Start with a specification that appears, almost verbatim, on a dozen retailer pages for the same product. The Dometic RTX 2000 is described as a 12V DC unit delivering 2000W, equivalent to 6,824 BTU. Elsewhere on the same page you’ll be told it draws only 19 amps in Eco mode, with a broader range of 10 to 58 amps.
Read those two sentences quickly and you come away believing you can get 2,000 watts of something for 19 amps of something else. A forum contributor put the problem plainly: electrical watts and thermal cooling watts are not the same thing, and the product description is confusing at best.
The 2,000W is cooling output. The 19A is electrical input. Neither figure is wrong. But they are also not measured at the same moment, and that second problem is the one that matters, because it runs through this entire product category and almost nobody names it.
The core problem: two numbers, two operating points
Every one of these units uses a variable-speed compressor. That is the whole engineering story — the compressor modulates its output rather than switching fully on and fully off. It is a genuine advantage and it is why the category exists.
It also means the unit has no single operating point. It has a range.
The headline BTU figure is measured at or near maximum output. The headline amp figure is measured in the lowest-output mode the unit offers. Divide one by the other and you produce an efficiency number that the machine does not achieve at any moment of its operating life.
Do the arithmetic on the RTX 2000 as marketed: 6,824 BTU divided by 19 amps at nominal 13V gives about 247W of input, which works out to 27.6 BTU per watt. That would be an extraordinary result — roughly double the rated efficiency of a good residential mini-split at design conditions, achieved by a unit six inches tall sitting in direct sun on a vehicle roof.
It isn’t real. The unit produces its 6,824 BTU somewhere much closer to the top of its 10 to 58 amp range. At 58A and 13V that’s roughly 754W, which gives about 9 BTU per watt. That is a perfectly respectable number. It is also about a third of the number the marketing arithmetic implies.
Your sanity check for any DC air conditioner spec
Take the maximum cooling figure. Take the maximum current draw. Multiply the current by 13V for a 12V system (batteries sit above nominal under charge and below it under heavy load; 13V is a fair working average). Divide BTU by watts.
If the result is much above 12 BTU per watt at full output, be sceptical. Rooftop units operate in hostile conditions — high ambient, direct sun, short duct runs, tiny condenser coils. Twelve is roughly what good equipment achieves at rated conditions. Numbers substantially above that are either measured in a laboratory at mild ambient, quoted from a reduced-output mode, or optimistic.
The corollary is more useful still: if a specification quotes current without stating the ambient temperature it was measured at, the figure is not information. A compressor’s draw at 75°F ambient and 100°F ambient differ enormously, because the work of rejecting heat outdoors gets harder as outdoor air gets hotter. This is precisely when you need the machine and precisely when it costs you the most.
Credit where it’s due: some manufacturers do state the ambient. Nomadic’s S1 mini-split is specified at 30 to 31 amps in Eco mode at 85°F ambient and 47 to 50 amps in Strong mode, with a rated power of 960W and a 100A fuse. That’s a usable specification. You can plan a system around it. Most of the category doesn’t publish to that standard.
What the units actually draw
Figures below are manufacturer-published. Watts are calculated at 13V for 12V configurations. The BTU-per-watt column uses maximum cooling against maximum draw — the honest comparison point.
| Unit | Cooling (BTU) | Eco draw | Max draw | Max input (W) | BTU/W at max |
|---|---|---|---|---|---|
| Dometic RTX 2000 | 6,824 | 19A | 58A | ~754W | ~9.0 |
| Nomadic Cooling X3 | 11,830 | 35–55A | 65–105A | ~1,365W | ~8.7 |
| Nomadic S1 (mini-split) | — | 30–31A @ 85°F | 47–50A | ~650W | — |
| Mabru RV 12000 | ~10,000–12,000 | ~21–22A | not published | — | — |
| Typical 120V rooftop, 13.5k | 13,500 | n/a (single speed) | 11–13A @ 120V | ~1,450W | ~9.3 |
The Nomadic X3 is published at 11,830 BTU with Eco draw of 35 to 55 amps and Max mode between 65 and 105 amps. Note that the top of that range, on a 12V system, is over 100 amps of continuous current — the 12V configuration ships with a 125A fuse. That is a serious cable run and a serious fusing problem, and it is the strongest practical argument for buying the 24V or 48V version of any of these units if your system supports it.
