Energy Harvesting Peltier Ring

[Sean] is by no means an electrical engineer, but when he discovered the magic of Peltier plates he knew he had to make a project with them. This is his Energy Harvesting Peltier Ring.

The effect he is harnessing is called the SeeBeck Effect — the process of generating electricity through temperature differentials. He has shown how peltier plates work to many people, and, as you can guess, most people think they are amazing (free energy wow!). Unfortunately, most peltier plates are rather large and bulky, so [Sean] decided he wanted to try to design something small enough that could fit on a ring. Just a proof of concept, to light a tiny SMD LED.

The tiny Peltier plate he found generates about 0.3V with a temperature differential of about 20C — not bad, but it won’t light up any standard LEDs at that voltage! He started looking into voltage steppers and discovered Linear Technology’s 3108 Ultralow Voltage Step-up converter and Power Manager — a surface mount chip capable of scaling 0.3V to 5V. The only problem? [Sean’s] never done surface mount soldering.

His first circuit was built on a prototyping board, and after it worked successfully, he designed a PCB using Fritzing. Another success! Prototyping complete, it was now time to try to downsize the PCB even more to fit on a ring. Realizing there was no way he was going to fit it on a single ring, he decided to make a double ring out of CNC machined aluminum. He made use of his school’s CNC shop and the ring came out great. It works too! The room has to be fairly cool for the LED to light, but [Sean] definitely proved his concept. Now to make it even smaller!

51 thoughts on “Energy Harvesting Peltier Ring

  1. I’m pretty sure that ring would be considered a weapon akin to “brass knuckles” in California, and most other States in the USA, and likely elsewhere.

    It’s a nice piece of fabrication, but I’d be very, very, careful where I wore that ring.

  2. I’m not allowed in my school’s CNC shop. RIP :L
    Also, Mister X This is a good vehicle through which we can test the law and potentially defend ourselves from the tyranny of bureaucracy.

    1. With all due respect SYNTRONIKS, there are plenty of other less “in your face” ways to accomplish the same goals, I’ve already tried your way, I always lost, but when I started working with authority figures in a reasonable and organized way, I was allowed to do things my fellow students couldn’t, all with the blessings of the Dean.

      I’m not implying that your way is wrong, it’s your path, it’s just that some of us have already figured out there’s a “bridge out” sign at the end and make attempts to save fellow travelers wasted time and energy. Be safe on your path, and never stop Hacking.

      1. :) Thank you — My first comment was unrelated to the second. Use of that machine shop is restricted to research projects and contracts for that one specific organization. There is a significant amount of resource wasting at my university. We have a machine shop in this very building I am in now but it is abandoned due to lack of supervision. No one wants (understandably) to assume liability for accidents.

      1. My circuit also accepts very tiny solar cells(low voltage) and piezo elements, so you could technically have them charge all day, maybe even in partially shaded areas – think calculator solar cells – then have the led come on when no light is detected. Also if the thing you stick them too vibrates a lot thats another way to generate current – Piezo.

          1. First thing to realize is that nothing about any of these forms of energy are currently “efficient” when it comes down to turning them into electric energy. But with that you realize that normally its just wasted energy in which case even an efficiency of 10% is useful.

            I have not experimented with piezos yet, but in the data sheet for the 3108 there are different setups for piezo/solar/thermo. but in general to hook up a piezo would just need a couple more components. One of the key features to making the chip work for thermoelectric is the 1:100 wound coil from coilcraft. When using the 1:100 its max input voltage is 0.5V, so if working with something larger than that you can step up to the 1:50 coil or the 1:20 coil all allowing more voltage to be provided to the chip.

  3. Awesome! As a next step, consider reading up more on printed circuit layout; it could really help you make the ring much smaller, possibly small enough to fit on one finger!

  4. With a little more tightening up of the circuit, one could build 3 smaller boards and stack them, then it could be a “real” one finger ring.

    None the less, it’s really nice to see someone go to the trouble to make this work, it’s quite an accomplishment to do more with less, and now others are seeing the potential of this and may carry on the work.

    1. Then go for it, and we can laser scan both to see which has higher accuracy.

      Also, do it in a similar amount of time.

      If you have access to a CNC why the hell not use it and save yourself time and get a better product.

