Cost, When It Wins, the Bench and the Decisions

5.1 What It Costs

Prices from the seller’s page on 19 October 2026 unless stated. Texas Instruments’ product pages showed no price when read, and Mill-Max publishes none; those rows say so rather than guess.

Table 1 — 5.1 What It Costs

Item	Pick	Price, seller	Where decided
6S pack, two	GNB 6S 450 mAh 80C long type, XT30	2 × $24.99, RaceDayQuads	LB-2
Balance charger	SkyRC B6neo	$39.99, Wrekd	LB-5
Charger supply	The wired gun’s GST60A24-P1J brick	$19.40 (DigiKey) to $24.86 (TRC), 2026-10-18 (Power Supplies Vol 2); nothing extra if the wired gun exists	LB-5
Pack protector	2 × TI BQ77905 and their FETs	at order time	LB-3
Rail switch	TI LM5060 and a 60 V N-FET	at order time	LB-4, WP-2
Fuse	DC-rated, valued after the LiPo dive’s test B2	at order time	LiPo Batteries Vol 4
USB lead	L-com CAUABHFX-3M, high-flex, 3 m	$40.29, L-com	WP-4
Breakaway, optional	Mill-Max Maxnetic 4-position pair	at order time	WP-6
Bench breakaway trial	Adafruit 5413 DIY magnetic connector	$6.95, Adafruit	Volume 2
Battery magazine	Printed, with clamp, boss and the service connector	filament; the service connector is the printed build’s	Volume 1
Priced total, without the optional items		about $130, plus the TI parts and fuse

Against the wired gun. Fully Wired Vol 6 puts the whole wired setup, brick included, at under about $40. The battery magazine and its lead cost about three times that, and two-thirds of it is the packs and the charger. The lead is the single most expensive line only because the proposal buys a flex-rated one; a generic marked 28/24 lead would cut about $30 and the flex rating with it.

Against the fully wireless gun. The same packs, charger and protection, plus a radio and a receiver, minus the lead. The Fully Wireless dive prices the radio; whatever it costs, it is added to this one’s total, not substituted for it.

In a kit. A kit ships no pack (LB-7), so a kit’s battery magazine is the printed parts, the protection board, the rail switch and the lead; the buyer brings the pack and charger. The LiPo dive’s shipping reasons apply unchanged (LiPo Batteries Vol 7).

5.2 The Three Architectures Side by Side

Table 2 — 5.2 The Three Architectures Side by Side

	Fully wired	Wireless power only	Fully wireless
Cable in play	Umbilical, USB + 4 × 20 AWG in a sleeve	USB lead, 4.5 mm	None
Far-end hardware	Brick, junction box, wall outlet	None	Dongle in a USB port
Latency	USB, 1 ms poll	Identical	USB plus the radio link
Recoil limit	Brick trip band (D16, MR-6)	Coil heat only	Coil heat only
Fault current	Brick-limited	Pack: fuse and protector mandatory	Pack: the same
Runtime, busy play	Unlimited	About 9.5 h per 450 mAh pack	About 5.6 h per pack (logic on the pack)
Firmware change	None	None (logic on VBUS)	Radio firmware; the D25/D30 trial
USB power rules	Bus-powered	Bus-powered	Self-powered when the USB lead is used for configuration (Volume 3)
Plug order	Fixed (D31)	None	None
Charging	—	Out of the gun (LB-5)	Out of the gun (LB-5)
Priced cost over the gun	under about $40	about $130 + TI parts	this, plus the radio, less the lead

5.3 When It Beats Fully Wired

• When the umbilical is the complaint. The wired gun’s tether is a sleeve over two cables, one of them 7.6 mm of 20 AWG power cable (Volume 2). The lead here is a 4.5 mm USB cable. A player swinging the gun feels the difference at the wrist.
• When there is no outlet at the gun’s end of the room, or the brick is the clutter. The wired gun needs a brick and a junction box near the PC; this gun needs a USB port.
• When autofire must be sustained. On a brick, one stage at a time and the trip band set the limit (D16, MR-6). On a pack, only the coil’s heat does (Solenoid Recoil Vol 4). A game that holds the trigger for a long burst kicks harder for longer on the pack.
• When “the gun disconnects when it fires” will not go away. The pack removes the shared-return mechanism entirely (Volume 3). A gun that resets only on the brick and never on the pack points straight at the brick’s wiring.
• When the gun is shown or carried. No plug order, no brick, no box: plug the lead into a PC and play.

