System Settings Sheet v9.6
Nobotic USV, Cape Dory 25D. Every configurable number aboard, in one place, with its source. Change a number here first, then in the device, then log it. Where a guide disagrees with this sheet, this sheet wins on setpoints and the guide wins on wiring.
1. Charging (the five main-bank MPPTs)
The SBMS0's EXTIO4 pauses all five controllers through VE.Direct remote on/off cables in parallel. The controller profiles below stay as the backstop if a plug shakes loose.
| Setting |
Value |
Where |
Why |
| Profile type |
User-defined lithium |
All 5x Victron SmartSolar (2x 100/20, 3x 75/15) |
LiFePO4 8S |
| Absorption voltage |
27.6 V (3.45 V/cell) |
same |
Conservative top; worst-case 300 mV group drift tops a group at ~3.7 V - aging territory, not damage |
| Float voltage |
27.0 V (3.375 V/cell) |
same |
|
| Temperature compensation |
OFF |
same |
LiFePO4 wants none |
| Absorption time |
30 min |
same |
Mike, 2026-09-05 |
| RX pin |
Remote on/off (VictronConnect, on all five) |
same |
The VE.Direct port carries the SBMS0 pause cable ASS030550320; this setting makes the controller obey it. Using the port for the cable rules out a GX device or dongle on that controller |
| Low-temperature charge cutoff |
Not set in the MPPTs; it lives in the SBMS0 (parameter 29 = 0 C, below) |
SBMS0 |
The 32 F stop is the whole cold plan; there are no heaters (deleted, do not re-propose). CONFIRM C1 |
| Bluetooth |
DISABLED after commissioning, on all five (too much watt-hour draw). With the pause cable on each VE.Direct port there is no per-string yield data anywhere; total charge current comes from the SBMS0 shunt |
all controllers |
Power draw first; attack surface second |
Emergency bank: SmartSolar 75/10, same user-defined lithium profile, absorption 27.6 V / float 27.0 V (the Dyness built-in BMS is the backstop above that). Victron Orion-Tr 24/24 float charger: output at 27.0 V (float only, monthly exercise per Build Guide v9.6 4.3).
SBMS0 watchman settings
| Setting |
Value |
Where |
Why |
| EXTIO4 |
Type 1 |
SBMS0 Automation menu |
The charge-pause output: drives the five VE.Direct remote on/off cables in parallel through a 1k 1 W resistor. Closed = charging allowed (CONFIRM C1, polarity per Dacian) |
| EXTIO3 |
Type 2 |
SBMS0 Automation menu |
Per the battery-system block |
| Parameter 30 |
1182 |
SBMS0 |
The 55 C high-temperature stop |
| Parameter 29 |
32 F (0 C), the low-temperature charge stop, entered as the SBMS0's own XT1 reading in ice water, not as a nominal zero |
SBMS0 |
3144 per the manual; confirm against firmware, pending C1 (Mike emails Dacian). XT1 thermistor on pins 3-4, against a cell face. The DROK 10K 3950 probe reads about 4 F low against the SBMS0's Murata curve near freezing, so calibrate in ice water (Battery Box Guide v9.6 Part 8 and Part 10) |
| XT1 ice-water reading |
To be measured at the Part 10 bench check: what the SBMS0 displays with XT1 in ice water (about 28 F if the 4 F offset holds), and what it displays with XT1 unplugged |
SBMS0 display |
Both numbers go here and in the punch list. The first sets parameter 29; the second lets the parser tell a broken lead from a cold pack |
| Save device settings |
REQUIRED after the above |
SBMS0 |
Or the EXTIO settings are lost at power cycle |
| USB serial to Pi A |
Optional telemetry only |
WiFi/USB add-on board |
No CAN, no RS-485. The Pi never controls charging; the BMS never controls loads |
2. Propulsion (primary VESC, Flipsky 75200 Pro V2.0)
| Setting |
Value |
Why |
| Battery voltage max |
28.0 V |
Above absorption, below the MOV ladder |
| Battery voltage min (cutoff end) |
22.0 V |
Pack floor |
| Battery current cap, normal |
~85 A (= 2 kW at 25.6 V) |
Mission ceiling |
| Battery current cap, CUT-BURST |
~180 A, command-gated, 10 s |
Build Guide v9.6 5.3 |
| Motor phase current cap, normal |
~180 A |
|
| Motor phase current cap, CUT-BURST |
~300 A |
The 75/300 rating; stall torque ~74 N-m |
| Regen termination |
27.8 V (corrected at v9.4 from 28.4 V, which sat above battery max) |
Taper from 27.6 V |
| Motor temp derate / shutdown |
60 C / 80 C |
Never raised |
| Backup VESC (cold standby), Boat 2 only (Boat 1 sails with one VESC) |
Motor and battery caps ~12-15 A, ~300 W, soft-start |
Cruise-only envelope |
3. Steering (Ewellix CAHB-22, single actuator; the Thomson Electrak HD is Boat 2 reference only)
The Boat 1 actuator is the Ewellix CAHB-22-F4E-3000529-BAA0PT-000 bought 2026-09-10: 24 V DC, 300 mm (11.8 in) stroke, self-locking screw, built-in position feedback, no CAN. Ewellix rates the rod for 5 km of double strokes; the voyage estimate was 11.2 km on the 190,000-move budget, so the move budget is halved by a wider deadband and a longer minimum interval. Confirm the rated-load basis with Ewellix before the numbers are fixed.
