Vehicle Power · Volume 4

The CPAP, Which Is the Load That Changes the Design

Introduction

UPDATED 2026-08-30 — and this volume’s argument won. 🔴 The CPAP now runs from an EcoFlow River 3 Max Plus (858 Wh), not a 200 Ah bank, and with the fridge out of the build the CPAP is the only load the electrical system exists for. This volume argued that a small dedicated pack at the tent was the right answer to the eight-feet-up problem; it turned out to be the answer for the whole rig. ResMed’s confirmed 10–25 W humidifier-and-hose-off figure is unchanged and is what makes the smaller pack viable: 3.9 nights at the top of the band. ⚠ Two new constraints: the pack’s 12 V port is 126 W, which permanently forecloses the humidifier; and LiFePO4 will not charge below freezing, so the pack cannot spend cold nights in the tent. → Weight and Balance vol 3.

Every other load in this build is discretionary. The fridge can be switched off. The radios, the gateway, the lights, the charging — all of it can be sacrificed for a night without consequence beyond inconvenience.

The CPAP is therapy, and it runs every night. That single fact changes the electrical design from an optimisation problem into a reliability problem, and it introduces a constraint none of the other loads have: it has to work eight feet up a ladder, inside a tent, disconnected from where the batteries live.

This volume covers both.

4.1 The machine, and its published numbers

The relevant combination is a ResMed AirSense 10 with the Air10 90 W DC/DC converter, model 370003 — the factory part that takes 12 V or 24 V from a vehicle socket and produces the 24 V the machine wants.

Table 1 — The machine, and its published numbers

SpecificationValueSource
AirSense 10, typical power53 W (57 VA)AS10 user guide, 90 W PSU
AirSense 10, peak power104 W (108 VA)AS10 user guide
DC/DC converter, typical51 WAir10 DC/DC converter guide
DC/DC converter, peak106 WAir10 DC/DC converter guide
Converter input12 V at 13 A, or 24 V at 6.5 AAir10 DC/DC converter guide
Converter output24 V at 3.75 AAir10 DC/DC converter guide

Those “typical” figures include the humidifier. ResMed states it directly in the converter guide: the device with the HumidAir humidifier consumes considerably more than the device alone, and using it will drain a battery faster.

Translated to a night: 51 W for eight hours is 408 Wh, about 34 Ah at 12 V — very nearly a second fridge, and the single largest overnight load in the vehicle.

4.2 The humidifier-off decision, which is the whole design

Turning the humidifier off is the difference between a battery bank that lasts one night and one that lasts two. It is worth roughly 22 Ah per night, which is more than every load in the truck except the fridge combined.

CORRECTED 2026-08-26. This volume previously recorded that “ResMed does not publish a humidifier-off figure.” That was wrong. ResMed give the AirSense 10 at 10–25 W — roughly 1–2 A at 12 V — with the humidifier and heated tube off, which works out to 6.7–16.7 Ah across an eight-hour night.

The estimate this dive had been carrying — 10 to 20 W, 7 to 14 Ah — was good. It sat in the middle of the published band and was slightly optimistic at the top.

The band is wide because prescribed pressure drives it. Ten watts and twenty-five watts are the same machine at different settings, so the remaining question is not what does an AirSense 10 draw but where in ResMed’s band does this prescription sit. That is a much smaller question than the one this dive opened with.

🔴 Measure it. It is the one number in this whole electrical design with real uncertainty, and everything else is sized around it.

The measurement is straightforward and the worksheet already exists: the CPAP project’s AS10-BENCH-SHEET.html in Tasks/CPAP/ was built to capture exactly this class of hardware-blocked measurement. What this project needs added to it:

  • Steady-state current at 12 V, humidifier off, at the actual prescribed pressure — not at a default.
  • The same with humidifier on, for comparison and to validate ResMed’s 51 W typical.
  • Startup surge, because it determines fuse rating rather than energy.
  • Total watt-hours across a full night, which is the number that sizes the bank.

Humidifier off is not a hardship here. The owner already turns it off most of the time, and in a rooftop tent in the Upper Peninsula the ambient humidity is generally high enough that the case for heated humidification is weaker than it is in a dry bedroom in February. A heat-moisture-exchange filter is the low-tech substitute and draws nothing at all.

4.3 The 13-amp number, which is a wiring decision

The converter’s input rating is 12 V at 13 A. That is a peak figure rather than a running one, but it is what the circuit feeding it must tolerate.

A shared 12 V accessory socket is not adequate. Factory sockets are commonly on a 10 to 15 A fuse shared with other loads, and a 13 A peak on a shared circuit is exactly the arrangement that produces intermittent, hard-to-diagnose faults — and this is the one load in the vehicle where an intermittent fault at 2 a.m. is unacceptable.

The CPAP gets its own circuit: its own fuse sized for the converter’s peak, its own adequately-sized conductor, and a socket or connector rated for the current. This is a small addition to the distribution block described in volume three and it should be planned in from the start rather than added later.

