Thermal experiment
Workshop enclosure airflow test with two 120 mm fans
A small electronics enclosure in the workshop had one 120 mm intake fan and one 120 mm exhaust fan. Both were nominally identical PWM models. I wanted to know whether running both at low speed was actually better than using only the exhaust fan, and whether the intake filter was causing most of the temperature rise.
The enclosure held a 35 W resistive test load for repeatability. Ambient temperature was measured 0.5 m from the intake. Inside, one DS18B20 sat near the intake edge and another was fixed 30 mm above the load. Fan speed was controlled from an ESP32. Each test ran for 50 minutes; I used the average of the last ten minutes after temperatures had mostly stabilized.
| Fan mode | Approx. duty | Intake ΔT | Load-area ΔT |
|---|---|---|---|
| Exhaust only | 45% | 2.1 °C | 8.7 °C |
| Intake + exhaust | 35% / 35% | 1.4 °C | 5.2 °C |
| Intake + exhaust | 55% / 55% | 0.9 °C | 3.6 °C |
| Two fans, filter removed | 35% / 35% | 1.0 °C | 4.4 °C |
The filter cost about 0.8 °C around the load at the lower fan setting. That is measurable but not enough to justify running without it in a dusty room. More importantly, two slow fans produced a much better result than one fan at a similar perceived noise level. The intake fan appears to reduce recirculation around the lower vents rather than merely adding airflow.
I also tested fan direction with smoke from a safe distance outside the enclosure. With exhaust only, air entered through both the intended lower vent and several cable openings near the top. With the intake fan running, most visible flow followed the intended bottom-to-top path.
Control rule
The final control is intentionally simple. Below 30 °C inside the enclosure both fans stay at 30% duty. Between 30 °C and 40 °C duty rises linearly to 70%. Above 40 °C both fans run at 100%. The controller also forces full speed if either internal temperature sensor has not updated for more than two minutes.
No attempt is made to hold an exact temperature. The enclosure only needs enough airflow to keep the temperature rise bounded while staying quiet most of the time.