| Setting |
Value |
| Drive |
Reversing 24 V DC at the central node (the 30 A relay modules on hand as a reversing pair, or a 24 V H-bridge), current-sensed; no CAN, no J1939 |
| Passage-mode deadband |
TO BE SET, wider than the +-10 deg carried through v9.5 (rod-travel budget) |
| Passage-mode minimum move interval |
TO BE SET, longer than the 3 min carried through v9.5 (rod-travel budget) |
| Passage-mode move style |
Whole-error moves; budget about 150 moves/day (half of the 300/day the Thomson was budgeted at) |
| Precision-mode deadband / interval |
TO BE SET on the same halving (was +-3 deg / 20 s); budgeted ~200 h per voyage |
| Storm / streamed JSD |
Rod held at center by the self-locking screw, drive de-energized, zero moves (no brake: the screw holds) |
| Authority schedule |
Deflection cut above ~8 kt boat speed |
| Position source |
The actuator's own feedback, cross-checked against encoder A on the stock; disagreement beyond a threshold set at commissioning raises an alert |
| Rudder rate |
About 2 deg/s at the 12 in tiller station (rod 0.40 to 0.51 in/s). Accepted 2026-09-16: at 2 to 3 kt the rudder's force comes from water speed, not swing rate, and storm mode parks the rod. Zero to 25 deg takes about 12 s |
| Jam-detection stall timer |
TO BE SET, and longer than the Thomson's. A move takes about two and a half times as long on this unit, so a timer tuned to the old actuator would false-trip on a healthy move |
| Jam response |
Commanded effort + no motion on either feedback -> the current-sense trip cuts the drive within seconds, then the de-energized-closed power-cut relay opens |
| Exercise |
Weekly full-stroke sweep |
4. Battery protection and environment
| Setting |
Value |
Where |
| Heater thermostats |
None. There are no heaters (deleted, do not re-propose). The SBMS0's 32 F charge stop (section 1) is the whole cold plan |
- |
| Charge-continue bypass |
Relay channel 8 closes after 10 min of silent watchman stream with the battery probes above 32 F; opens when the stream returns; shore-commandable both ways. A dead SBMS0 pauses charging for at most 10 minutes; charging continues after the bypass |
Central node; BMS Software Change Notes v9.6 sec 2 |
| Cell-temperature trip, level |
150 F (65 C). Any of the six DS18B20 pack probes above 150 F: isolate the bank, HIBERNATE, Iridium alert. Set by Mike 2026-09-15. The reasons: the EVE MB31 cell sheet allows discharge up to 140 F, so a trip at the rated limit could fire on a hot still day in the tropics with the cabin near 104 F and the pack a few degrees behind; the SBMS0's own charge stop at 131 F (parameter 30) already stops charging well below it; and LiFePO4 cells do not start heating themselves until far above 150 F, so 150 F still gives hours of warning. |
Pi + central node; BMS Software Change Notes v9.6 sec 4; Routing Guide v9.6 11.7 |
| Cell-temperature trip, rate |
A rise of more than 10 F in five minutes on any probe: the same action. It is the real detector of a cell fault |
same |
| Cell-temperature trip, release |
When every probe has read below 120 F for one hour |
same |
| Cell-temperature warning, alert only |
Any pack probe above 131 F: shore alert, no action. Matches the SBMS0's own charge stop at 131 F (55 C, parameter 30) |
Pi; section 6 |
| SoC source |
The SBMS0 watchman's own shunt, read over its USB serial link (settled 2026-09-16). It is the only current sensor on the boat. If the stream goes stale the Pi sheds on a fixed conservative schedule and raises the quiet-watchman alert; stale numbers drive nothing |
BMS Notes v9.6 sec 6 |
| SoC load-shed ladder |
Shed at 70 / 50 / 30; everything dischargeable off below 5% SoC |
Pi + central node relays |
| Routine SoC target / floor |
90% / 10% |