4.4 🔴 The connector hazard, which is specific and easy to get wrong

The Air10 family uses a three-contact DC plug — an HP-style connector around 7.4 / 5.0 mm — and the centre pin is widely reported by the community to be an identification or “power-good” sense line sitting at roughly 3.3 V. This behaviour appears in no ResMed document. Two independent community sources, including a cord-building discussion, describe it as a sense pin and state that a plain two-wire feed will not run the machine.

Two consequences:

Use the genuine ResMed Air10 DC/DC converter, model 370003. It is the factory part, it handles whatever the connector requires, and it is the reason a “12 V power cord” either works reliably or does not. A generic 12 V-to-24 V cord that ignores the sense pin may run, may run intermittently, or may not run at all — and the failure will be discovered at a campsite.

Do not assume the pin exists, and do not assume it does not. 🔴 If any cord or adapter is ever built for this machine, meter the plug before cutting it — once the plug is off the brick, what it presented on its contacts is unrecoverable. That warning, and the reasoning behind the 5S-then-boost pack architecture, is already worked out in the CPAP project’s build documents and should not be re-derived here.

Note also that the AirSense 11 and AirMini use a different, incompatible plug. A cord that works on one is not a spare for the other.

4.5 The problem none of the other loads have: it runs in the tent

The batteries live in the cab or the bed. The CPAP runs in a rooftop tent, up a ladder, on top of a 26-inch rack (⚠ the awning uprights do not add 18 in — see Bed Rack vol 1) with awning uprights beside it. This is a genuine installation problem and it deserves a real answer rather than an extension cord.

Three approaches:

Run a dedicated 12 V feed up to the tent. A properly-sized, fused conductor from the distribution block, up a rack upright, into the tent, terminating in a connector at the mattress. It is permanent, it is powered by the main bank, and it is the tidiest answer. The costs: the cable has to survive the tent opening and closing every single time, it has to be routed where a fold-out section will not pinch it, and a 13 A peak over that distance wants real copper.

Put a dedicated battery in the tent. A small LiFePO4 pack or a modest portable station — a few hundred watt-hours, ten to twenty pounds — lives with the CPAP, is carried down and recharged from the truck during the day, and is completely independent of everything else. This is the strongest option for a medical device, because the failure of any part of the vehicle’s electrical system does not reach it.

Both. A tent feed for normal use, and a pack as the fallback. For a device where failure means no therapy, this is not over-engineering.

The recommendation is the dedicated pack, with a tent feed added if the routing proves clean. It is also the option that connects directly to work already done: the CPAP project’s own battery-pack build — the 5S-then-boost architecture, the harvested OEM plug standardised to a generic 5.5 × 2.5 mm mating pair, the DeWalt 20 V-max pack as an off-the-shelf 90 Wh cell stack with its own BMS and charger — is precisely this device, and it was costed at roughly $80–105 on top of a tool battery against $339 for a commercial CPAP battery.

That project and this one have converged. What was a bench exercise is now a load-bearing part of the overlanding electrical design.

4.6 Redundancy, because it is a medical device

The design rule from volume one, restated as practice:

  1. Reserve one night of CPAP capacity explicitly. A monitor reading 30 Ah remaining means one more night of therapy and nothing else — not “30 Ah of margin”.
  2. The humidifier stays off in the field by default, and the HME filter is the substitute.
  3. There is an independent fallback — the dedicated pack — that does not share a failure mode with the truck’s electrical system.
  4. The truck itself is the last resort. Idling the engine to charge is inefficient and unpleasant, and it is a real option that exists, which is worth knowing at 2 a.m.
  5. Know the actual numbers. The difference between “I think there’s enough” and “there is 40 Ah, which is three nights humidifier-off” is a shunt-based battery monitor, and for this load that monitor stops being a nicety.

4.7 What to record

  • The measured humidifier-off draw at the prescribed pressure, at 12 V. ✅ No longer a blocker — ResMed’s published 10–25 W band bounds it, and the numbers here sit inside that band. The bench measurement now refines rather than unblocks, telling this build which end of the band it lives at.
  • The measured humidifier-on draw, to check ResMed’s 51 W typical against this machine.
  • The startup surge, for fuse sizing.
  • A full night’s watt-hours from the shunt, with the fridge running too.
  • Which cord is actually in use, and whether it is the genuine model 370003 converter.
  • How the tent power was solved, photographed, because it is the detail nobody documents.

The CPAP itself — clinical menus, settings, the connector investigation, and the battery-pack build — lives in the CPAP project at Tasks/CPAP/, including six ResMed primary PDFs with text extractions in refs/. Read those before searching the web; the numbers in this volume came from them. The distribution, fusing, and connector practice this load plugs into is volume three of this dive.

References

  • ResMed — AirSense 10 user guide (Americas), 90 W power supply specifications. Local: Tasks/CPAP/refs/as10-user-guide-amer.txt.
  • ResMed — Air10 90 W DC/DC converter guide, model 370003. Local: Tasks/CPAP/refs/air10-dc.txt.
  • ResMed — battery guide, for the general approach of sizing with a contingency margin.
  • Community reporting on the Air10 three-contact DC plug and its centre sense pin — not corroborated by any ResMed document; treat as unverified until metered.

Comments (0)

  1. Loading…

Comments are held for moderation — nothing appears until approved.