Blog · · 5 min read
DIY BBQ temperature controller: probes, a small fan and a loop that does not overshoot
How a controller holds a charcoal smoker at 110 to 135 °C for ten hours: thermistor or thermocouple probes, a small fan with a flap, a PID loop that copes with an open lid, and electronics kept away from the heat.
Low and slow is a simple recipe with a hard requirement. A charcoal smoker has to hold between 110 and 135 °C for ten hours or more, while the wind picks up, the sun goes down and the fuel burns away. A controller does it with a probe, a small fan and a bit of logic.
It is also a good first electronics product: a sensor, a motor, a control loop and a box that lives next to a fire.
The fire only has an accelerator
A fan can add air, and air makes the fire hotter. Nothing in the system can take heat away. Stop the fan and the fire slows, but the steel, the ceramic and the meat keep their heat, and a heavy cooker takes a long time to lose even 20 degrees.
So the controller need not be fast, but it must avoid overshoot, because getting back down takes far longer than it took to get up.
It also means the fan must be the only way in for air. Close the bottom vents, open only the fan port and leave the top vent partly open. If the cooker can draw enough air on its own, the controller can push the temperature up but never down. A flap that falls shut when the fan stops closes that leak for almost nothing.
Thermistor or thermocouple?
Most food and pit probes use an NTC thermistor, a resistor whose value falls as it heats up, sealed in a stainless steel tip. Thermistors are cheap, give a large signal and, after calibration, read to within about a degree.
Their weakness shows up where a smoker works: the curve flattens at higher temperatures, so one degree changes the resistance much less at 120 °C than at 25 °C. Choose the fixed resistor in the voltage divider close to the probe's resistance at cooking temperature, so the reading sits mid-range. Then convert resistance to temperature with the Steinhart–Hart equation, fitting its three coefficients from three known points.
Probes of the same model can read slightly differently, so check each one. Ice water gives 0 °C. Boiling water gives 100 °C at sea level, about one degree less per 300 metres of altitude. A third point against a good reference thermometer in hot oil covers the cooking range.
A type K thermocouple handles well over 1,000 °C, the right sensor for a pizza oven or the fire itself. But its signal is tiny, about 41 microvolts per degree, and it measures the difference between the tip and the point where the wires meet the board. It needs a converter chip with cold-junction compensation, and standard-grade wire is only specified to about ±2.2 °C.
The cable is the weak spot: pinched under a lid, it sits in the hot gas, so check its rating, not only the tip's. And since many probes use small audio-style plugs, any probe fits; store which type is plugged in, or a different curve gives wrong numbers without warning.
How big a fan?
Smaller than you think. Commercial controllers ship fans of roughly 6 to 10 cubic feet per minute for kamados and about 15 to 20 for kettles; only large offsets and stick burners need several times that. An oversized fan turns every pulse into a flare-up.
PWM speed control gives smoother air than on and off, but small blowers stall below a minimum duty. Below that, use pulses: full speed for a few seconds, then a pause. A short pipe keeps the motor away from the hot steel, and a hood keeps rain and grease out.
A PID loop with patience
A charcoal pit reacts to a change in airflow over minutes, not seconds. Sample every few seconds, but tune for the slow response: a modest proportional gain, a slow integral, and a derivative on the measured temperature that brakes when the pit climbs fast.
The lid is the real test. Open it and the pit probe drops sharply. A plain PID loop sees a huge error, drives the fan to full speed and winds up its integral. When the lid closes, the fire is roaring and the cooker overshoots for a long time. The fix is lid detection: recognise a drop faster than any fire could cause, freeze the integral, hold the fan at its last steady value and resume once the temperature recovers. Many commercial controllers do this.
A fresh fire in a cold cooker overshoots easily, so limit the fan or ramp the setpoint during start-up. And when a probe reads open or shorted, turn the fan off, not on. A dead fire is a nuisance; a fan stuck at full speed is a hazard. If the fan runs flat out and the pit still falls, raise an alarm: the fuel is probably gone.
Electronics next to a fire
Do not mount the box on the lid. Ordinary commercial-grade chips are rated from 0 to 70 °C, and a box on hot steel or in full sun can pass that. A side table, shade and a longer cable are cheaper than heat-rated parts.
Lithium cells should not be charged below 0 °C or above about 45 °C, so a controller charging in the sun beside a smoker may not charge safely. A 12-volt adapter that also powers the fan is simpler. On a battery, do the power budget first.
Long probe cables pick up noise from the fan's switching; an RC filter, averaging and a separate fan ground keep readings steady. Seal the box, point the connectors down and add a drain hole.
Test it before the first brisket
Check every probe in ice water and boiling water before the season. Then do a dry run with fire but no meat: log pit temperature and fan duty for three hours, open the lid at a fixed time and see how far it overshoots after closing. Unplug the pit probe and confirm that the fan stops. Note the vent settings that worked; they are part of the product too.
Follow the build
We have started this controller as an open-hardware project: DIY BBQ Temperature Controller. The concept comes first, and the project grows step by step with the interest it gets. It should also fit the BBQ Modular Grill as an extra module (how that grill works).
The short version
Make the fan the only way in for air. Use thermistors for pit and meat, a thermocouple only for the fire. Tune slowly, add lid detection, and turn the fan off when a probe fails. Keep electronics and battery away from the heat, and do a dry run before cooking.
To take your own grill gadget from idea to design, start at makemachine.ai/try.
Watch hardware get built by AI. Follow public projects on makemachine.ai, read every result in full and get an email when there is something new.
More from the blog
- Will my electronics enclosure overheat? A five-minute heat estimate for a sealed box
- Joining 3D printed parts: screws, heat-set inserts, snap fits or glue?
- Free or paid AI 3D modeling? What you really pay for, and a 30-minute test to compare tools
Open-hardware projects built with AI
- DIY Automatic Plant Watering System: An open-hardware watering system for indoor plants: a small pump, a refillable tank and a soil moisture sensor in each pot keep up to six…
- DIY BBQ Temperature Controller: A temperature controller for charcoal grills and smokers that makers can build themselves. A pit probe, two meat probes and a small fan on…
- ZählerSchau: ZählerSchau is a small, battery-powered unit that clips onto an electricity meter in Germany (water meters via a clamp adapter), takes a…
- Roadstoy: Roadstoy is a remote-controlled toy version of the new Tesla Roadster that can briefly jump and flip using compressed air. It drives on an…
- The forever Washing machine: The forever washing machine is a washing machine offered as a service instead of a product. Customers do not buy the machine but pay for…
- Bike Light: I want to build a bike light that needs no battery. Magnets sit in the spokes close to the hub, and a coil on the fork picks up their…
- Rooficle: I want to develop a light, efficient and affordable e-bike that can be used legally on public roads in Singapore. Its core is an…
- BBQ Modular Grill: I want to design a modular grill. Instead of one sealed block, it is made of separate parts that can each be bought on their own online…
- Go FANUC Go: A mobile base that moves an existing FANUC M-710iC industrial robot (about 560 kg, with a milling spindle at the wrist) out of a room and…
- Jump & Run Robot: I want to build a small, low-cost toy humanoid robot that people can put together themselves. It should be about 30 cm tall, or as small…