Build Guide v9.6 4.1 |
| Cell temp caps |
Cap 80% SoC above 35 C cell temp; cap top-off above ~25 C |
Build Guide v9.6 4.1 |
| Main fuse and motor branch |
300 A Class T at the bank; 250 A MEGA on the motor branch under it |
Build Guide v9.6 4.5 |
| MOV ladder classes |
~35 / 40 / 45 / 56 V DC plain thermal-disconnect modules, no status contacts; the PV SPDs are plain too, no remote signalling |
Above the 29 V ceiling band |
| Bilge pump breakers |
5 A Type-1 auto-reset. Every small branch that can take a 24 V stud-type auto-reset breaker has one (bought 2026-09-07); blade-fuse blocks are out |
Build Guide v9.6 3.6 / 4.5 |
5. Cameras and network
| Setting |
Value |
| Network switch |
ONE switch on the boat: Teltonika TSW202 at the aft junction (bought 2026-09-07). ALWAYS ON from the house branch, never on a switched-load relay channel (shedding it would blind Starlink and cut the Pi-to-Pi link). The Netgear GS308EP and GS305EP are out; the 5-port switch in the Pi A box is removed |
| TSW202 port map |
1 to 5 the five cameras; 6 the PoE light driver; 7 Pi A; 8 Pi B; SFP 9 Starlink on a 10GTEK SFP-T copper module; SFP 10 spare. Ports 1 to 8 carry power; a copper SFP module carries no power |
| TSW202 firmware |
PIN IT. Record the version the bench test ran on and fly that exact version (Teltonika renamed a setting between 1.01 and 1.05). Version: to be recorded at the bench test |
| Port control |
Teltonika Web API turns a port off and on; read the state back over SNMP (read only) after every command, so the order out and the answer back travel different paths |
| PoE switch supply |
Mean Well DDR-120B-48 set to EXACTLY 54.0 V, trimmer sealed with paint. It stays: cameras plus lights fit with room over |
| Switch boot state |
ALL ports ON |
| Camera IR LEDs |
OFF in settings |
| Camera duty |
One watching, others slept, per-port PoE control from Pi A through the Web API |
| Interior PoE lights |
Port 6, the PoE Texas driver with two disc lights; 25 W or 12 W mode, both fit. Kept, for colour on the interior dome camera |
| Foredeck camera |
A weatherproof (IP66) 180 degree PoE fisheye, Reolink Duo 3 PoE or Duo 2 PoE class, to buy. The Reolink FE-P as bought is an indoor unit; it goes back or serves indoors |
| Underwater camera UV-C timer |
~5 min ON / 10 min OFF hardware cycle |
6. Alerts (shore)
| Alert |
Threshold |
| SBMS0 serial stream stale |
~1 h silence, escalate after 24 h (eyes lost; the pause is hardware and does not depend on the stream) |
| No-charge health check |
No charge current for one full solar day with sane bus voltage. It catches a dead SBMS0 only if the charge-continue bypass failed to close; with the bypass working, charging continues after 10 minutes and the tell-tale is the flag below |
| Charge bypass active |
Reported daily whenever relay channel 8 is closed (a silent watchman stream for over 10 minutes) |
| Group spread / excursion |
Spread threshold set at commissioning; any group above 3.45 V/cell average |
| Bilge duty trend |
Rate-of-rise on pump duty cycle |
| Watchdog |
WD-A cycles Pi A rail on 5 min heartbeat silence |
| Steering feedback disagreement |
Actuator feedback against encoder A, threshold set at commissioning (section 3) |
| Link-open (wing) |
Commanded rod position against measured wing angle disagree: the compliance link has opened under wind load; the pod reduces the tab command on repeated alerts (section 8) |
| Cell temperature trip |
Any pack probe above 150 F, or a rise of more than 10 F in five minutes on any probe: the bank isolates, the boat enters HIBERNATE and an Iridium alert goes out; release after one hour with every probe below 120 F (section 4) |
| Cell temperature warning |
Any pack probe above 131 F, the SBMS0's own charge-stop temperature: warning only, no action (section 4) |
