Fire Power

Seebeck Powered Phone Charger

Johann Wyss

Issue 20, March 2019

This article includes additional downloadable resources.
Please log in to access.

Log in

Keeping your tech running while you’re away from civilisation and mains power can be challenging. We’ll show you how to convert some heat from the campfire or stovetop into power to charge your USB Smart devices.


If you have ever been fortunate enough to leave civilisation for more than a few days, perhaps on an adventure experiencing the great outdoors, then you most likely found a new appreciation for the many modern conveniences we have. The most obvious for me (being the technology lover that I am) is just how easy it is for us to find power to keep our tech charged. USB power adaptors in almost every room in the house, in our workplace, in the car, at tables at the airport and now in bus seats.

However, if like many Australians you love getting out and about camping and hiking, then this device may just be for you. It allows you to charge your electronic devices such as phones and cameras by converting thermal energy into electrical energy.


More and more of our tech is powered or charged by USB these days. At the same time, bright screens and power intensive apps seem to drain the battery faster then ever, even with the latest in battery technology.

We also tend to forget that rechargeable batteries have a limited number of charge cycles before they lose performance. Many of us are always reaching for a charging lead wherever we can, or leaving home with a power supply in our backpack or handbag, just in case that percentage remaining indicator changes colour.

If you’re going “off grid” for an extended period, away from mains power then how do you charge your phone? Sure, you can take a powerbank with you, but what happens when that goes flat?

One solution is a solar charger or powerbank, which are great, but of course, rely on many hours of full sunlight.

Another method is converting heat into power using the Seebeck effect. Wikipedia describes the Seebeck effect as a phenomenon in which a temperature difference between two dissimilar electrical conductors or semiconductors produces a voltage difference between two substances.

This brings us to Peltier devices that are available from many electronics retailers. We’ll show you how to generate power to charge your USB tech by applying heat to the Peltier. The heatsource could be from coals on your campfire, candles, or from your gas stovetop, for example. A 5V boost module will be used to keep the output voltage at 5 volts, and we’ll describe how to store some of the power using a small charging module and Li-ion battery.



construction diagram DIAGRAM CREDIT:

The heart of this charger is the Peltier or Thermoelectric cooler/solid-state heat pump.

As the current flows through the series connected P-N junctions, one side of the device begins to absorb heat and transferring it to the opposite side. This process creates the temperature differential. One side will get cold and the other side hot.

To assist this process, it is mandatory to remove the heat from the hot side as quickly as possible. This is usually done using a heatsink, which when done correctly, will allow the cold side to reach temperatures several degrees below ambient temperatures.

Unlike thermoelectric coolers on the market, we are using the Peltier in reverse. The heat from our heat source excites the barrier regions into conduction, allowing a current to flow due to the potential difference of the PN array.

This method may not seem initially very useful for charging your electronic devices, however, it isn’t difficult to turn a Peltier heat pump into a Seebeck generator.

By applying a large temperature difference across the outer plates of the Peltier device we generate a small voltage and current. The bigger this difference, the more power we can generate, provided of course that you don’t exceed the absolute maximums of the device.

It’s worth mentioning at this point that Peltier devices are incredibly inefficient (usually lower than 5% efficiency!), and using them as a generator i.e. in ways they are not specifically designed to work is especially inefficient.

Experiment 1:

Testing the Peltier

Parts Required:Jaycar
1 × Peltier Device^ZP9100
1 × 5V USB Boost ConverterXC4512
1 × CPU Fan with Heatsink-
1 × Adhesive Heatsink CompoundNM2014
Tea Light Candle
Third Hand
12V Power Source to Power CPU Fan
Digital Multimeter
USB Power Tester

^You will find out by Experiment 3 that a larger peltier is required for adequate power output.

The first steps in my quest for a Seebeck phone charger were to test the principals and performance of a Peltier. I wanted to see what voltage and current I could get with different heating and cooling methods. My first experiment was simple, using a CPU fan with a heatsink attached to the cold side of the Peltier. I used a common tea light candle as a heat source, and a multimeter to measure the voltage directly across the Peltier.

