Generating Electricity from Heat to run a Prop

JohnR2010

I'm looking for a power source that would generate 10 to 200mA at 2.8v from heat to temporarily run a propeller. I'm working on a fireplace flew temperature monitor and I would like to pull the power straight from the heat in the flue. I have done some searches on the Internet and man there is a lot of Smile out there. I don't want a do it yourself science project, I'm looking for a reliable solution. I thought I would post a question here and see if the forum has any recommendations on a place to start.

Thanks!

Publison

This is proven technology that you could reverse engineer. :

http://www.biolitestove.com/products/campstove/

Not cheap though to extract the components.

LoopyByteloose

200ma at 2.8VDC is 0.56 watts.

Could it be that thet TEG (Thermoelectric Generator in that BioLite device is simply a Peltier effect junction that is being used in reverse?
It is called the Seebeck Effect.

http://en.wikipedia.org/wiki/Thermoelectric_cooling
http://en.wikipedia.org/wiki/Thermoelectric_generator

If so, just DIY a few to the side of the stove.

Tracy Allen

You have to have a large temperature difference in order to extract that kind of power. The heat source has to be in the hundreds of degrees and the heat sink has to be fan cooled. Fastening something to the outside of the flue will give you maybe a 50°C differential, and only a few milliwatts of power. A heat exchanger inside the flue could do better, but you'd have to experiment. There are industrial things that do this but I don't know of any you could just go out and buy for your fireplace flue.

Here is a photo of the Biolite campstove teg arrangement. You can see the copper stake that goes into the hottest part of the fire and projects through an opening in the stove body. The insulating pad sits against the outside of the stove. Heat is carried up in intimate contact to one side of a TEG while the other has a form-fit fan cooled heat sink visible through the slots. The outflow of the fan goes into the firebox to help oxygenate the fire. Electronics step the voltage up to 5V for the USB charge connection. There was a lot of engineering that went into that.

kwinn

Not sure if these would work for you, but there are thermoelectric modules that hold a pilot light gas solenoid open as long as the pilot light stays on. Try your local hvac distributor.

tonyp12

You say you need 200mA, what is if for?
Radio Transmission I guess.

Are you trying to sample the temperature and send it with Radio to receiver?
If this is the later, a CC430F5123 with two AA batteries should last 5years, so maybe go that route?
It have built-in CC1101, but I think you want Zigbee so a maybe: http://www.ti.com/product/cc2530/description

Or do this:
https://devworksinprogress.wordpress.com/2011/06/18/project-zigbee-temperature-sensor/

JohnR2010

The closest thing I can find so far is http://tecteg.com/store-thermoelectric-power-module-selection-purchase/, their lowest unit is $38 and at my temperature difference I don't think one will generate enough power. I will keep looking. I'm fairly sure I can operate down around 40mA (.112watts). I can get by if I can harvest enough power to power up the propeller every 10 minutes or so. It doesn't have to be running all the time for my application I was planning on putting it to sleep so I will only be drawing around 75uA when sleeping.

With an energy harvest circuit and a Ultra Cap I wonder if I would be able to build up a charge to power up the propeller and have it check in. I have zero experience with an energy harvester circuit. I see Sparkfun has one that can output 3.3v https://www.sparkfun.com/products/9946, does anyone have input on this? I also see Linear Technology says they have ICs that can harvest energy from a Thermocouple. I was planning on using a Thermocouple to measure the temperature anyway. Looks like I have my reading materiel for the afternoon.

JohnR2010

Tony, that is exactly what I'm trying to do. With 3 AA batteries I'm currently getting a year or better life with a propeller and a xBee ZB (as an FYI it is using a LDO regulator YOU RECOMMEND in a previous post THANK YOU its working great!!)

I hate to make the jump away from the Propeller but it is on my list to do for the coming year. When I make the jump I was planning on doing it with my first Bluetooth project and I see TI has a Bluetooth System-on-Chip CC2540. I will also take a look at the cc2530. Even talking about this makes me feel like I'm cheating on my wife?? I do love her so!!

Thanks Tony!!

tonyp12 wrote: »

You say you need 200mA, what is if for?
Or are you trying to sample the temperature and send it with Radio to receiver?
If this is the later, a CC430F5123 with two AA batteries should last 5years, so maybe go that route?
It have built-in CC1101, but I think you want Zigbee so a maybe: http://www.ti.com/product/cc2530/description

tonyp12

My Best guess, A fireplace that is used for 2hrs a day for 4months out the year.

Thermistor hooked up to the Enable on LDO, if temp is low the whole system is shut off.
http://mouser.com/Passive-Components/Thermistors-PTC/_/N-796na?P=1yzmno7Z1yzeki4Z1yzekj6&Ns=Pricing%7c0
Maybe have to apply additional 100K resistance as to get Thermistor standby down to 1uA

When fireplace is in use, Send data for 1second 60 times an hour at 50mA, sleep the other 59minutes an hour at 0.1mA
Equal an average of 1.87mAh a day, 224mAh for the season.

1500mAh batter should last 5years accounting for a 2% self-discharge , use alkaline or non-rechargeable (primary) Lithium batteries

Sir Gawain

Here is a link to Mr. Teslonian. It is the first of many for a completely homemade manner of getting electricity from heat.
He has follow up videos on a grander scale. I really like this guy !!

https://www.youtube.com/watch?v=zzEK5eODr6I

Tracy Allen

I agree about using a battery, and conserve power between transmissions.

Here is a photo of an arrangement we made here (in cooperation with Johnathan Cedar, founder of Biolite) to capture heat energy to run a data logger with wireless from the heat of a flue. An aluminum block is cut with a 4" radius to fit tightly the contour of the flue, and the TEG is sandwiched between that and a heat sink in a chimney arrangement to help with convection cooling. The data logger with XBee is fastened on the outside of the chimney.