    2. What can I say, I love robots doing human’s work. No Luddite here haha
      The holes had to be specific diameter.

      Also if you read my article, I think I put in there that I somehow manages to measure my ring finger incorrectly. So the dremel made an appearance to make the hole bigger.

    1. Those are incredibly small and expensive. I ordered two HV56 ( ) because I thought I was good enough to solder the 0.3mm pads on them (I was dumb and naive enough to not order the lead version). I was wrong. The first one, somehow the solder pads came right off one side, so there goes the first $100. The second one split in half due to the heat of soldering, another$100. I was very very frustrated. I will probably use one in the future and get the leads version but for right now my $40 12mmx6mm peltier from digikey will do.

      I’m focusing on designing a silver heat sink at the moment, and figuring out how to isolate the heat from my ring to it without an ugly rubber band.

      I will make it smaller.

  5. Nice job on the ring!
    This could be used to make other wearable electronics, like a wristwatch, that powers from your body heat. Or it may “charge” a primitive cellphone..

    1. The good thing about my first development board I designed is you can customize the output voltages; on top of that you can even take things like miniature solar panels or piezo elements to generate currents too. Unfortunately the amount of current required to charge a phone is huge and because the voltages are being up-scaled so much the currents are almost zero. Linear Tech. does have a new chip I’m going to look into where its sole purpose is to use very low voltages and charge Lithium batteries. I’m thinking of putting two piezo elements in my shoes and seeing if I can charge a phone via walking or running.

      I’ve experimented with my larger peltier module and a candle with two heat sinks but still wasn’t able to charge a phone. The circuit needs to be improved. Kind of put the project on hold for a moment while my finals are taking place.

  6. Make it out of gold or gold plate the next one, aluminum can leave a dark residue on the skin, you might even consider embedding the circuit between two plates to protect it and give you extra surface area for the cold side. It’s really cool

    1. Pure silver has the highest thermal conductivity and I was planning on making the heat sink out of silver. The real issue I’m dealing with now is finding a way to properly secure and isolate the heat-sink from the heat of the ring(probably just going to use M1 screws and very tiny nylon insulators). Right now its held on with an ugly rubber band.

  7. As azraph suggested, I think he could overcome the “cold room” problem a little better if it had more thermal mass on the cold side, but would require a larger heatsink area, which would mean a bigger ring. But then again a heatsink might give it more stylish flare and make it more face ripping as a weapon (in spirit with the brass knuckles above)depending on design of the heatsink.

  8. Challenge accepted. Bringing out the old germanium radio energy harvester grid and will take some pics this time. It definitely peaks at 5pm with all of the various chatter in the airwaves. Been reading these articles on HaD and I have some new ideas to make it more efficient. It used to take three days to charge the deep cycle with a trickle store feeder but that was enough to drive the small sprinkler system for its short spring cycles. After the initial charge, it seemed to be doing pretty well and then one day it just stopped producing at all and it was relegated to under the house. Probably just a bad something or other and this ring inspires me to dig it back out for another run now that I am 4 years wiser in the arts lol. Kudos to the builder for inspiring me to get off my rear and back to some energy tinkering :)

    1. Do bring out your “germanium radio energy harvester”, along with a video demonstrating it’s performance. For myself a field/signal strength meter doesn’t show much power available. Then again the closest broadcast tower is 10 miles away as RF flies, and I wasn’t using a large antenna.

  9. Anyone have an idea of the efficiency (or lack thereof) of the Seebeck effect? I know Peltier cooling devices are notoriously inefficient. I was wondering if the reverse effect was better or even worse.

    1. I had a look at this a few years ago – see

      In short – a few percent.

      That data was measured on a module which was intended for
      cooling purposes, NOT for power generation. As a result, the
      maximum hot-side temperature is limited by the melting point
      of the alloy, and the alloy used is specifically meant for heating/cooling
      purposes, not for generation. There are generator-specific modules
      available, which use better materials (higher melting point) but
      they are rather expensive, and the efficiency isn’t that much greater.

      I measured mine using a water-cooled heatsink on one side (constant
      cold temperature) and a heated aluminium block on the other with
      a thermocouple. I had previously measured the thermal resistance
      of the entire module by seeing how long it took for an insulated
      metal block to cool down via conduction through the module, and this
      enabled me to work out the thermal power flowing, and hence the
      overall thermal/electric efficiency.

  10. I was rushing all my replies and I forgot to say…

    Its a WIP, but the state its in right now is enough to give people an understanding of what is happening.