The Mabru row has gaps because the maximum draw isn’t consistently published. What is published, by a retailer, is the claim that the unit uses roughly the same power as the Dometic RTX 2000 while producing nearly twice the cooling capacity. A professional van builder’s response to that class of claim is worth repeating: Mabru looks best on paper, but the numbers would represent genuinely new ground in air conditioning if accurate.
That is the correct posture. Not “it’s a lie” — refrigeration engineering does improve, and marine-derived compressor design is a real thing. But a doubling of efficiency over a competitor’s current product, in a mature field, announced on a retail product page rather than in a test report, is a claim that should be independently verified before you spend $3,000 on it.
The comparison that actually matters: DC versus inverter
Here is where the category’s marketing does the most damage, and where the truth is more modest and more interesting than either side admits.
A conventional 13,500 BTU rooftop unit draws between 1,300 and 1,600 watts running. Run it off a battery through an inverter and you pay a conversion penalty — real-world DC-to-AC conversion with a quality inverter lands around 87%.
120V rooftop, from battery:
1,450W AC-side ÷ 0.87 = 1,667W battery-side
13,500 BTU ÷ 1,667W = 8.1 BTU per watt
DC rooftop, from battery:
754W battery-side (no conversion)
6,824 BTU ÷ 754W = 9.0 BTU per wattThe DC unit is roughly 11% more efficient at the battery terminals. Not double. Not triple. Around eleven percent.
That figure surprises people who have absorbed the marketing, and it should reframe the purchase. Eleven percent, on its own, does not justify a $2,500 to $3,500 premium over a conventional unit and an inverter you probably want anyway for other loads. The same builder forum makes the point directly: the price of a conventional unit plus an inverter is comparable to a DC unit, and most people spending that much already want a 120V system.
If eleven percent were the whole case, this category would not exist.
What you’re actually buying
It isn’t efficiency. Once you strip out the specification-sheet arithmetic, the real advantages are these, and several of them are worth real money:
Variable-speed modulation. A single-speed compressor cycles hard on and hard off. A modulating one holds temperature at partial output. Over a full night this is a genuine energy saving that doesn’t show up in any peak-draw comparison, and in mild conditions it’s the difference between a unit that sips and one that gulps. It is the strongest technical argument in the category.
Near-zero startup surge. No 3,000W spike, so no soft starter, no oversized inverter, no tripping breakers when the compressor kicks in at 2am. If you were going to buy a 3,000W inverter purely to start an air conditioner, some of the DC premium is offset immediately.
Noise. Consistently and substantially quieter, because the compressor rarely runs at full speed and there’s no hard start. In a vehicle you sleep in, this is not a minor consideration.
Roof height and footprint. Low-profile units free up roof area for panels, which — given how much array you need to support cooling at all — is more consequential than it sounds.
No inverter in the loop. Fewer components, fewer failure points, less standby draw.
What you’re giving up: cooling capacity per dollar, straightforwardly. A $900 conventional unit produces 13,500 BTU. A $3,500 DC unit produces 6,824. If the goal is cooling a 30-foot trailer in Texas in August, the conventional unit wins on raw output and the DC unit will not close the gap.
The rebadging question
One more thing you should know before spending. A builder with multiple DC units across a fleet noted that several branded products in this category are off-the-shelf units sourced from Chinese marketplaces, and that influencer marketing drove much of the category’s growth.
We can’t verify that claim unit by unit, and we’re not going to assert it about any specific brand. But it should shape two things about how you buy:
Warranty and support are the product. If the compressor and control board are commodity parts, then what you’re actually paying the premium for is a company that answers the phone in year three and can send you a board. Ask, before you buy, what the warranty covers, how long it runs, and whether spares are stocked domestically.
Test data beats spec sheets. Some retailers now publish side-by-side heat-chamber comparisons rather than reprinting manufacturer figures. Those are worth ten specification pages. Where a vendor won’t publish or point to independent testing, treat the numbers as marketing.
Operating limits worth reading
The RTX 2000 is rated for ambient temperatures between 41 and 126°F, and retailers recommend a battery bank of at least 180 amp-hours.