The cold side of a Peltier is normally the side with its identification print on it. If you are unsure, an easy way to find the distinct hot and cold sides is to apply 12V to it for a few seconds (Red is positive, black is negative). One side will get very hot and the other will stay cool. The hot side is the side we heat up and the cold we need to keep cool.

The heatsink compound was enough to hold the peltier securely against the heatsink. I used a 12V power supply to power the fan, and I used a third hand to hold the assembly above the candle.

With this simple experiement, the multimeter displayed a consistent voltage output of around 2 – 2.5 volts.

Next, I attached a DC-DC USB Boost Module, designed to step-up a voltage between 2.5-5V to 5V. This module managed to produce a current of around 90mA at 5V, which is around 0.5W of energy. To effectively measure the current, I used a Uni-T UT658 USB tester in conjunction with a generic USB electronic constant current load. This is a USB device that I sourced online that allows me to “dial” in a desired constant current to test projects, just like this one.

I would increase the current demand on the constant current load whilst monitoring the voltage. When the voltage dropped to 4.2 volts the current was recorded as the maximum operational current output. 4.2 volts was selected, as this is the nominal lithium voltage, and the minimum voltage required for the TP4056 lithium charging IC to operate correctly.

Considering this low power output, I thought about trying three Peltiers in series to generate enough current, which brings us to the next experiment.

Experiment 2:

Testing for More Current

Parts Required: Jaycar
1 × Peltier Device^ZP9100
1 × Adhesive Heatsink CompoundNM2014
2 × Rectangular Cake Tins-
Hookup Wire-
Tea Light Candles-
Dremel or similar metal cutter-
Drill with 10mm drill bit-

^You will find out by Experiment 3 that a larger peltier is required for adequate power output.

To increase the current, I experimented using three of the peltiers in parallel. I used a regular cake tin and glued the three peltiers to the bottom. I then sandwiched the peltiers to another cake tin which had a cutout in the side to allow candles to burn underneath. See below for details on how to cut the hole.

I used cold water to dissapate the heat on the cold side because I couldn’t install CPU fans like in my first experiment. The concept was as the water heats up you tip it out and add new cold water (or snow if you were lucky enough).

I added 8 candles and took measurements similar to my first experiment.


Unfortunately, the power output was still not nearly enough. I only recorded about 200mA of stable current when using the three Peltiers in parallel which is only 1W. This is less than I was expecting as in my previous experiment a single Peltier of this size produced 90mA. I was expecting to be closer to 270mA or 1.35W.

This discrepancy could be caused by uneven heating across the Peltier elements or perhaps even inaccuracies in the measurement devices. Whilst 1W is enough to slowly charge your mobile phone I was less than impressed and wanted to get more efficiency from the device.

Wikipedia says a standard tealight candle produces about 32W of energy.

Since we are using four tealight candles, our input power is about 130W.

Efficiency is calculated using n = Pout/Pin ×100

This means our device has an efficiency of:

(1W/130W) × 100 == 0.77% efficient.

After some research online, I decided to try the experiment using a larger Peltier with significantly higher rated power input.

Experiment 3:

Using a Higher Power Peltier

Parts Required: Jaycar
1 × 68W Peltier DevicesZP9100
1 × Adhesive Heatsink CompoundXC4512
1 × CPU Fan with Heatsink-
8 × Tea Light Candles
2 × Rectangular Cake Tins
Drill with 10mm Drill Bit

I purchased the higher rated Peltier device to confirm if it would meet my expectations. I applied a similar approach as my first experiment, but this time, I used a large tin can for the hot side to avoid heat escaping. I removed one end from the can and drilled several ventilation holes. I used heatsink compound to glue the hot side of the Peltier device to the closed end of the tin can.

As shown above, I used some corregated cardboard to provide a little thermal resistance. We want all the heat to transfer via the Peltier, so having that barrier helps slow down the hot sides heating the cold side. Next, I glued the CPU heatsink and fan to the cold side of the Peltier device.