5.4 When It Beats Fully Wireless

• Latency and reliability. The data path is the wired gun’s. There is no radio, pairing, interference or dongle, and the D25 ESP32-S3 trial has not yet said whether a radio link is good enough.
• No firmware change. The private fork stays one firmware (D10, D30).
• Runtime. With the logic on USB, the same pack lasts about 1.7 times as long in busy play (Volume 4).
• USB rules kept by construction. A fully wireless gun that also has a USB port for configuration is self-powered and must gate its D+ pull-up (Volume 3); this one need not.

5.5 When to Stay Wired

• The first build. D35 and D17 put the wired printed MP7 first, and nothing here argues otherwise: the wired gun’s brick limits the current of a stage that has not yet been proven, and a pack does not.
• When the magazine will not hold the pack. Everything in this dive assumes the pack fits the magazine (LB-6), which waits on the magazine scan’s measurement. If it does not fit, the pack goes elsewhere in the shell and the battery magazine stops being a self-contained module; the one-source-per-rail rule (WP-3) must then be kept another way, such as the LM74700 pair of Volume 3.
• When a pedal is wanted. The pedal’s two lines ride the wired umbilical (CI-7); the USB lead carries none (Volume 2).
• When nobody wants to manage lithium packs. Charging, storage charge, fire-safe storage and inspection are the LiPo dive’s whole Volume 6. A wired gun has none of it.

5.6 The Bench Programme

Each test uses the lab’s programmable supplies and the Analog Discovery Pro ADP3450 (D9). Run the LiPo dive’s stage-1 bench first, with a programmable supply standing in for the pack (LiPo Batteries Vol 7).

Table 3 — 5.6 The Bench Programme

#	Test	Set-up	Pass
W1	Rail follows the lead	Programmable supply as the pack, LM5060 board with EN from VBUS; ADP3450 on VBUS, EN, GATE and V_SOL (×10)	V_SOL ramps after VBUS within the LM5060’s soft start, and falls when the lead is pulled; no V_SOL without VBUS
W2	Standby current	The same, lead out; DMM in series with the supply in the microamp range	About 21 µA typical, under 30 µA; against about 4.6 mA with EN forced high
W3	Fault latch clears on re-plug	Force a V_DS fault (a short through a small resistor on V_SOL after the switch)	GATE latches off; pulling and replacing the lead restores V_SOL
W4	Ground offset during autofire	Gun on the lead from the PC that runs the ADP3450; CH1 Pico GND, CH2 host USB shell, math CH1−CH2; coil current on the 0.1 Ω bench shunt	A flat line at the DC offset of Volume 2 (about 0.1 V at 3 m of 28 AWG); no step at the shot
W5	Lead flex	The lead in a printed copy of the magazine exit, swung ±90° for a few thousand cycles; continuity of the four conductors on the ADP3450 logic channels	No dropout of 1 µs or longer
W6	Breakaway (if fitted)	Maxnetic housing in the lead near the PC; D+ and D− on two channels at 100 MS/s while tapped, twisted and pulled; pull force on a luggage scale	No SE0 of 2 µs or more in use; separates at 4 to 6 N, before the USB-A plug moves
W7	No back-drive (Option B only)	Pico on the buck with the lead plugged into a powered-off PC; D+ to ground at the host end	With the fork’s GP24 gating, D+ falls to 0 V within 10 s of VBUS going; without it, it sits near 3.3 V through 1.5 kΩ (the driver behaviour of Volume 3)
W8	Rise time across the discharge	Programmable supply at 25.2, 22.2 and 19.8 V; IPROPI captured for one shot at each	The rise lengthens by about a fifth at 19.8 V against 24 V; decide whether it is felt
W9	A logged session	The pack in the gun for an evening; V_SOL once a second, shot count from the debug output	Replaces Volume 4’s illustrative shot rates with measured ones

5.7 Decision Rows Proposed

For DECISIONS.md, all Proposed:

Table 4 — For DECISIONS.md, all Proposed

#	Part / choice	Candidates	Pick	Why, and why not the others	Vol
WP-1	Logic supply in the wireless-power-only gun	VBUS (bus-powered); pack via buck + ORing (self-powered); pack only	VBUS; the pack feeds V_SOL only	Keeps USB 2.0 §7.2.1’s VBUS and pull-up rules by construction; no firmware change; busy runtime 9.5 h vs 5.6 h. The pack route needs GP24 gating of the D+ pull-up (TinyUSB’s RP2040 driver forces VBUS-detect on), a power switch, and drains the pack; worth it only on 100 mA host ports	3
WP-2	Rail-switch enable	always on; mechanical switch; GPIO; VBUS	LM5060 EN from VBUS through 10 kΩ, 100 kΩ to ground	~21 µA standby vs ~0.10 W (flat in 3.3 days); no switch to forget; re-plug clears a latched fault; rail cannot precede the logic. GPIO: no spare pin (FW-6)	3, 4
WP-3	V_SOL sources	brick and pack in parallel; back-to-back FETs; LM74700-Q1 ideal-diode pair; one per magazine	One source per magazine: the battery magazine has no brick input	Parallel: the brick charges the pack at up to 2.5 A (5.6C) through the LM5060 FET or its body diode. Back-to-back: still conducts when on. LM74700 pair: correct, kept for a magazine that must take both	1, 3
WP-4	USB lead	unmarked thin; 30–32 AWG power; coiled; generic 28/24; high-flex 28/24	3 m high-flex USB 2.0, 28 AWG pair, 24 AWG power, PU, 4.5 mm: L-com CAUABHFX-3M ($40.29); BU2HF-2824BK bulk for a kit	41 mV ground offset at 3 m; published construction and a 10,000-cycle claim. 30/32 AWG exceed USB’s 125 mV ground budget at 2.5 m / 1.5 m at 150 mA; coils pull the gun	2
WP-5	Gun-end termination	GX16 wired USB-only; USB-C receptacle; captive	Captive in the battery magazine, clamped on a boss, terminated in the service connector	Nothing to mate or wrench where the gun kicks; the magazine is the detachable unit. A GX16 socket would accept the 24 V umbilical; USB-C waits for Stage 2	2
WP-6	Breakaway	none; at the gun; mid-lead; near the PC; Adafruit 5521 / DIY pogo; Mill-Max Maxnetic	None for the first build; if fitted, a Maxnetic 4-position pair near the PC	At the gun it must hold the plug through every shot (~2.9 N needed vs 4.45 N minimum, unmeasured acceleration) and any 2 µs dropout disconnects USB. Near the PC it releases before a 10 N USB-A plug. Adafruit parts publish no force or cycle figure	2
WP-7	Charging from USB while playing	yes, VBUS boost charger; yes, USB-C PD (BQ25756); no	No: out of the gun (LB-5)	Energy suffices (+0.7 W net on USB 2.0 in busy play), but 6S needs a boost and balancing in PETG, the charger must obey USB’s 100 mA and suspend limits, and it opens the low-side sneak path; 9.5 h per pack and a magazine swap make it unnecessary	4
WP-8	Logic ORing parts (Option B, fully wireless, Stage 2)	Schottky; DMG2305UX P-FET; LM66100; TPS2116	LM66100 on a Pico module; TPS2116 priority mux on the Stage 2 board	Schottky drops 0.3 V and shares the load; the P-FET gives VBUS priority, the opposite of the low-power-port case	3

5.8 Questions for the Builders

1. Adopt the battery magazine as the power module (WP-3)? It depends on the magazine scan’s outside width (LB-6) and closes LB-6 against MR-7 for the battery gun.
2. The rumble stage’s standing load. MR-3’s IN1-high wiring keeps the DRV8251A awake at about 67 mW on the pack (Volume 4). Is the rumble motor kept in the battery gun, and if so, is it worth finding a pin to let that stage sleep?
3. Flex-rated lead or generic first (WP-4)? $40.29 for the L-com lead against a few dollars for a marked 28/24 cable, upgraded if it fails at the exit.
4. A breakaway (WP-6)? Is the gun played next to a console or laptop that could be pulled off a shelf?
5. Low-power hosts. Will the battery gun ever be played from a handheld or a battery-powered computer whose ports may supply only 100 mA? If so, Option B (WP-1) becomes the design, with its fork change.

5.9 What Could Not Be Verified

• The gun’s peak acceleration at the solenoid’s stop. The breakaway argument (Volume 2) rests on an estimate of a few g to about 70 g.
• L-com’s 10,000-cycle flex rating has no stated test method; “about 1,000 cycles” for standard cable is L-com’s own comparison.
• Prices for the BQ77905, LM5060, LM66100, TPS2116, LM74700-Q1 and DMG2305UX, and for Mill-Max Maxnetic parts: none shown on the makers’ pages; distributor pages refused automated reading.
• The TinyUSB copy in the pinned 4.5.4-tusb core was not read line by line; the master branches of TinyUSB and Adafruit’s Arduino port both force VBUS-detect on.
• The logic’s current (100 to 150 mA) is Power Supplies’ estimate, not a measurement; the kit camera, OLED and IMU are unmeasured.
• The magnetic connectors’ behaviour under shock: no maker found publishes a discontinuity figure.
• The enable divider values (10 kΩ, 100 kΩ) are a starting point, not a data-sheet calculation.
• Adafruit 5521’s separation force and current per pin: not published.