# Current limiting a dc-powered heating element

What would be the best way to limit the current into a 2.1Ohms, 12V max heating element fed with a range of known dc voltages (5V to 12V) with also known maximum input currents (2A to 5A)? The load can only be switched from the high side as the low side is shared with a thermocouple.

My first experiment was using a +/- 40 kHz to 400 kHz pwm signal to control a high side mosfet, with the duty cycle as a way to modulate the average current consumption. That led to big inrush currents, even at 1% on time.

So my second idea was to used a buck converter. Given a known input voltage and current sourcing capability (let's say 5V, 2A = 10W) and a estimated efficency of the buck (90%), that gives a certain amount of power that can be fed to the heating element (9W). Finally using ohms law, $V = \sqrt{P * R}$, I can determine the output voltage of the buck (in this case 4.34V).

The second idea works quite well, however I'm aiming for a compact solution, and the coil and the input/output capacitor end up quite having to be quite large when running simulations.

So I thought I would reach out to the community for any other possible solutions. Thanks in advance for your inputs!

• Which temperature is the heating element supposed to reach? For which purpose it is employed? How is it set-up physically? – Lorenzo Donati Apr 8 '18 at 20:27
• @LorenzoDonati It's a soldering tip, supposed to reach approx 350°C. – Nicolas Schurando Apr 8 '18 at 21:21

Since the buck regulator works, would you consider increasing the switching frequency for the sake of reducing component size?

Is powering the heater with AC an option? You could use phase control to regulate the voltage. You could also control for a ratio of full cycles on to full cycles off so that you are always switching on/off at zero, which eliminates transients.

• I did actually consider rising the switching frequency, and moved from a TPS54821 at 1.6MHz to a LM73606 at 2.2MHz which helped, but it still feels not enough. And I believe such a frequency is reaching the limit of what manufacturers seem to offer. – Nicolas Schurando Apr 9 '18 at 19:52
• AC would probably be best, and in fact I believe this is how most of soldering irons work, but in this specific case the whole point is using DC. – Nicolas Schurando Apr 9 '18 at 19:53

If you’re regulating current and not voltage you don’t need a large input or output capacitor. Basically, do solution #1 but add an inductor and diode (you may also need a snubber).

• I understand your reasoning but there a few things I am unsure about. Using a buck with voltage or current control, am I not still regulating voltage and not current? If I remove output capacitors, wouldn't that lessen the ability of the buck to self-regulate? And if I remove input capacitors, wouldn't that affect the buck performance as well but also other components share the same input rail? – Nicolas Schurando Apr 11 '18 at 9:29
• If the buck converter is providing constant current, the inductor is what smooths the current. The capacitor doesn’t help. If you remove the input capacitor it will have the same effect on the rail as the PWM option. – τεκ Apr 11 '18 at 13:06