7. Navigation (heading from the IMUs; two receivers for position only)
| Setting |
Value |
Where |
Why |
| IMU heading offset |
To be measured at setup; degrees, one value per BNO085 (three). The angle of each IMU's forward axis relative to the bow; zero if mounted square. Heading = IMU yaw corrected by this offset, fused with the track |
Boat software config file |
The ZED-F9P moving-base heading pair is deleted (Mike, 2026-09-08); the two NEO-7M receivers give position only. Measured once at setup against a known bearing |
| GPS receiver |
hiBCTR NEO-7M, one per island (bought 2026-09-09 as a 4-pack; two are Boat 2 stock). TTL serial at 3.3 to 5 V, SMA antenna. Single-band GPS with GLONASS, accurate to a few yards, which is all the boat needs |
- |
D25; Mike 2026-09-15 |
| Wind sensor primacy |
Post unit (Calypso CMI1018, UART to the central node) primary; wing-tip unit (CMI1017, RS-485 to the pod) the cross-check, resolved into the boat frame through the pivot encoder |
Boat software |
Mike, 2026-09-11. Disagreement flags icing and fouling |
| Wind motion correction |
Tilt correction from IMU A attitude on both sensors (a 2D ultrasonic tilted 20 degrees under-reads by about 6 percent, one-sided). The swinging correction needs rotation rate plus one tape-measure offset from the centre of motion to the post head: to be measured |
Boat software |
Mike, 2026-09-11 |
8. Wing trim: compliance link and exercise
The compliance link in the tab pushrod is the fail-safe, a wind-load fuse (Wing Trim Actuator Installation Guide v9.6 secs 6 and 7). It resets itself when the wind eases.
| Setting |
Value |
Where |
Why |
| Link preload |
4 lbf initial, tuned at sea from the link-open alert; at least 0.5 in of travel each way |
The link's spring caps, witness-marked |
Compliance Link Build and Setup 2026-09-15 |
| Test 1, rod locked |
Rod locked at full extension; pull the trailing edge of the tab toward straight with a hanging scale: it starts to move at about 1.5 lbf and reaches near straight |
Bench, before the wing closes and at each refit |
Replaces the old power-off spring-straight test, which is deleted |
| Test 2, hinge spring |
Pushrod disconnected; the hinge spring returns the tab to within 2 degrees of straight from both sides. The Thomson Electrak MD has a self-locking acme screw, so there is no back-drive for the spring to beat |
Bench |
The hinge spring only centers the tab; it never fights the actuator |
| Exercise sweep |
WEEKLY full-stroke sweep in very light wind, not daily; if no calm window comes in a week, a small health check instead. The MD's brushed motor and potentiometer are why the sweep matters |
Pod software |
Mike, 2026-09-08 |
| Loop rate |
A few degrees at a time, a few times a minute |
Pod software |
Wing Power System v9.6 |
| Actuator protocol |
J1939 CAN to the Thomson Electrak MD (MD24A025-0050CNO): control PGN 61184, status PGN 126720 every 100 ms, 120 ohm terminators at each end, on the pod Nucleo-G431's own CAN through the same transceiver breakout as the central node. The MAX3485 in the pod serves only the CMI1017 wind sensor. No switched feed: the MD sleeps under 2 mA |
Pod Nucleo-G431 |
D17; Mike 2026-09-15 |
Sources: v9.6 Change List 2026-09-15 (Part 6, Mike's answers of 2026-09-15 evening; D15 steering, D16 and D17 wing trim, D20 temperature trip, D21 power distribution, D22 and D23 network, D25 heading, D27 XT1), Mike's battery-system block of 2026-09-05 (SBMS0 pause, heaters deleted), Build Guide v9.6 secs 3.4, 3.6, 4.1, 4.2, 4.5, 5.1, 5.3, 8.4, Electronics and Wiring Plan v9.6 sec 6, Serial and Network Summary v9.5 (port map, firmware pin). Commissioning checklist: verify every row on the bench, initial and date it.