Most cooling problems are not caused by a bad fan — they are caused by a fan chosen for the wrong reason. A model picked purely on “CFM” often runs loud, stalls against a dense heatsink, or fails two years before the product warranty ends. This guide walks through the six decisions that actually determine whether a DC fan works in your enclosure.
Step 1: Calculate the heat you need to remove
Start with watts, not with fan size. Every enclosure has a heat load: the sum of losses from the power supply, the processor, the motor driver and any other component that converts electricity into heat. A common rule of thumb for a sealed enclosure is:
Required airflow (CFM) ≈ 3.16 × heat load (W) ÷ allowable temperature rise (°C)
For example, a 60 W load with a 15 °C rise needs roughly 12.6 CFM of actual delivered airflow. That number is a starting point, not a final answer: filters, grills, tight bends and dense heatsinks all reduce real flow. Engineers typically add 30–50% margin to cover these losses. If you are unsure about the heat load, measure the internal temperature rise with the enclosure running closed for one hour; the difference between ambient and internal air tells you what the fan has to fight.
Step 2: Match the fan to the physical space
Enclosure geometry decides your frame size more than thermodynamics does. Standard DC axial fans run from 30 mm to 150 mm in square frames, with thicknesses from 10 mm to 38 mm. The classic trade-off is simple: a larger, thicker fan moves the same air at lower speed, which means less noise per CFM. If your product allows a 92 mm fan instead of an 80 mm fan, take it — the acoustic benefit is usually larger than the cost difference.
Also allow for the air path. A fan pressed against a solid wall or a sealed panel will recirculate its own exhaust. Leave at least one fan-diameter of free space on the intake side, and make sure the exhaust has an open area of at least 60–70% of the fan’s face area.
Step 3: Separate airflow from static pressure
Airflow (CFM or m³/h) describes volume moved against almost no resistance. Static pressure (Pa or inH₂O) describes the fan’s ability to push air through resistance. A dense heatsink, a filter mat or a long duct needs pressure; an open chassis needs volume. Most 12 V axial fans deliver their rated CFM only at zero pressure, which never happens in a real product.
The honest way to choose is to compare the fan’s P-Q curve with your system’s resistance curve. The intersection is your actual operating point. Without a curve, a practical rule is: if you can see through the airflow path easily, an axial fan is fine; if the air must squeeze through fins or a filter, choose a high-pressure model or a blower.
Step 4: Choose voltage and control method
DC fans are commonly available in 5 V, 12 V, 24 V and 48 V. Match the rail you already have — adding a dedicated 12 V supply for one fan rarely makes sense. If the fan speed must follow temperature, specify a 3-wire (tach) or 4-wire (PWM) version rather than dropping the supply voltage, which shifts the operating point unpredictably and can stop the fan at low voltage.
Where noise is critical, PWM with a soft-start function avoids the audible “kick” at power-on. Where the fan is inaccessible, a locked-rotor alarm output is worth the extra wire: it lets your controller report a failed fan instead of silently overheating.
Step 5: Decide bearing type and expected life
Sleeve bearings are cheap and quiet when new, but their life drops sharply at high temperature. Dual ball bearings cost more and are slightly louder, yet they tolerate heat and any mounting orientation. For products with a five-year service life, or for fans mounted vertically, dual ball or hydraulic bearings are usually the safer choice. Ask for L10 life data at your actual operating temperature rather than a generic “50,000 hours” figure measured at 25 °C.
Step 6: Validate with a prototype, not a datasheet
Before release, run three checks: measure internal temperature at the worst-case ambient, measure noise at the distance your user will experience (often 1 m, not 1 m in an anechoic room with the fan free-standing), and run a 72-hour continuous test to catch early bearing noise. If the fan is speed-controlled, verify start-up at the lowest duty cycle and check for resonance at mid-speed, where blade-pass frequency often excites a plastic panel.
What to send a supplier
A good inquiry shortens the process dramatically. Send the enclosure drawing or a photo, the supply voltage, the measured or estimated heat load, the maximum acceptable noise, the ambient temperature range and the annual quantity. With those five items, a competent manufacturer can propose two or three models the same day — and tell you honestly which one will not work.
Five mistakes that show up again and again
Choosing by frame size. “We have an 80 mm hole, so we need an 80 mm fan” ignores the heat load entirely. Size is a constraint, not a specification. Trusting the headline CFM. As explained above, catalogue airflow is measured at zero resistance; the delivered figure in a real enclosure is often 40–60% of it. Ignoring the intake. A fan pressed against a perforated panel with 35% open area is fighting itself, and no amount of extra speed fixes it. Over-speccing the bearing but under-speccing the environment. A dual-ball fan in a 75 °C enclosure still ages quickly if the lubricant is rated for 60 °C. Testing at room temperature only. Thermal problems appear at the temperature extremes of the real installation, not on a comfortable bench.
Each of these mistakes has the same root cause: treating the fan as a catalogue item rather than as part of a thermal system. The fix is procedural — write down the heat load, the resistance, the ambient range and the noise target before opening any datasheet.
Working with a supplier efficiently
The best suppliers will push back on an unrealistic brief, and that is a good sign. If you ask for 60 CFM from a 40 mm fan at 20 dBA, a serious manufacturer will tell you the physics does not allow it, while a trading company will simply quote a number. Ask for measured curves, request a sample with test data, and validate in your own enclosure before committing to tooling. A supplier who supports that process is worth a small price premium; one who discourages it is a risk you will discover later, at the worst possible moment.
Need help selecting the right fan?
Send your size, voltage and airflow requirements — our engineers reply within 24 hours.