Last time, we covered storing and charging a 3000 Farad supercapacitor to build a solar-powered, portable spot welder. Since then, I’ve made some improvements to the charging circuit and gotten it running. To recap, the charger uses a DC-DC buck converter to convert a range of DC voltages down to 2.6 V. It can supply a maximum of 5 A though, and the supercapacitor will draw more than that if allowed to.
After some failed attempts, I had solved that by passing the buck converter output through a salvaged power MOSFET. A spare NodeMCU module provided pulse width modulated output that switched the MOSFET on for controlled periods of time to limit the charging current. That was fine, but a constant-voltage charger really isn’t the right way to load up a capacitor. Because the capacitor plates build up a voltage as it charges, the current output from a constant-voltage charger is high initially, but drops to a very low rate in the end.
Before Lunar New Year, I had ordered two 3000 F, 2.7 V supercapacitors from China for about $4 each. I don’t actually remember why, but they arrived (unexpectedly) just before the holiday.
Supercapacitors (often called ultracapacitors) fill a niche somewhere between rechargeable lithium cells and ordinary capacitors. Ordinary capacitors have a low energy density, but a high power density: they can store and release energy very quickly. Lithium cells store a lot of energy, but charge and discharge at a comparatively low rate. By weight, supercapacitors store on the order of ten times less energy than lithium cells, and can deliver something like ten times lower power than capacitors.
[Mike Rigsby] has moved a train with a coin cell. A CR2477 cell to be exact, which is to say one of the slightly more chunky examples, and the train in question isn’t the full size variety but a model railroad surrounding a Christmas tree, but nevertheless, the train moved.
A coin cell on its own will not move a model locomotive designed to run on twelve volts. So [Mark] used a boost converter to turn three volts into twelve. The coin cell has a high internal resistance, though, so first the coin cell was discharged into a couple of supercapacitors which would feed the boost converter. As his supercaps were charging, he meticulously logged the voltage over time, and found that the first one took 18 hours to charge while the second required 51 hours.
This is important and useful data for entrants to our Coin Cell Challenge, several of whom are also going for a supercap approach to provide a one-off power boost. We suspect though that he might have drawn a little more from the cell, had he selected a dedicated supercap charger circuit.
Clearly a believer in the old adage, “Go Big or Go Home”, [Ted Yapo] has decided to do something that seems impossible at first glance: starting his car with a CR2477 battery. He’s done the math and it looks promising, though it’s yet to be seen if the real world will be as accommodating. At the very least, [Ted] found a video by [ElectroBOOM] claiming to have started a car with a super capacitor, so it isn’t completely without precedent.
Doing some research, [Ted] found it takes approximately 2,000 W to 3,000 W at 14 V to start the average car engine. This is obviously far in excess of what a coin cell can put out instantaneously, but the key is in the surprising amount of potential energy stored in one of these batteries. If the cell is rated for 1000 mAh at 3 V, [Ted] shows the math to find the stored energy in Joules:
According to the video by [ElectroBOOM], he was able to start his car with only 6,527 J, and [Ted] calculates it should only take about 9,000 J on the high side from his research. So as long as he can come up with a boost converter that can charge a capacitor with high enough efficiency, this one should be in the bag.
[Ted] has started putting together some early hardware, and has even posted the source code he’s using on a PIC12LF1571 to drive the converter. He notes the current charge efficiency is around half of what’s needed according to his calculations, but he does mention it was an early test and improvements can be made. Will it start? If it does, this is some awesome Heavy Lifting.
There’s some interesting technology bundled into this energy harvesting wristwatch. While energy harvesting timepieces (called automatic watches) have been around for nearly 240 years, [bobricius] has used parts and methods that are more easily transferable to other projects.
Unlike early mechanical systems, this design uses the versatile BPW34 PIN photodiode (PDF warning). PIN photodiodes differ from ordinary PN diodes in that there’s a layer of undoped ‘intrinsic’ silicon separating the P and N doped layers. This reduces the utility of the diode as a rectifier, while allowing for higher quantum efficiency and switching speed.
They are typically used in the telecommunications industry, but have a number of interesting ‘off label’ applications. For example, the BPW34 can be used as a solid-state particle detector (although for detecting alpha particles you’re better off with something in a TO-5 package such as the Hamamatsu S1223-01). The fast response speed means you can send data with lasers or ambient light at high frequencies – a fun use for an LED lighting system or scrap DVD-RW laser.
Some common solar panels are essentially large PIN photodiodes. These are the brownish panels that you’ll find in a solar-powered calculator, or one of those eternally waving golden plastic neko shrines. They specifically offer excellent low-light performance, which is the basis of the energy harvesting used in this project.
Consider for a second the Internet of Things. A vast network of connected devices, programmable matter, and wearable electronics can only mean one thing: there’s going to be a ton of batteries. While changing the battery in a smoke detector may seem tolerable, changing the batteries in a thousand sensor nodes is untenable. The solution to this problem is self-contained sensor nodes, and right now the best power source for mobile devices is probably solar.
For his Hackaday Prize entry, [Shantam Raj] is building a self-contained sensor node. It’s a Bluetooth device for the Internet side of this Thing, but the real trick to this device is solar energy harvesting and low power capabilities through optimized firmware.
Basically, this system is a low-power SoC with Bluetooth. The power from this device comes from a small solar cell coupled with a very efficient power supply and some new, interesting supercapacitors from Murata. These supercaps are extremely small, have high storage capacity, low ESR, and fast charging and discharging. The test board (seen in the video below) provides a proof of concept, but this device has a problem: there’s a single ‘sanity check’/power LED on the board that consumes 4 mA. The microcontroller, when running the optimized firmware, only consumes 1 mA. Yes, the LED thrown into the prototype that only serves as an indication the device is on is the biggest power sink in the entire system.
This project is great, and it’s exactly what we’re looking for in The Hackaday Prize. If the Internet of Things ever happens as it was envisioned, we’re going to be buried under a mountain of coin cell lithium batteries. Some sort of energy harvesting scheme is the only way around this, and we’re happy to see someone is working on the problem.
What do you do when you find a small horde of supercapacitors? The correct answer is a spectrum of dangerous devices ranging from gauss guns to quarter shrinkers. [Rinoa] had a less destructive idea: she’s replaced the battery in a laptop with a bank of supercapacitors.
The supercaps in question are 2.7 Volt, 500 Farad caps arranged in banks six for a total of about 3 watt-hours in each bank. The laptop used for this experiment is an IBM Thinkpad from around 1998. The stock battery in this laptop is sufficiently less advanced than today’s laptop batteries. Instead of using a microcontroller and SMBus in the battery, the only connections between the battery and laptop are power, ground, and connections for a thermocouple. This is standard for laptops of the mid-90s, and common in low-end laptops of the early 2000s. It also makes hacking these batteries very easy as there’s no associated microprocessors to futz around with.
With all the capacitor banks charged, the laptop works. It should – there isn’t a lot of intelligence in this battery. With one bank of six supercaps, [Rinoa] is getting a few minutes of power on her laptop. With a stack of supercaps that take up about the same volume as this already think Thickpad, [Rinoa] can play a few turns of her favorite late-90s turn-based strategy game. It’s not much, but it does work.