DIY Stovetop 100w TEG build, design input wanted

[natdak]

I am brainstorming designs for a DIY 100w TEG that I want to build this winter. I think it will be a good solution because our primary heating source is a wood fired cooking oven that is on all day during the harshest part of winter and a thermoelectric generator will provide us with constant and reliable power when the panels are covered in snow.
What are your thoughts on a simple design?

My thoughts is to have device that will fit on one ring of the cook stove and can be placed to the side easily when not in use. The ring is cast iron with fins on the bottom that directly touch the flames. I can mill and polish the ring to make a good enough contact, if needed. Im also considering whether the tegs should be clamped with an extra plate or exposed to be in direct contact with the ring. What do you think?

Since I cant get circular TEGs I would like to approximate the shape by using square ones that can extend off of the ring a little bit. Would it be a good design to simply glue/clamp the TEGs to the underside of a sealed cooking pot with an water inlet and outlet attached? With maybe a heat sink glued on the inside..

Ive heard that you need silver solder.. What about just clamping the wires together? Or leaving the wires long and extend them to a cooler spot to make the join?

What is a good charge controller? Ive seen threads here that are indicating that solar mppt controllers work. Cheap is fine, even if its a bare circuit board I can make a box for it. It would be nice with a display telling me watts and kwh.

 

[Don B. Cilly]

Though I must say I love the idea - it would close the circle on my theoretical all-weather situation (sunny, great, the panels are charging, rainy, good, the cistern is filling up, windy, even better, the turbine is spinning, cold... well, cold, the panels are charging a bit more, so not too bad) in practice... considering a 100W TEG is over $1000 and a 100W PV panel under 100... I don't know...
-

 

[natdak]

Check out this website, Thermonamics. They sell completed thermoelectric generators and individual modules:

江西纳米克热电电子股份有限公司--Thermoelectric Generators

江西纳米克热电电子股份有限公司

www.thermonamic.com

江西纳米克热电电子股份有限公司--TEHP Series is the Bismuth Telluride based thermoelectric modules for converting heat into electricity. It is designed to optimize its performance for around 250-300C heat source

江西纳米克热电电子股份有限公司

www.thermonamic.com

I have been in contact inquiring about prices and asking some questions. This is what they had to say:


I didnt get the price for TEG-HL100 12v/24v (http://www.thermonamic.com/pro_view.asp?id=877) which appears to be a totally different design to the ones I got a quote for despite having the same model number.
They were too expensive for me, but probably quite affordable for many. So I want see if I can build one for less money.
I have asked about the prices of their individual modules and will get back to you.

 

[natdak]

Though I must say I love the idea - it would close the circle on my theoretical all-weather situation (sunny, great, the panels are charging, rainy, good, the cistern is filling up, windy, even better, the turbine is spinning, cold... well, cold, the panels are charging a bit more, so not too bad) in practice... considering a 100W TEG is over $1000 and a 100W PV panel under 100... I don't know...
-

I think it can have its place in the northern regions. I am in Sweden. 6 months of winter, 6 hours of "daylight" at the least, up to 2m of snow(that I cant clear from my panels), -30c and where Im situated I dont get direct sunlight on the panels for 4 months of the year. But at exactly that time, the woodstove is on all day, I could do with a radiator in my bedroom, the only wasted energy is the water pump, its totally passive and quiet and wood is a safe and rather dense store of energy. When considering I get the wood for free and a TEG lasts up to 100,000 hours (70 years of seasonal use theoretically) Im thinking it can be a great solution. Even better if I can make it cheaper!

 

[Don B. Cilly]

I could do with a radiator in my bedroom

Easy enough... see my signature ;·)
-

 

[fblevins1]

In your own way, you are doing exactly what I am doing, which is to make use of an existing fire source to generate energy. I know that TEG's are not cost effective if you have to make fire specifically for generating electricity, you are not going to break even, but....if you are burning the wood anyway.... The biggest problem I anticipate is cooling the cool side, because it is that delta temperature that generates the energy, if either side cool or hot overpowers the other, then you don't have the temperature differential required. I am going spend 20 watts of that generated power to operate a cooling fan to hopefully keep the tower heat sink at a cool enough temperature (relative to the hot side temp) to charge the Jackery Box which I think has a MPPT type controller on the power input for the wall or to the panels. Then in turn use the Jackery to power what I need (240 watts is it's limits) Anyway, bla, bla, bla. Moving on.

Though I must say I love the idea - it would close the circle on my theoretical all-weather situation (sunny, great, the panels are charging, rainy, good, the cistern is filling up, windy, even better, the turbine is spinning, cold... well, cold, the panels are charging a bit more, so not too bad) in practice... considering a 100W TEG is over $1000 and a 100W PV panel under 100... I don't know...
-

The actual Peltier modules are dirt cheep, my son bought me a box of 10 for well under a hundred bucks. From there I just cobbled together some metal plates, a tower heatsink and a small fan to keep air flow going over the heatsink. Not sure what the cost would have been if I had purchased all the components separately. The pictures shown were very early in my figuring out what was the best heat sink/module combination. I went with the 8 module parallel 12 volt configuration on a the brass looking heat sink with two fans and a jackery box with built in controller, ergo a multi-use power box. I went with a 1/16 inch contact plate and a 1/4 inch pre-heat plate. But I mounted the modules on backwards and my first test was a complete failure. More to come and I think I need to pay attention to this thread. Sorry about the rambling...it can't be helped...I think it's an ADHD thing.

