We updated the instructable with a new implementation you can build without spending all that money on those expensive seebeck generators. This one uses Peltiers in reverse to generate rather than cool. They aren't as efficient but they are much more affordable. Plus most of the parts required are made from scrap.
'''How to build a Steampunk WALL E'''
First Intro Video
Companies such as BMW are investing in Thermoelectric Generators to make their cars more efficient by replacing the alternator. Thermoelectric Generators convert wasted heat from the engine into electrical power. In this instructable we show how you can use the same technology right now at home to collect heat energy from car exhausts, waste oil burners and even our hands. We can power electronics, joule thiefs, super caps, Lego Car and anything else you can imagine.
This is our Epilog Laser Contest Entry. If you like it please vote for us.
Step 11'''How to build a Steampunk WALL E'''
Step 6Fire Powered Lego Car (Video)
Step 10Thermoelectric Joule Thief (LED that lights up from oven heat)
Step 1Whats a Seebeck? Is it different from a Peltier?
Step 2How much power does a Seebeck generate
Step 3Building an inefficient 5V regulator
Step 5Thermoelectric Implementations
Step 7More Efficient Switching Boost circuit / Info for designers
Just interestingThermoelectricity from Tellurex
Stick around and learn about the the difference between peltier and seebeck units, and what electrical circuits you could use to make a green impact on our environment by utilizing wasted heat.
Keep on reading and enjoy our videos :).
Step 1: How Thermoelectric Units Work
Here is a Peltier unit freezing water in 30 seconds.
Here is a mislabeled (whoops) Seebeck unit powering a motor. The thermoelectric unit generates electricity as heat passes through it from the fire to the ice water. The unit has a thin layer of thermal grease to help conductivity. It sits on a aluminum bar in ice water keeps the other side at 32F so we had a baseline. The top of the thermoelectric element is covered with metal to protect it from being scorched. It turns out we really didn't need it.
Our team member Chris LoBello was recently featured for taking first place in Polytechnic University Inno/Ventions competition with this idea.
Some Fun Photos Below of our test setup
Step 2: Specifications (warning Details and Science)
Chargers for other devices are rated at around 0.5W to trickle charge and 10W to fast charge.
Our seebeck unit is rated to generate 6 Watts so its more than enough to charge an iPod.
So how do we do this? Well we need a heat source and a cooling source. We can easily attach this to a car or a motorcycle since the exhaust easily goes over 200F and the wind over the element would cool it.
Phones and MP3 players have internal charging circuits. If you supply the right voltage and enough current it will start charging. Most phones and iPods need 5 volts so we decided to use a 5V regulator. The regulator will make sure the voltage never gets too high.
If we use a seebeck unit with an efficient regulation circuit we could charge an iPod at the same rate as an outlet charger. We could even charge batteries or a super capacitor and use it as a battery booster when the iPod dies.
We had designed a boost charging circuit but it is not quite ready for prime time. Sorry :-( For now, we will show you how to use a 5V linear regulator instead. Its not as efficient but it works.
It was the G1-1.4-219-1.14 $75 from tellurex.
Step 3: Parts List (5V Regulator)
Heat sink - Radioshack or from an old computer. http://www.radioshack.com/search/index.jsp?kwCatId=&kw=heatsink&origkw=heatsink&sr=1
Thermal Paste - We got ours from dealextreme.com. Its not the best, but it was cheap and offered free shipping. A higher grade thermal paste would have improved it. http://www.radioshack.com/product/index.jsp?productId=2526910
5V Regulator, LM7805 - This can also be purchased from Radioshack, as you can see, our's was. http://www.radioshack.com/product/index.jsp?productId=2062599
0.33uF Capacitor - Radioshack or your favorite electronics store. (optional...ish)
0.1uF Capacitor - Radioshack or your favorite electronics store. (optional...ish)
USB female plug - We used a USB plug salvaged from a motherboard, but you can use anything that fits the device you would like to charge. Its actually easier to cut up a USB extension cord that has the USB receptacle plug.
Seebeck unit - http://tellurex.com
Step 4: Build It
This would not work with a single seebeck unit because this kind of regulator needs more than 5Vs and is very inefficient. Two seebeck units connected in series would be enough to satisfy its input voltage.
A lot of people also seem to have an issue with trying to fool an iPod touch to think it is connected to a computer. We built a break out board and tied D- to 5V directly and D+ with a 10K to GND. It worked on our's, the iPod's USB state machine is fooled and we have been charging it this way for a while.
Here are some other iPod chargers people have made.
We are working on a boost circuit that will allow you to charge an iPod with just one seebeck unit, but it won't be ready before this contest is over.