    I believe Ill make it smaller with modifications, also make the led be on more of the time, whether it be blinking or not. Using a TEG module instead of a TEC will help also add current/voltage, just need to hold off on the spending for a bit(TEGS are usually double the cost of TECs and I ruined two TEGs already – See above). The effect is great when you go outside at night, really shows off the potential(thanks cold canada).

    Thanks again, and comments welcome.

  11. Is it Cyber Troll Tuesday or something? A lot of off topic garbage today…

    Anyhow, neat project: I’ve always had interest in thermoelectric generators and this made me grin. I’m sure some of the legitimate comments above will help you to figure out a good way to minify the ring a bit further.

    Great work for a hobbyist!

  12. There is some cool stuff going on with graphene as heatsinks and energy pipes.

    search for graphene batteries , and the heatsink search will show that it can work as an inductive circuit for heat – no phys contact neccesary.

    pretty freaked

    here is a water cooler, would work better with alcohol, because everything does….

    But graphene does it without even having to be in contact.
    Was studied to transfer thru a spaceship body to the exterior.

    Not it..

  13. I reall need to have a blog open when I doing the internet thing to make notes in. I recall sometime back there was an effort by some company to use the heat of a motor vehicles exhaust to create power from peltier devices.. Working on the theory that if successful alternators could be smaller, saving weight and fuel. No doubt I bookmarked it , but I can’t find it.

  14. Cool project!
    I did something similar using a Peltier and LTC3109 to power an energy-optimized Arduino derivative (Mosquino). The big drawback I find with Peltier-type devices is that they are very thermally conductive, and a lot of heat has to “pass through” to get any work done. The upshot is that unless the heat/cold sink on the passive side is very large, it equalizes toward the active-side temperature much more quickly than it can dissipate heat/cold, so the power production when a thermal source is present is short-lived. I’ve looked briefly into thermoelectric elements designed specifically for energy generation (rather than cooling), but the costs and overall lack of hobbyist-buyable product threw me off. Do they offer any substantial advantages over garden-variety Peltier elements?

    1. The TEGs are definitely advantageous over the general run of the mill TEC module. But currently in my ring I’m still using a TEC, specifically this one..

      I haven’t done any testing on how much voltage this one is generating as I had a smaller one before so I just threw this in the ring the other day. So as I mentioned above, I attempted to make a HV-56 TEG module from Nextreme to work, it was 3.1mmx3.1mm and when I initially tested it I was generating way more voltage than any peltier I’ve used(for its size). The data sheet says about 0.16v for a 10 degree difference. Unfortunately as I mentioned above, I burned out both of them trying to solder leads to them. I’m not really sure what happened to the one but it seems that something corroded away the solder pads on one, and the other one basically split in half after a lot of frustration and heat. It was a costly mistake but I was nearing the end of the project and decided to splurge on the expensive TEG modules, next time Ill pay the extra $15 and get it with leads(what was I thinking!) – Live you learn.

      I would love to see what you were able to accomplish with your project and the Mosquino. I’m going to be looking into linears ultralow voltages Lithium battery charging chip next.

      Another problem with mine is if its really cold, your fingers don’t like to hold heat, even if you’re basically a human heater like me.

    1. Well, no, as long as the cold side keeps losing heat to the room, and the warm side keeps absorbing heat from Sean. That depends on how good the heat sink is, as well as the temperature of the room and Sean’s skin. It could carry on working forever.

      The power will presumably start off low, as Sean heats the ring up, then higher as the heat transfers to the heat sink thru the Seebeck chip. Then lower again as the heat sink warms up to it’s “hot” temperature.

  15. I Have too admit, I’m a little pissed someone finished this before me. I have been working on 3108 based ring with two hv56 units for over a year now. And now it looks like I’m just copying someone’s work. In fact i promised one of these to my gf last may. Maybe a co-op on designs? I have some ideas for you.

    1. Hey I’m always willing to bounce Ideas, I figured there had to be someone else out there with this thought in mind.
      I wish I had not broken my HV56s though. The ring would be way better right now, Ill have to wait for some funds before I buy them again.

      You’re welcome to email me directly through my account, or the IdleHandProjects account.

      I have a good idea of what I’m going to do about the heatsink now, just need to manufacture it. It involves no sharp fins to catch on pockets.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.