# Wet lab box (new box A18), 2026-10-14

Mike asked (2026-10-14): "yes, lay out the wet lab box." Claude's layout, standing unless Mike says otherwise. Sources: Flow-Through Wet Lab Design v10.2 secs 7, 10, 11, 12; Electronics and Wiring Plan v10.2 secs 4, 5, 7 and A3, A15, A15a; the order tracker as of Oct 14.

## What it is for

The wet lab runs a cycle every hour (Wet Lab Design sec 10): open valve 2 with the pump off, start the pump, flush, measure, then stop the pump and shut valve 2 together; dose every 6 hours; on a clog, shut valve 2 and blast with the Gulper. Today one relay at the central node (channel 2) switches the whole wet lab line on or off, and nothing switches the parts one at a time or reads the pressure sensor, the intake temperature probe, the flow meter or the valve position wires. This box does that.

It is a small node like Release Node B: a NUCLEO-G431KB on the **house bus**, taking orders from Pi A over CAN (the boat's data wire). Pi A runs the cycle; the node does what it is told and refuses anything unsafe.

## Where it goes

Near the wet lab, outside its drip tray and above the tray's rim, glands facing down with drip loops. Within about 5 ft of the USB hub box if it can be (the two FTDI cables are 6 ft; longer runs get extended on screw terminals, below). Nothing in it is wired to the lightning copper.

## Power

- In: the wet lab's own 16/2 line (15 A breaker at the bar, switched by central node relay channel 2), to the XWST converter's input.
- The XWST (24 V steady, 10 A, on hand) sits inside this box and feeds everything in it.
- Claude's pick: Pi A turns relay channel 2 on about a minute before each burst and off after it, so the whole wet lab, node included, draws nothing between bursts. Nothing needs power while idle: valve 2 is shut when unpowered, and valve 1 stays where it is.
- Every load gets a fuse (5 x 20 mm slow-blow, QBGTFAK kit on hand) in a DIN fuse holder off the XWST output: (1) the Gulper, 5 A (a guess; read its label); (2) the Iwaki pump and the Kamoer dosing pump together, 2 A; (3) both valves, 2 A; (4) the DO sensor and the SBE 45, 1 A; (5) the node and the flow meter, 1 A.
- A 1.5KE33CA surge diode across the XWST output.

## Inside the box

| Part | Job | Status |
|---|---|---|
| NUCLEO-G431KB on an AITRIP Nano screw carrier | The node's brain | **Buy the Nucleo** (both on hand are used: central node and Release Node B). Carriers: six bought, two used |
| SN65HVD230 CAN board | Talks on the house bus | On hand: four, two used |
| Taidacent RS-12V/2S | Surge part on the CAN drop | One of the two bought Oct 7 for the DO and SBE 45 lines, which have no job now (coming Oct 21 or later) |
| Klnuoxj 24 to 5 V converter | 5 V for the Nucleo and the pressure sensor | From the 8 on the shelf (3 used) |
| HiLetgo 24 V 8-channel relay module | Switches each wet lab part | One of the two spare modules (4 on record, 2 used at the central node). Check the shelf |
| Fafeicy 8-channel opto board (24 V inputs) | Reads the flow meter pulses and the valve position wires | On hand: five bought, one used |
| XWST 24 V 10 A converter | Wet lab power | On hand |
| 1N5408 diodes x3 | One across each pump motor at its terminals (stripe to +), so the relay contacts do not spark | On hand (CHANZON) |
| Resistors: 10k and 18k (pressure divider), two 4.7k (I2C pull-ups) | See the wiring map | MelkTemn kit on hand |
| DIN fuse holders x5 | One per fused group | **Count**: part of the DIN fuse holder shortfall in the box parts check |
| 1.5KE33CA | Surge diode on the 24 V | On hand |
| DIN rail and about 30 screw terminals | Every field wire lands on a terminal first | Rail and terminals: count with the rest |
| Membrane vent | The box breathes through it | **Count**: part of the vent shortfall |

## Relay channels (all normally open: a dead node turns everything off)

| Ch | Switches | Note |
|---|---|---|
| 1 | Iwaki pump | Node runs it only while valve 2 reads open |
| 2 | Kamoer dosing pump | Run time set on the bench for 5 mL |
| 3 | Gulper air pump | Node runs it only while valve 2 reads shut |
| 4 | Valve 1 OPEN line | Never on together with ch 5 |
| 5 | Valve 1 CLOSE line | Never on together with ch 4 |
| 6 | Valve 2 power | Powered = open; unpowered, it springs shut |
| 7, 8 | Spare | Later sensors |

Valve 1 is the stays-put 5-wire valve: read its wire colors off its label (which wire is open, close, common and the two position wires) before wiring.