Step 5: Implementations
A seebeck mounted on a motorcycle. In this case, the heat sink is much larger than it needs to be. Three seebeck units can actually be mounted under that heat sink.
Since exhaust pipes are generally round and seebeck elements are rigid, it would be a good idea to attach them to the largest diameter part of the exhaust pipe and use smaller seebeck units. This would provide a greater surface area and result in higher efficiency. Hose clamps secure it to the pipe.
Another area we can attach these would be heaters or a fireplace in a house. These get pretty hot in the winter while the air temp would be cold enough to generate power. A very large super capacitor with an efficient boost regulator can be charged to store energy for a quick boost to your phone or mp3 player when needed.
In the summer the seebeck unit could be placed in the central air conditioner heat exhaust. Air conditioners actually expel heat as well, that's why they need something on the outside. Perhaps someone could install this into the central air conditioner to power a LED porch light in the summer. The circuitry would be similar to a solar charger.
Exhaust vents from your stove
Just an idea, we never looked into this one.
Charging Super Capacitors or batteries
If you connect this up to a super capacitor or batteries you could collect the power you generate and bring it with you.
Kryptonite recommended we attach this to a Joule Thief. With it we were able to power an LED easily. The boost circuit we designed on Page 7 can do it just as well, but the Joule Thief circuit is nearly free since you can make it from spare parts. The temperature difference required was so low we could put a piece of ice on one side and our hand on the other to generate enough power to light a white LED (60F difference). It gets cold though, so body heat can't power it indefinitely. You can instead leave the hot side down on a radiator and its a free night light whenever the heat is on.
Step 6: Other Ideas (Fire Powered Lego Car)
This car is a dangerous fire hazard. Ours can even go faster but common sense told us, "Hmmm....a rolling ball of fire might be dangerous and Lego Bricks don't mix well with fire."
Good thing we didn't listen.
Check it out
Step 7: Schematics, Power Specs, Other Things for Designers
I included a schematic of the first boost converter design MAX856. It was good for a LED flashlight or charging a super cap but not powerful enough to charge an iPod.
I am not going to specify the parts we used since we don't plan to continue using it.
We are looking towards using the MAX1522-MAX1524 from Maxim IC but it will take time to source new parts and we don't want to give out an untested schematic. Any recommendations for a good efficiency boost with a 0.8V to 5V input range? Needs to be over 85% efficient for a 500mA output.
Step 8: Frequently Asked Questions
1. Is a seebeck the same as a peltier unit?
Almost. You can use any thermoelectric in either way but a peltier is optomized to cool while a seebeck is optomized to generate power. A peltier could be used to generate power but as our video shows its not as efficient as a seebeck is. It took around 700F Delta to generate 5V with a peltier, the seebeck only required 200F. You can buy peltiers on Ebay for a lot cheaper than a seebeck. To get the same power you will have to use more peltiers so it would mean you would need more heat and more cooling. that may defeat some of your efficiency. Plus at 700F the wires tended to melt off for us.
2. How would you improve the design and draw more power?
Well just like batteries or solar panels you can stack more in parallel to draw more current, or more in series to get more voltage. If manufacturers could make these in flexible strips we could wrap them around pipes like the engine exhaust and that could generate some serious power. See the extras page for more info on this.
3. How is a blowtorch green?
It isn't, that is why we recommend using a wasted heat source that would otherwise be vented into the environment.
4. Do you need a heat sink?
Not always, but it will work better. They won't generate the full power if you don't heat sink the other side.
5. Are you really planning to make a store out of the laser cutter?
Yes we are, we were planning to launch www.splitreaction.com on July 20th as a blog, but since our main idea fit this Instructables contest we decided to jump start and rush to finish it to make the deadline.
6. Why does your current boost circuit not work for iPods?
We built that boost circuit for a super cap powered LED flashlight. Its only designed for 100mA. It will allow a charged super cap to output 5V steadily from 5V to 1.8V input. Since I-pods and phones need more current our current boost circuit wont charge them. We do plan to improve it.
7. When will you be providing the circuit board design?
Hopefully by July 20th our full blog will be up and we will have perfected a boost circuit.
8. Will a Minty boost work with the thermoelectric element?
Maybe. We don't own one. Perhaps we can collaborate.
9. OMG these seebecks are expensive.
Yes they are, but over time they compensate for the initial cost of the purchase. Not to mention, the prices should drop dramatically when large manufacturers see the benefit and start creating them in larger quantities.
10. Why do you use a regulator?
The voltage is very unstable and varies with changes in the temperature, without regulation the element could go way beyond what you want and possibly damage your device.
11. Can it charge a super cap?
Yes it can, and with a efficient boost regulator it could power almost anything you could think of.
12. Are you going to sell kits?
Maybe if enough people request it we can order parts in bulk and kit it out.