 

[Delmar]

The biggest problem I anticipate is cooling the cool side, because it is that delta temperature that generates the energy, if either side cool or hot overpowers the other, then you don't have the temperature differential required.

If you are operating a woodstove then it must be cold outside. How do you transfer the Peltier heat outdoors?
* Mount on a long heatsink that penetrates the exterior wall.
* Liquid thermosiphon using an antifreeze solution. Water cooled computer heatsinks are easy to find.

 

[fblevins1]

If you are operating a woodstove then it must be cold outside. How do you transfer the Peltier heat outdoors?
* Mount on a long heatsink that penetrates the exterior wall.
* Liquid thermosiphon using an antifreeze solution. Water cooled computer heatsinks are easy to find.

I want to keep the cooling as simple and as passive as the set up will allow for. Mine is not for regular use but rather one of many items in my emergency survival kit. The idea is since I would have to be using a fire for cooking and desalination anyway, I'd might as well shag a little power from the process. The sun is not always shining and the wind is not always blowing so the three sets up are solar, wind and heat power generation. Once I get mine properly wired, I will know fairly quickly if low power cost means of shunting the heat away from the cool side is practical. I just got lazy and distracted once my first attempt was a failure due to mounting the Peltier modules upside down.

 

[natdak]

Very very awesome, both of you! Don, thats exactly what I had in mind, just with modules glued to the bottom of the pot so I have a lot to learn from you.
So your water pump uses 44watts of power. Thats quite a lot. does it need to be that big?
Have you considered putting modules on your pot?
What does your radiator look like and how big is it? Does one need a seriously beefy radiator to get the temperature right down? Does the temperature need to be right down?

And fblevins1: how hot are your tegs rated for?

 

[fblevins1]

Very very awesome, both of you! Don, thats exactly what I had in mind, just with modules glued to the bottom of the pot so I have a lot to learn from you.
So your water pump uses 44watts of power. Thats quite a lot. does it need to be that big?
Have you considered putting modules on your pot?
What does your radiator look like and how big is it? Does one need a seriously beefy radiator to get the temperature right down? Does the temperature need to be right down?

And fblevins1: how hot are your tegs rated for?

Not hot enough, if I can manage to keep the cold side around 20 C then we are looking at around a delta 30 C which won't be a lot of power but then again with 8 modules in parallel....maybe. Well at this point I am just going to mount again and see what happens. It was really kind of retarded how I reversed the modules in the first place. Basically I turned the heat sink upside down and forgot to stand on my head while mounting the modules.

 

[Doggydogworld]

This page has a fun video that compares a TEG and Stirling stove fan.

TegMart sells stove generators of different sizes, can give you some ideas. Maybe buy a small one and experiment.

 

[fblevins1]

This page has a fun video that compares a TEG and Stirling stove fan.

TegMart sells stove generators of different sizes, can give you some ideas. Maybe buy a small one and experiment.

Oh my gosh, TegMart is still out of stock on just about every product? Man I heard about supply chain shortages but I checked their site back in May and it was the same situation. I really want one of their charge controllers but no dice. I hope we are not seeing the end of them, because that is a very long time to not have product. Cheers

 

[Don B. Cilly]

So your water pump uses 44watts of power. That's quite a lot. Does it need to be that big?

It's not, really. I have 13kWh of LFP. But you can probably find one that uses even less.
I tried simple convection, and even a small Stirling engine to drive a pump, but in the end the electric one was simple and effective.

Have you considered putting modules on your pot?

As I said, I have, but they just seem too expensive for the power they can generate. I would probably put them on the chimney tube if I tried though.

The radiator is quite small, two elements. But one can use a bigger one, or even more than one, the stove generates a lot of heat.
It was used to heat the room, not to cool any modules. But the modules do need a "cool" side, they work on temperature differential.
-

 

[natdak]

It's not, really. I have 13kWh of LFP. But you can probably find one that uses even less.
I tried simple convection, and even a small Stirling engine to drive a pump, but in the end the electric one was simple and effective.



As I said, I have, but they just seem too expensive for the power they can generate. I would probably put them on the chimney tube if I tried though.

The radiator is quite small, two elements. But one can use a bigger one, or even more than one, the stove generates a lot of heat.
It was used to heat the room, not to cool any modules. But the modules do need a "cool" side, they work on temperature differential.
-

I guess it isnt much power for you, but in my situation trying to generate 100w, a 44w pump is too much!