Run that recommendation through the arithmetic. A 180Ah 12V lithium bank holds about 2.3 kWh usable. At Eco draw of roughly 247W, that’s around nine hours — which is presumably where the “all night” claims originate. At the top of the range, around 754W, the same bank gives you three hours.
Which figure applies to you depends entirely on the outside temperature, and outside temperature is the variable nobody controls. Plan on the pessimistic number. Working out what the rest of your loads do to that same bank is the point of a proper load calculation, and it’s worth doing before the air conditioner arrives rather than after.
Measure your own unit, because the spec sheet isn’t about your rig
Everything above concerns published figures. What your unit draws in your vehicle is a different number, and it’s the only one that matters for planning.
The measurement is straightforward if you have a shunt-based battery monitor, which you should. Note the state-of-charge or amp-hour counter at bedtime and again at wake-up, and divide by the hours elapsed. That gives you average consumption across the whole night, duty cycle already baked in — which is precisely the figure vendors can’t give you because it depends on your insulation, your parking spot and your climate.
Do it on three nights: a mild one, a typical one, and the hottest one you experience. Those three numbers define your operating envelope far better than any specification.
If you want instantaneous draw rather than the average, watch the monitor’s amps reading with everything else switched off. You’ll see the unit ramp up, hold, and modulate down as the cabin cools. The spread between the top and bottom of that range is the variable-speed behaviour you paid for, and seeing it is the clearest confirmation that the compressor is doing what it should.
Two habits worth adopting. First, log the ambient temperature alongside each reading — without it the number is as meaningless as the manufacturer’s unqualified figures. Second, record whether the unit was fighting direct sun. A rooftop condenser in full afternoon sun is rejecting heat into air far hotter than the shade temperature your weather app reports, and the consumption difference between shaded and unshaded parking is routinely 20% or more.
The verdict
Buy a DC air conditioner if you want quiet, modulated cooling in a low-profile package without building your electrical system around a startup surge — and if you’ve accepted that you’re paying a premium for those properties rather than for a large efficiency gain.
Don’t buy one because a specification implies you’ll get 6,824 BTU for 19 amps. You won’t, at any ambient temperature where you’d want the unit switched on.
And before spending on either type, look hard at the thermal envelope. Every BTU you don’t let into the vehicle is a BTU you don’t have to pay a compressor to remove, and insulation and ventilation are where the cheap wins live. The full energy picture for battery-powered cooling — bank sizing, recharge deficit, system voltage — is covered in detail separately, and it’s the calculation that should come before the product choice, not after.
Air conditioning is also, in cost-of-ownership terms, one of the largest discretionary line items you can add to a rig once you count the batteries, array and cabling it drags along behind it. If you’re building a budget, it belongs in the five-year total, not the accessories column.
Frequently asked questions
Is a 12V air conditioner more efficient than a 120V one? At the battery terminals, by roughly 10–12% once inverter losses are accounted for. Variable-speed operation adds further real-world savings over a full night. It is not the two-to-three-fold advantage the specification arithmetic suggests.
How many amps does a 12V air conditioner draw? Depending on unit and output mode, anywhere from about 10A to over 100A. Eco-mode figures in the 19–35A range are the ones quoted in marketing; maximum figures are two to three times higher and are what you’ll see on a hot afternoon.
Can I run a 12V air conditioner from a 100Ah battery? Briefly. A 100Ah lithium battery holds around 1.3 kWh usable — roughly five hours at Eco draw in mild conditions, or well under two hours at higher output. Manufacturers typically recommend 180Ah as a floor, and that recommendation is optimistic.
Do these units work above 100°F? Most are rated to operate well above that, but rated operation and adequate cooling are different things. Output falls and consumption rises as ambient climbs. A unit that holds a comfortable cabin at 85°F may simply hold the line at 105°F.
Should I buy the 12V, 24V or 48V version? Whichever matches your system, but if you have a genuine choice, take the higher voltage. Drawing 100A continuously through 12V cabling is expensive in copper and unforgiving of poor terminations. The same power at 48V is a quarter of the current.
Are DC mini-splits better than DC rooftop units? They free the roof for solar and often reach lower interior temperatures, at the cost of a more involved installation and an exterior condenser that has to go somewhere. For builds where roof area is the binding constraint, they’re worth serious consideration.