13. Will the $15 350W Peltier from Ebay work? It seems bigger than the 6W you use.
Well it will... kinda... but not as well as a seebeck one would. They use different metals to make them. Also the 350W number is how much energy it consumes to cool, not how much power it can generate.
14. I don't know how to solder or do any of this, but I still want to try it out. Where should I start?
It's not the easiest thing to do, but start off by browsing Instructables. They have plenty of How-tos that will build you up to the expertise you need to try this one.
Step 9: Extras
This video talks about an idea that is very similar to what we have discussed in previous steps. We think this is great, but we think we can do better.
Catalytic converters are already pretty efficient. They need to warm up in order to operate at full efficiency. If we attached seebeck units it may draw away heat needed to warm it up. So lets just let them do what they do best (reduce toxicity of emissions).
Our idea is to attach seebeck units to the exhaust manifold. By placing the seebeck units on the manifold we get the hottest exhaust region. More heat differential means more electric power generation. Also being inside the car would protect it from the elements.
With future improvements, we would like to have seebeck units in a flexible strip. They could be wrapped around exhaust pipes and then perhaps the technology could replace the alternator entirely, even free up some HP for the engine.
It's too soon to tell without some more testing.
Note: Although these units are very rugged, they are unfortunately not bullet proof. We recommend making sure the cold side is cooled for best results in efficiency and protection of the units under extreme conditions.
Step 10: Reader Suggestions: Joule Thief
A Joule thief is a transformer feedback single transistor inverter. We found some details via hackaday on this site. Apparently it is a circuit a lot of people are using to drain the last remaining volts out of a battery. The one we built worked down to about 0.30V.
Here we have it using the waste heat from our toaster oven to turn on a LED
We even tried to use the circuit with our body heat and ice for the cooling side. It worked but it sure made us feel cold 'cause it drew our body heat away. Yikes!
Step 11: WALL-E
Step 12: Practice With a Paper WALL-E
Step 13: Download the Paper WALL-E and Double It in Size.
We found a roll of sheet metal left over from when my roof was repaired. It's dirty, it's rough, it's so perfect for WALL-E. After we printed the pieces we cut them to fit the sheet metal, making sure all the bend locations were the same. We used very little of the spray-on glue since we needed to peel the paper off later on. We tried to tape but it wouldn't stay.
Step 14: Cut It Out
Now when you cut the pieces out you will need to use gloves and goggles. The edges will be very sharp.
Note: we removed the front chest piece for WALL E and the inner side of the treads. We needed this space so it was sacrificed. If we were to do it again we probably would have removed the back piece. Originally we wanted the front door to open but the altoids tin proved to be too large.
Step 15: [bend It / JB Weld It]
You should be able to use your experience from the paper model to know which parts to bend.
We used JB weld and a lot of clamps to attach everything together. JB weld is a type of Epoxy adhesive. Its $5 at Home Depot and can handle the high temperatures we are working with. It takes about 24 hours to fully cure so you need to be patient. We also used it as a filler in some locations since the model has gaps on the treads.
Step 16: Paint It
Step 17: Build the Chassis
Step 18: See If It All Fits.
Step 19: Cut a Lego Wire
Step 20: Assemble Thermogenerator
A easy way to make a heatsink is to use the top cover of an altoids tin. We attached our cover to the peltier with mini binder clips. We added some thermal grease between the cover and the peltier. To make sure the cold side is cool enough we put 1 ice cube on top of the cover. The ice cube melts so don't put more than one; the tin can't hold that much water and it will spill when the ice melts.
The two Peltiers are attached in series.
(make sure the hot side is facing down)
Step 21: Assembling the Microcontroller (optional)
If you are daring enough to consider programming it here are some useful references. We seriously recommend prototyping the system on a breadboard before you solder it. Different peltiers generate different voltages.
We used the Pic kit 2 and programmed it in the development board then pulled the PIC out and placed it into the WALL-E.
We won't be going into the details because its big enough to be an instructable of its own. In fact it is! You can search instructables.com for how to program PIC micros.
Our WALL-E PCB has two 2n3904 transistors that drive the two motors. We have 100 Ohms resistors connected to the base of each transistors. Two output pins of the PIC turn the transistors on and off. This allows wall E to spin or go forward. To get WALL-E to drive backwards would require a full H-bridge.
Step 22: Assemble It All
Step 23: Make a Candle
If you see that your WALL-E is struggling to move you can try adding more then two wicks. Be careful, the Lego Bricks could melt or the wires could come off if you make the candle burn too hot.
Make sure your prototype works initially. We used a stove to test because it was reliable and consistent. We only did that test twice to measure the power. Once you get an idea if it works you can stick to just using a more efficient standard power supply to power it.