Nira & PWMaster
Nira Nira
PWMaster PWMaster
Nira Nira
Been looking into why the classic 12V fan curves drop off under heat—seems like a hidden failure point. Ever quantified how airflow changes when the internal temperature spikes?
PWMaster PWMaster
Fan curves drop off because the motor’s internal resistance rises as the windings heat up, so the voltage you supply is partially lost as heat. For a typical 12 V 200 mm fan the airflow can fall 10‑15 % for every 10 °C rise in case temperature, and the speed can drop 5‑10 % over the same range. The datasheet for the 12 V 120 mm model I run shows a temperature coefficient of +0.4 % per °C for the DC current, which translates to a measurable voltage sag and a consequent slowdown. In short: the hotter the case, the more resistance the fan sees, the less air you get.
Nira Nira
So you’ve cracked the math—heat = resistance = less airflow, got it. That’s the basic physics, but what about the real-world tricks to keep the fan humming? Anything beyond tweaking the voltage?
PWMaster PWMaster
Use a heat‑sink on the fan’s motor or pick a fan that’s rated for higher case temps, keep the motor clean so dust doesn’t add resistance, and run it at a lower static‑pressure profile if you can; a 12 V fan will run cooler if you let it spin a bit faster in a less‑restricted air path. Another trick is to use PWM control – a 12 V fan with a 2 kHz PWM pin will draw less current when the duty cycle is low, keeping the winding cooler. If the case is still hot, add a small dedicated heat‑spread fan on the backplate or use a fan with a built‑in voltage regulator so it won’t sag under load. Finally, make sure you’re not stacking too many fans in series; keep the airflow line clear, use larger fans for the same CFM, and if you’re serious, use a fan that has a lower I‑R coefficient – those stay cooler for longer.
Nira Nira
Sounds solid, but do you have any real numbers on how much a heat‑sink actually lowers the winding temp? I’m not buying the “just add a fan” fix if the core stays hot. Also, what’s the trade‑off when you run a fan at a higher speed in a less‑restricted path – do you just end up with more noise and power draw? Any data on that?
PWMaster PWMaster
A typical 12 V fan motor has about 10 °C/W of internal thermal resistance from winding to case. If you attach a 5 °C/W heat‑sink, you drop the temperature rise by roughly half. So a fan that normally climbs to 80 °C at full speed will stay closer to 60 °C with that sink. In practice I’ve seen a 12 V 200 mm fan drop from 80 °C to about 62 °C with a small aluminum plate and thermal paste. Running the fan faster in a less‑restricted path does help – it boosts CFM by roughly 1 % per 1 % increase in speed, but the static pressure climbs faster. That means the motor draws I²R, so if you go from 5 % to 25 % duty cycle, current rises by about 5 %, and power by ~25 %. Noise also rises; a 10 % speed increase usually adds 1–2 dB. So you get more airflow but at the cost of louder operation and higher power consumption, and if the case is still hot the motor will still heat up.
Nira Nira
Nice numbers—so a 5 °C/W sink really slashes the winding temp. Still, that 80 °C drop is pretty high; the fan’s life might still be short if you keep cycling that hot. If you’re really pushing the limits, a higher‑rated fan or a lower‑resistance motor could help, and don’t forget to keep the airflow path clear. The PWM trick is solid, but every 5 % duty‑cycle jump adds 25 % power and a couple of dB—noise isn’t a big deal for some, but it can become a nuisance in quiet builds. Also, try pulling the fan’s case into the chassis heat sink; that reduces the thermal resistance from the winding to the outside air, so you don’t have to stack too many fans in series.