A 600W 24V switching power supply pulls 25A at full load. That number is the whole problem. It is also the number most spec sheets bury somewhere below the marketing copy, and it is where most outdoor LED wall projects quietly go wrong.
Most of the sizing advice floating around treats wattage as a single figure you match against a load total. It isn’t. A 600W unit behaves differently from two 300W units in ways that have nothing to do with the wattage — thermal behaviour, fault domain, inrush current, and the cost of the copper feeding it all shift. This piece works through those differences with real numbers, and flags the places where a datasheet is probably not telling you the whole story.
The case for a single large unit is straightforward. One CPS-600-xx measures 199 × 99 × 50 mm. Two 300W units in the same range would occupy more total volume, need two AC feeds, two mounting positions, and two sets of cable runs. If you are feeding a dense LED wall from a single cabinet, one 600W unit is genuinely tidier.
The case against is less obvious but usually more important: fault domain. A single 600W unit is one point of failure for the entire load it serves. When it goes, everything downstream goes dark at once. Split the same load across two 300W units and a single failure costs you half the wall, and — more usefully — tells you exactly which half to look at.
So the honest test isn’t “how many watts do I need.” It’s “how much of the installation am I willing to lose at once, and how fast can someone get to it?”
Where 600W genuinely wins:
Where it loses:
That last one deserves emphasis. A 600W supply running at 90% efficiency dissipates roughly 67W as heat. Not 6W. Sixty-seven. At 199 × 99 × 50 mm, the total external surface area is about 0.069 m² — which puts you near 970 W/m² of heat flux if the enclosure has no forced airflow. That is a lot of heat in a very small box, and it is the single most common reason these units fail early in the field.
Most 600W sizing conversations skip straight past this, and it is usually the thing that kills the design.
600W at 24V is 25A. Push 25A down a cable and the cable resists. What you lose is voltage, and your LED modules see that loss as dimming at the far end of the run.
Copper resistivity is roughly 0.0175 Ω·mm²/m at room temperature — and it rises with temperature, so treat that figure as optimistic. For a 10m run you also have to count the return path, so the effective conductor length is 20m.
Take a common 2.5mm² cable:
That is not a workable number. Most LED systems want drop held to around 5% or better — about 1.2V on a 24V system.
To hit 5% at 25A over 10m:
So a single 600W feed over ten metres wants roughly 7mm² of copper — call it 8 AWG territory. Expensive, stiff, and awkward to terminate.
Now split the same load into two 300W supplies, each placed near its own half of the wall. Each carries 12.5A, and the runs get shorter as well as lighter. At 4mm² over 10m:
Under 5%, with less than half the copper.
This is the strongest practical argument against a single large supply on a spread-out load — and it has nothing to do with the supply. It’s Ohm’s law. A 600W unit is the right answer for a dense load close to the cabinet and the wrong answer for a load spread across ten metres, regardless of what the wattage math says.
A 600W supply at 90% efficiency draws about 667W from the mains and turns roughly 67W into heat. That heat has to leave through the case.
The CPS-600-xx case is 199 × 99 × 50 mm. Its total external surface area:
67W spread over 0.069 m² is about 970 W/m² of heat flux with no airflow.
For comparison, run the same arithmetic on a CPS-250-xx — 160 × 98 × 30 mm, about 0.047 m² of surface, roughly 28W dissipated at 90% efficiency. That’s about 594 W/m².
So the 600W unit concentrates roughly 1.6 times the heat flux of the 250W unit in the same kind of enclosure. Not unmanageable — but it means the thermal path has to be designed rather than assumed, and it’s why the “just use a bigger supply” instinct tends to produce early failures.
Here is the thing about “600W” on a nameplate: that rating assumes a specific ambient temperature, usually somewhere around 25–40°C, with a specific mounting orientation and unrestricted airflow. Inside a sealed outdoor cabinet in July, none of those assumptions hold.
The derating behaviour is fairly consistent across this class of unit. Output capability typically holds flat to somewhere around 40–45°C ambient, then falls progressively — commonly to roughly 80% of nameplate by 50°C, and lower still as you approach 60°C. Put plainly: a 600W unit in a hot cabinet may deliver 480W, or less.
Why does this matter more at 600W than at 150W? Because the absolute margin is bigger. Losing 20% of a 150W supply means losing 30W, which you probably had headroom for anyway. Losing 20% of a 600W supply is 120W — that is the difference between a design that works and one that browns out on the first hot afternoon.
Two practical rules follow from this:
Size against the worst-case ambient, not the average. If the cabinet sees 55°C on a summer afternoon, size for 55°C. Ambient averages are useless here because thermal failure isn’t caused by averages.
Ask where the temperature was measured. Internal cabinet temperature and external ambient are different numbers, and suppliers quote both. A rating taken at the supply’s own air intake inside a ventilated cabinet is far more useful than one taken in free air on a bench.
The counter-intuitive part: adding a bigger supply to “be safe” can make thermal behaviour worse, not better. A 600W unit running at 50% load generates less heat than a 200W unit running at 95% load in the same box — but it also has a larger thermal mass and often a lower-efficiency operating point at light load. Oversizing is not automatically a safety margin. It is a different operating point, and you should check what the efficiency curve does there.
