BeiDou B1C | 5 Hz | 28 | 160 MHz | 199: 36, 200: 175999, 201: 36 | - |
Notes:
- S3 was intended to be GPS + Galileo + BeiDou B1C @ 160 MHz, but somehow reset itself to GPS + Galileo + GLONASS @ 80 MHz
- S3 was the only device to drop any frames, presumably because the ESP32 was running @ 80 MHz
Statistics
Click the GPY filenames for charts showing SOG, Sats, HDOP, and sAcc.
| File | Description | Mean | Median | Stddev |
|---|---|---|---|---|
| SY2_2607140033.gpy | GPS + Galileo + BeiDou, 28 sats @ 5 Hz | 0.024 | 0.021 | 0.013 |
| D5_2607140031.gpy | GPS + Galileo + BeiDou, 28 sats @ 5 Hz | 0.024 | 0.021 | 0.014 |
| ubxGPS_2607140032.gpy | GPS + Galileo + GLONASS, 32 sats @ 5 Hz | 0.024 | 0.021 | 0.014 |
| D3__2607140033.gpy | GPS + Galileo + BeiDou, 28 sats @ 5 Hz | 0.024 | 0.021 | 0.013 |
| D1_2607140032.gpy | GPS + Galileo + BeiDou, 28 sats @ 10 Hz | 0.042 | 0.039 | 0.022 |
| D2__2607140033.gpy | GPS + Galileo + BeiDou, 28 sats @ 10 Hz | 0.042 | 0.039 | 0.023 |
| SY1_2607140034.gpy | GPS + Galileo + BeiDou, 28 sats @ 10 Hz | 0.047 | 0.045 | 0.025 |
Speed Over Ground (SOG)
The 5 Hz devices all performed significantly better than the 10 Hz devices.
The same was also evident in the median values for SOG.
The performances from the perspective of standard deviation were similar to the mean values.
Speed Accuracy (sAcc)
The mean “speed accuracy” at 5 Hz is reported as being WORSE than 10 Hz.
The median “speed accuracy” at 5 Hz is reported as being WORSE than 10 Hz.
The standard deviation for “speed accuracy” at 5 Hz is also WORSE than 10 Hz.
Since we know the true Speed Over Ground (SOG) is zero, we can be sure that the Speed Accuracy (sAcc) is misleading during these tests.
Observations
- S3 (ubxGPS) was intended to be GPS + Galileo + BeiDou B1C
- Somehow it was switched to GPS + Galileo
- S3 (ubxGPS) was also the only device to drop any frames
- Presumably caused by the ESP32 running @ 80 MHz
- SY2 dropped saw a few “mini spikes”, similar to previous tests
- Perhaps due to the M10 using “balanced” power mode
- Performance of all 5 Hz devices comparable
- SY1 not as good as D1 and D2 @ 10 Hz, maybe baud rate?
- 5 Hz accuracy is significantly better than 10 Hz
- There is no obvious advantage to 10 Hz for speed sailing
- sAcc is very misleading
- It says that 10 Hz is most accurate, which is NOT correct
Conclusions
- 5 Hz is significantly more accurate than 10 Hz
- Phase 7 results are consistent with phases 4, 5, and 6
- Speed Accuracy (sAcc) cannot be taken literally
- 5 Hz was better than 10 Hz, but sAcc said the opposite!