With the fire running full tilt in this particular stove, the surface of the flue went up to ~100°C, but the hot side of the TEG only up to 70 something, and the temperature difference across the TEG with the passive ventilation hovered around 20°C. Both the voltage and the current into a matched load are proportional to the temperature difference across the TEG, so the power is roughly proportional to the square of temperature. The power output approached 50mW peak, but during most of the stove operation was closer to 10 to 25 mW. Fading to less than 1mW as the fire went out.



We tried a combination of chips to harvest that energy, LTC3105 for the peak production, LTC3108 for the long tails, down 50mV from the TEG, and the LTC4071 for LiPo charge management. In the end, the TEG did not produce enough to run the logging and wireless protocol we wanted. The project ended up relying on a larger battery pack.

LoopyByteloose

I would be tempted to use a small steam engine tied to a DC motor to generate my own. After all, there are a lot of tiny model steam engines that would work nicely in this context. About the only real concern is keeping the engine supplied with very clean fresh water.

Of course, you might consider something more exotic.. a Sterling engine. They actually seem to operate well on a small scale. Some have actually been commercialized in the range of a few horsepower. The problems of engineering become more and more substantial as one tries for greater power outputs.

http://www.ebay.com/bhp/stirling-engine

en.wikipedia.org/wiki/Stirling_engine

https://www.youtube.com/watch?v=s2I9OoZs4Wc

Heater.

Exotic? Us teenagers were building Sterling engines in school back in 1970 something. A standard metal shop exercise in lathe and mill work. And in our case sand casting.

With a blow lamp on one end they generally managed to turn over. The better made ones, like mine:), achieved quite a lick. But I still can't imagine getting enough power out of them unless they are somewhat bigger and better engineered.

But as we are talking flew gasses here that implies a lot of heat and high temperatures are available. Which Stirling engines love of course. So perhaps that is a solution. It's not going to be simple or cheap though.

Duane Degn

Tracy Allen wrote: »

the temperature difference across the TEG with the passive ventilation hovered around 20°C.

Tracy, It was very interesting to read about your experiment. This subject comes up from time to time here on the forum but this is the first time I've seen some hard numbers of the power output of one of these devices. So much for the idea of putting a TEG generator behind solar panels as I've seen suggested.

Did you try actively cooling the cold side of the TEG? With the small power output I think it's unlikely active cooling would generate more power than it required but as you pointed out in post #4 the Biolite gadget uses active cooling. Of course the Biolite has its hot side pulling heat right from the heart of the flame.

Thank you for your posts. I found them very interesting.

JohnR2010

Thanks Tracy I was thinking someone would have tried this before. Thanks for the detailed response you have saved me a great deal of time!!

CuriousOne

LTC3108, when connected to standard 40x40mm 12v peltier module, module lies on table, and you put two fingers on it, generates about 1.5vdc @0.2mA. When module bottom is put into running water, and module top is painted black and exposed to direct sunlight, it outputs abot 5VDC and 5mA current.

JohnR2010

Thanks for the LTC3108 reference I did a quick search and found this chip http://www.linear.com/product/LTC3108. Do you have more details on the 12v peltier module. There seems to be a bunch of these on eBay with various descriptions. Do you have a recommendation for a model number and distributor? Can you combine the peltier modules to feed one LTC3108?

CuriousOne

I have all data, as well as PCB drawing and part numbers at work computer, will post them tomorrow.

Modules for sure can be combined - actually a single module is a array of peltier elements wired series-paralel.

JohnR2010

Awesome looking forward to it!

Tracy Allen

Here's another type of stove. This is the kind that has a copper bar going from the one side of the TEG right into the heart of the fire, and forced air cooling on the other side of the TEG. The air used for cooling then goes into the fire box to create a draft circulation to make the fuel burn hotter and cleaner. You can see the circulation in the flames.

This is the stove in operation in our parking lot. The collaborators in the photo are Ilse Ruiz Mercado, now at UNAM, and Jonathan Cedar founder of Biolite. This was an early prototype for Biolite, and the casing of the TEG unit on the side of the Stovetec rocket stove was 3D printed. Jonathan was helping us with a UC Berkeley project on measuring performance and adoption of improved stoves. Here we were testing the data logger and the range of the XBee wireless.

I took the stove on a trip up to Yosemite Nat. Park and this is a graph of a typical cooking session. Once the fire gets going, it produces enough power to run the blower on its own. This was also hooked up to the data logger (Prop with 16-bit ADC for thermocouples) and XBee wireless to view real time data. The temperature difference from hot to cold side of the TEG in this case hovered around 130°C during the cooking (orange line), and generated up around 3.5W (blue line).


Last year Biolite made a TEG firepit that generates 1000W to power a Christmas tree in Brooklyn.

JohnR2010

This is awesome! Thanks! 1kw from a firepit is very cool!!

Tracy Allen

Here is more on the LTC3108. Curious One, I'd like to hear about your experience too. I can say that the LTC3108 is not suitable for use with a 12V source though. It reaches its maximum power output at about 0.8V, and it will never provide much more than 5mA. It is meant for low-level energy harvesting, as from a single solar cell or a TEG operating in the range of 5 to 40 °C differential.

Here is my circuit board and schematic. This also includes an LTC4070 LiPo charge regulator.
........



And performance data...


This is fine for a µpower data logger even with an occasional high power burst. A 10mA to 200mA requirement would definitely have to come in bursts to meet the average that could be supplied by this chip. The LTC3105 is another harvester chip, capable of higher output and MPPC starting at about 250mV input.