The 199 × 99 × 50 mm footprint of a CPS-600-xx is compact, which is exactly why it gets installed badly. People see a small box and tuck it into a small gap.
Convection needs room. A unit dissipating tens of watts needs an unobstructed air path below and above it, and the path has to actually connect to something — a vent low on the cabinet and a vent high on the cabinet, or a fan. Sealing a hot supply into a dead air pocket is the fastest way to guarantee a warranty claim.
The 50mm mistake is this: leaving clearance on five sides and forgetting the sixth. In a 199 × 99 × 50 mm unit, the 199 × 99 face is the largest surface and the primary heat path when it’s mounted flat. Block that face — by mounting directly against a cabinet wall, or stacking another device on top of it — and you have removed most of the cooling capability while leaving what looks like adequate clearance everywhere else.
Specific guidance:
At 24V output, 600W is 25A. At 12V, it’s 50A. Those are output-side numbers, and they determine your cable gauge — at 50A over any real distance you are looking at substantially heavier copper than most installers budget for.
On the input side, the calculation most people skip is inrush. A switching supply draws a large current spike for a few milliseconds at turn-on as its input capacitors charge. That spike doesn’t show up in steady-state specs, and it’s what actually trips breakers.
The failure mode looks like this: the installation tests fine with one supply, someone adds three more on the same circuit, and now the whole string trips randomly at power-up. Nothing is wrong with any individual unit. The breaker was sized for steady-state current and never checked against the combined inrush.
What to do about it:
This is also the strongest argument for splitting 600W into smaller units in some designs: four 150W supplies have four separate inrush events you can stagger, versus one large one you can’t.
The cost comparison is closer than it looks. The two-unit approach costs more in copper and mounting hardware, but less in downtime and diagnosis. On an installation where a truck roll costs real money — which is most outdoor installations — the redundancy usually wins.
The exception is genuinely dense, short-run, cabinet-mounted work where all the load is within a couple of metres and access is easy. There, one 600W unit is simpler and cheaper and there’s no real argument against it.
Confirm each of these against the actual datasheet before signing off. The values below for the CPS-600-xx are from current published specs; anything marked with a placeholder needs a confirmed figure.
Two things to demand in writing rather than accepting verbally: the derating curve and the warranty exclusions. Both are where expectations and reality tend to diverge, and both are much easier to resolve before an order than after a failure.
If your wattage, voltage, or enclosure constraints don’t match a standard unit, the custom configuration route is usually faster than trying to make a standard part fit — and for a 600W build in particular, getting the thermal design right from the start is worth more than saving on unit price.
Ready to size a specific installation? Send the wattage, voltage, cabinet dimensions and quantity and we’ll work the derating and breaker math against your actual numbers.
How many amps does a 600W 24V power supply draw? 25A at full load (600 ÷ 24). On a 12V model the same 600W is 50A, which is why 12V systems at this wattage need significantly heavier cable.
Can I run a 600W supply at full load continuously? Not in every environment. Continuous full-load operation depends on ambient temperature and airflow; check the derating curve for your actual cabinet conditions rather than assuming nameplate output is available year-round.
Is one 600W better than two 300W? It depends on fault tolerance. One 600W is tidier and uses less copper; two 300W units give you partial redundancy and faster fault localisation. If the installation is hard to access, split it.
Why does my breaker trip at switch-on but not during operation? Almost always inrush current. The input capacitors draw a brief high-current spike at turn-on that steady-state specs don’t capture. Size the breaker against inrush, and stagger multiple supplies.
Do I need a fan in the cabinet? Depends on total wattage dissipated and cabinet volume. A single 600W unit at 90% efficiency sheds roughly 67W; whether passive venting handles that depends entirely on your cabinet geometry and the ambient range.
Published specs referenced: CPS-600-xx 199×99×50mm, IP20, CE/FCC/CB/UL, 5-year warranty, MOQ 50pcs. Efficiency, derating curve, inrush current, mounting orientation and clearance figures pending confirmation from the factory datasheet.
Need robust energy efficiency for automation machinery? Discover our high-performance enclosed AC-DC converters built for heavy industrial environments.
The 20% headroom rule is not about future expansion. It is about keeping the driver between 60% and 80% load, which is where both the efficiency curve…
24 V is not automatically the better choice. Below about 3 metres and 30 W per run, 12 V modules cost less and you gain nothing from 24 V. The math be…
IP68 is a worse buying decision than IP67, and the reason is uncomfortable: IP68 has no standardised test. Under IEC 60529, IP68 means only “dee…
100% burn-in is the dividing line in this industry. Full-load burn-in costs roughly 3–5% of unit cost in electricity, rack space and labour. Factories…
Efficiency is the number everyone compares and the number that matters least. Going from 90% to 93% on a 200 W sign running 12 hours a day saves about…
Housing customisation is the worst-value lever on this list. It costs the most (2,000–8,000), pushes MOQ to 3,000, adds 30 days, and your customer nev…