## Inputs to the Nucleo

| Signal | Path | Pin type |
|---|---|---|
| Flow meter pulses | Meter powered at 24 V, pulse wire to an opto input | Digital, counted (about 11 pulses a second at 1.5 L/min, a guess from typical turbine meters) |
| Valve 1 open, valve 1 shut | Position wires through two opto inputs | Digital |
| Valve 2 open, valve 2 shut | Position wires through two opto inputs | Digital |
| Line pressure | Sensor on the node's 5 V; its 0.5 to 4.5 V output through 10k on top, 18k to ground, so the pin sees 2.9 V at most | Analog |
| Intake temperature (Blue Robotics Celsius) | Its own 3.3 V, ground, SDA, SCL from the Nucleo, 4.7k pull-ups to 3.3 V | I2C |
| CAN | SN65HVD230 | CAN |

## What the node does on its own (fail-safe)

- At power-on, no message from Pi A for 60 s: it runs one burst on its own (the solo burst, Flow-Through Wet Lab Design sec 10). In the middle of a burst, no message for 60 s: all relays off. The pump stops and valve 2 springs shut.
- Pump only with valve 2 reading open; Gulper only with valve 2 reading shut; valve 1's open and close lines never both on.
- Each valve stroke is timed from its position wires and reported: that is the wear gauge the Wet Lab Design asks for.

## Cable junction for the two science sensors

The DO sensor's cable and the SBE 45's cable each carry power and data together. Both come into this box: power lands on its fused terminal, and the data wires carry straight on to the USB hub box on the FTDI cable's bare end (RS-485 for the DO sensor, RS-232 for the SBE 45). If the run is over about 5 ft, extend it on screw terminals here with leftover 22/12 shielded cable. This settles where those cables join, which no doc said.

## Holes (14 plus a vent, all on the bottom face)

| # | Entry | Gland |
|---|---|---|
| 1 | CAN house drop (cut Micro-C cord) | PG7 |
| 2 | 24 V in, the wet lab's 16/2 | PG9 |
| 3 | Iwaki pump | PG7 |
| 4 | Kamoer dosing pump | PG7 |
| 5 | Gulper | PG9 |
| 6 | Valve 1 (5 wires) | PG9 |
| 7 | Valve 2 | PG9 |
| 8 | Flow meter | PG7 |
| 9 | Pressure sensor | PG7 |
| 10 | Celsius probe | PG7 |
| 11 | DO sensor cable in | PG7 |
| 12 | DO data out to the hub box (FTDI RS-485 cable) | PG7 |
| 13 | SBE 45 cable in | PG7 |
| 14 | SBE 45 data out to the hub box (FTDI RS-232 cable) | PG7 |
| V | Vent | M12 |

PG7 x10, PG9 x4, one vent. Read each cable's diameter against its gland before drilling (PG7 grips about 1/8 to 1/4 in, PG9 about 5/32 to 5/16 in).

## The box itself

Plastic IP67 with a mounting plate, about 12 x 8 x 5 in (Claude's estimate from the parts: relay module about 5.4 x 2.2 in, opto board, Nucleo on its carrier, the XWST, a DIN rail with five fuse holders and about 30 terminals). Claude's pick: the TICONN IP67 box bought Oct 3 ($68.39, size not in the mail, no job), if it is at least that size. That frees the second drogue box question to stay with fiberglass, as the Drogue Box Guide asks. If it is smaller, buy a TICONN box about 12 x 8 x 5 in.

## To buy

- NUCLEO-G431KB x1. Not on Amazon as far as found. Mouser: https://www.mouser.com/c/?q=NUCLEO-G431KB ; DigiKey: https://www.digikey.com/en/products/result?keywords=NUCLEO-G431KB . About $13 to $20 (seen on distributor listings 2026-10-14, not confirmed).
- An N2K single tee for the house backbone, ahead of Release Node B's end tee: count the shelf first (seven 2-packs of tees in the order log).
- Everything else: on hand or in the counts already listed in `Reference/Box Parts Missing 2026-10-14`.

## Left out on purpose, and what stands in for it

- Pump current and dosing-reservoir level (Wet Lab Design sec 11 lists both): no current sensor is fitted on the boat, and no level sensor is named. Stand-ins (Claude's pick): valve stroke time from the position wires shows valve wear; the flow rate shows the pump's health; the reservoir level is counted from the doses given.

## To write in later

Electronics and Wiring Plan: a new A18 and the house bus layout (sec 5.2, 5.4); Wet Lab Design secs 10 and 12 ("the board that reads them" becomes this box); Box Wiring Sheets: a new sheet A18; BOM 15.11; System Settings Sheet (node ID, the 60 s silence rule, relay channel 2 timing).