Given the larger distance between the flame and the Peltier device I opted to use an alcohol burner that you find in bain-marie food serving trays. This produced a larger and hotter flame compared to the tea light candle and therefore should improve the power output.


In this configuration, the device was able to produce 5V with a higher current, peaking at about 200mA or about 1W of energy out of the boost converter.

This confimed that I now had a much more ideal Peltier, compared to the lower wattage one I tried in the previous experiments. This higher wattage device would have produced upto about 3W in the previous configuration of three Peltier elements, which is much more compared to the three I used in the previous experiment that only produced about 1.3W of power.

Therefore, I recommend you use the 68W device if you plan to build this project or want to get similar results in your own creation.


It's worth noting at this point I was completely flustered. I could see that the device was producing 5V and 200mA of current, yet despite this, anytime I connected a newer iPhone to the device it would refuse to charge. I again had to go online to research the charging process for Apple products and learned that Apple will not allow a newer iPhone to charge unless specific D- and D+ voltage conditions are observed by the phone.

In many cases a voltage divider across the pins is used, however, I learned that shorting the D+ and D- pins together is an accepted practice. I did this to the USB boost converter and it instantly allowed the newer iPhones to charge. With the USB data pins soldered together as shown here, this is all one needs to charge a smartphone from the Peltier device. It does, however, charge very slowly!

The Main Build:

Seebeck Powered Phone Charger

build photo
Parts Required:Jaycar
3 × Peltier DevicesZP9104
1 × 5V USB Boost ConverterXC4512
1 × Adhesive Heatsink CompoundNM2014
1 × Lithium Charger^XC4502
1 × Lithium Battery^SB2308
1 × Battery Holder^PH9205
2 × Male Pin Header StripHM3212
1 × SPDT SwitchST0336
2 × Rectangular Cake Tins
Hookup Wire
Tea Light Candles
Electrical Tape
Dremel or similar metal cutter
Drill with metal cutting drill bits
Access to a 3D printer
Permanent Marker
Nail Punch and Hammer
Metal File
High Temperature Silicon

^Devices marked with the ^ are considered optional if you want to make a rechargeable version. If you don’t have a portable battery bank I highly recommend you add these components as you may not want your phone close to a heat source for extended periods.

For my final build I decided to make it entirely self-contained with a rechargeable battery built-in. Making it rechargeable allows you to charge your device without the need to have your device and cables around an open flame. You can, of course, skip using a rechargeable battery depending on your application.

My ultimate goal was for my final build to use a PCB and be neatly housed in a 3D printed enclosure. As a proof of concept, I installed all of the parts onto a perfboard and soldered tracks underneath.


top view

Note: This image was taken while desperate to get the iPhone to charge. The Boost Converter was connected directly to a power supply. This was used to verify that the issues charging the iPhone were not related to current but rather the D+ and D- pins.

I soldered all the components to a 70 × 90mm perfboard. Using single pins from a pin header strip to make the connection between the small modules and the PCB. This configuration has the Peltier devices in parallel which provides the same voltage and higher current.

I soldered pins on the modules so they could be mounted on the perfboard.

I have provided the Eagle Board files for anyone who wishes to mill, etch or manufacture the PCB rather than using perfboard. You can download these from the resources section of the website. Please note that to cater for different boost modules, I have added a 2 pin header so you can physically wire the solder pads of the module to the board in the event the pin positions are not 100%. or your device has SMD pads compared to through hole.


I will guide you through how I prepared and assembled the cake tins with the Peltiers. You may find different tools and methods to get the job done.

You will need to choose which cake tin will be on the top (cool side) and which will be the hot side. The hot side will need a hole in the side to allow placement of the heat source and airflow.



For the cake tin assembly, I first used sandpaper to remove the Teflon and paint from the mating surface of the bread tins. This is where we need to adhere the Peltier device. I figured that the Peltier would adhere best with direct contact to the metal and would perhaps increase the thermal conductivity.