The pump in this video is 5w and had 15ft of head. Its probably that making the noise in the video however. I also wonder what the device and green hose in the corner is.

So do you have 3m or so of radiators?
What pressure is it running at? Whats the advantage of having it pressurized? Is it an isolated system?

Ive asked Thermonamic for the price of their 40mm(11w), 56mm(21.6w) and 80mm(17.9w) high temperature modules and Im waiting for a reply. It might not be super expensive, I have a feeling. As shown above a premade, 100w generator costs 600$ for one, 540$ for 3 and when I asked a couple of years ago, 10 pieces were discounted to 300$ per piece!!!

 

[Don B. Cilly]

Sure, but you don't need the pump to cool the modules and generate energy.
My rig is meant to heat the room.

I could do with a radiator in my bedroom

 

[fblevins1]

If you are operating a woodstove then it must be cold outside. How do you transfer the Peltier heat outdoors?
* Mount on a long heatsink that penetrates the exterior wall.
* Liquid thermosiphon using an antifreeze solution. Water cooled computer heatsinks are easy to find.

I am starting to warm up to your water cooled suggestion as it turns out that some of the units you linked me to are reasonably priced and of course amazon was good enough to suggest a nice 4.3 watt submersible water pump. And I have to wonder which would provide better cooling, a 5 watt water cooler or a 20 watt dual fan blowing over a tower heat sink. Air is easier but water might provide faster cooling and that is important because the Peltier module is thin and even if you get the unit to cool well, it has to be able to keep the cool side cool at the place of reaction or the module doesn't work. I am going to wind up with something that looks like a rig from ghost busters if I keep this up. Thanks again.

 

[Delmar]

And I have to wonder which would provide better cooling, a 5 watt water cooler or a 20 watt dual fan blowing over a tower heat sink.

Water can absorb lots more BTUs, but air is easier. Can you duct air from a cool location? Use one fan to keep the duct air moving. Mount the duct vertical for a chimney effect.

 

[stienman]

Since I cant get circular TEGs I would like to approximate the shape by using square ones that can extend off of the ring a little bit. Would it be a good design to simply glue/clamp the TEGs to the underside of a sealed cooking pot with an water inlet and outlet attached? With maybe a heat sink glued on the inside..

Ive heard that you need silver solder.. What about just clamping the wires together? Or leaving the wires long and extend them to a cooler spot to make the join?

TEGs should have a relatively stable temperature across their surface - if you leave a portion of it "open" rather than pressed up against a heatsink/metal plate, then over time you'll find it fails due to thermal expansion and it's very brittle nature.

Clamping is preferred, again because thermal expansion means the TEG and item it's attached to will need to move relative to each other, otherwise the TEG will crack over time. I have seen some say that thermal epoxy worked fine for them, but I don't know enough about their application or yours to say either way. Use some bolts with some shoulder washers, tighten them evenly so the pressure is constant across the TEG(s), and use thermal paste to improve heat transfer to/from the TEG surface.

If you can't protect the wires from the hot side, you can use crimped connections. Make sure they are tight enough to exclude oxygen, otherwise they will fail over time with the increased temperature - even inside insulation.

-Adam

 

[fblevins1]

Water can absorb lots more BTUs, but air is easier. Can you duct air from a cool location? Use one fan to keep the duct air moving. Mount the duct vertical for a chimney effect.

Definitely going to have to have a flexible duct to move air as the device will be too close and the fans would effectively move hot air rising back down into the heat sink. But, the first part of the testing is simply engage the system, and see how many watts it will deliver to a hungry battery pack and for how long before the delta temperature can no longer be sustained and then....go from there. I don't like the design of the tower sink I am using because there appears to be too much metal mass around the base plate.

 

[fblevins1]

TEGs should have a relatively stable temperature across their surface - if you leave a portion of it "open" rather than pressed up against a heatsink/metal plate, then over time you'll find it fails due to thermal expansion and it's very brittle nature.

Clamping is preferred, again because thermal expansion means the TEG and item it's attached to will need to move relative to each other, otherwise the TEG will crack over time. I have seen some say that thermal epoxy worked fine for them, but I don't know enough about their application or yours to say either way. Use some bolts with some shoulder washers, tighten them evenly so the pressure is constant across the TEG(s), and use thermal paste to improve heat transfer to/from the TEG surface.

If you can't protect the wires from the hot side, you can use crimped connections. Make sure they are tight enough to exclude oxygen, otherwise they will fail over time with the increased temperature - even inside insulation.

-Adam

I am glad I am piggybacking on this thread. Lots of problems await our efforts. At least the testing I am doing can have the source heat adjusted and measurements made to the contact plate to see which heat setting provides a consistent surface temperature somewhere around the operating temperature of hot side. I just wish the 12 volt modules didn't have such a low maximum temperature because that makes passive cooling a tough chore as the delta temperature is marginal to begin with. I would start back in sooner on my project but I am preparing to say goodbye to my best friend of 17 years. My little Corgi mix. (insert telenovela thread ad on)