I then cut one side of one of the bread tins using a Dremel, drill, 10mm drill bit and a 3D printed template.



Download and print the 3D print template from the resources section on our website. Use electrical tape to secure it level against the side of the cake tin. Using a marker or scribe, make four dots in each hole in the template, and around the circumfrence. Remove the template and tape.

tin bottom


Using a hammer and nail punch or large nail, make four deep impressions in the cake tin. These will help to drill holes in the following step.


tin bottom 2

Drill four pilot holes using a 3mm drill bit and then enlarge the holes using bigger and bigger drill bits until the holes are 10mm. This will decrease burring and deformation of the tin.


tin bottom 3

Using a Dremel or similar, cut out the panel along the template lines you applied. Use a file if necessary to remove any sharp edges.

Make sure you're wearing safety glasses, its very easy to cause irreparable damage to your eyes when using high velocity cutting tools.


Once the two tins were prepared, I attached the hot side of three Peltier elements to the bottom tin using the heatsink adhesive. I used heavy textbooks on top to keep constant pressure as the glue cured overnight.

I then created a gasket out of pressed cardboard to fit around the outside of the elements. This will help prevent heat radiating from the bottom directly into the upper pan which would decrease performance. I used pressed cardboard, but other materials may be more suitable, just make sure its thickness is suitable, and that it can withstand prolonged high temps exceeding 250°C. I have made a template to assist in cutting out the gasket outer dimensions.

I secured the gasket in place using a small amount of silicone capable of withstanding high temps, however, this step isn’t essential as once we secure the top plate the gasket can’t move.

It was then a matter of attaching the top/cold pan to the Peltier coolers using the same heatsink adhesive. Again, heavy text books keep pressure on the two parts as the adhesive cured overnight. Sadly, I didn’t align my tins perfectly, despite my efforts, so they are slightly off center.


To test the device, it was a simple matter of attaching a drained 18650 lithium cell into the socket and applying candles to the bottom of the tin.

I found four candles worked the best, as more tended to cause the candles to burn significantly quicker as the ambient heat inside the bottom tin would cause the entire candle to liquefy.

I also tried using alcohol burners, these worked very well at producing a great amount of heat, but they don’t last very long.

For optimum results you could use charcoal in the bottom tray as the heat source and keep the top container full of cold water.

The colder the water on the top the greater the thermal difference and the higher the charge. Therefore, using this device in snowy conditions would be optimum use as you could fill the top container with snow and water slurry.

This device does work as intended as an emergency battery charger, but admittedly needs more R&D to make it more effective. In its current form, it does take many hours to fully charge the internal lithium cell. During this time the water needs to be refreshed as it heats up quickly. It would, however, still be useful to keep in your camping kit as it could very well get you out of a jam.


render crop

I designed a 3D case with a bracket to make the device sit on top of the cake tin. The 3D print files are available in the resources section. You don’t necessarily need a 3D printer to create this device as you can mount the PCB any number of ways by using a mill or just wood working skills, etc.

Since I have access to 3D printers, I chose to make the enclosure using that method and will include the design files on our website.


I always attempt to design my 3D prints so they don’t require supports. Sadly, it doesn’t always go to plan though. The base of this device needs to be printed vertically and this creates a long horizontal lip, which is printing in the air.

I recommend using Flashprint from Flashforge to print this component because Flashprint has treelike support structures, which minimise the amount of waste support material.


The lid needs to be printed upside down on its flat as seen below and shouldn’t need support material.


Experiment to find ways to improve the efficiency for higher current output. Perhaps using different materials, and methods to make the hot side retain more heat and the colder side kept cooler.

Larger Peltier modules could be considered, or replaced with a fit for purpose Thermoelectric generator, which works more efficiently and provides significantly higher charging currents. These TEG elements are however, much harder to source.

This is a great proof of concept ready to explore and expand on. There is limited efficiency in this idea and we're aware of that, but sometimes, heat is more abundant than power, so it's worth exploring.