# How to turn a TEG output into something I could use?

Status
Not open for further replies.

#### Matienzo

##### Member
Hello there,

I'm a beginner with electronics so I apologize in advance for my basic questions.

I'm working on a project that uses a Peltier device (TEG) as the power source. The conditions under which it is used allow me to harvest about 1V and 130mA. The motor I want to power on a cycle - off and on with intervals of about 30' or a minute- works using a minimum of 1.5V and ~100mA under the load.

I think that if I could store the electrical energy in something like a capacitor and then release it with the desired voltage and current it could work.

My questions are:
How do I tell the capacitor to release the energy once it is full?
How do I transform the stored energy into a useful, constant output? 1.5v ~100mA

The main constrain: the only power I can use is the power coming from the peltier.

I'm a ME and my knowledge about circuits is rather rustic so I would appreciate if you guys/gals keep your explanations or instructions as basic as you can Any resources, references, suggestions, ideas or schematics are welcome.

Paulo

#### spec

##### Well-Known Member
Hello there,

I'm a beginner with electronics so I apologize in advance for my basic questions.

I'm working on a project that uses a Peltier device (TEG) as the power source. The conditions under which it is used allow me to harvest about 1V and 130mA. The motor I want to power on a cycle - off and on with intervals of about 30' or a minute- works using a minimum of 1.5V and ~100mA under the load.

I think that if I could store the electrical energy in something like a capacitor and then release it with the desired voltage and current it could work.

My questions are:
How do I tell the capacitor to release the energy once it is full?
How do I transform the stored energy into a useful, constant output? 1.5v ~100mA

The main constrain: the only power I can use is the power coming from the peltier.

I'm a ME and my knowledge about circuits is rather rustic so I would appreciate if you guys/gals keep your explanations or instructions as basic as you can Any resources, references, suggestions, ideas or schematics are welcome.

Paulo
Hy Matienzo,

Welcome to ETO. I see you are from Argentina care to tell us which part and put it in your profile page under location so that it displays in your member window at the left of your posts.

Don't worry about asking any question, never mind how basic- we all have to learn at some stage.

(1) Connect a high value aluminum electrolytic capacitor across the Peltier device.
(2) Use a low voltage input boost (voltage step up) converter to convert the 1V output from the Peltier device to charge a Lithium Ion (LiIon) battery to 4.1V maximum. A suitable battery would be a CR123 (3.6V nominal at 1 amp hour capacity).
(3) Use a buck (voltage step down) converter to convert the voltage from the LiIon battery to a constant 1.5V to drive the motor.
(4) Use a battery cut-off circuit to disconnect the motor from the battery when the battery voltage drops to 3V. This is to protect the battery.

The way that this configuration would work is that the Peltier device would continuously charge the battery in the presence of heat.

You could turn the motor on and off as you like, but if the battery voltage ever dropped below 3V the motor would no longer be powered until the battery voltage had risen to to say 3.6V.

A fully charged CR123 LiIon battery would power your motor via a buck converter for around 18 hours.

All this may sound complicated, and there would be some design work to get the system to operate as required, but standard passive components, chips, and modules are available quite cheaply for all the functions.

Let us know if this approach suits you and, if so, I will give a detailed design some further thought, and I'm sure that other ETO members will come up with some ideas too.

spec

Last edited:

#### Matienzo

##### Member
Hi Spec,
I just updated my profile so you can see I'm from the middle of nowhere I'm currently going to school in Arlington, Texas.

Thanks a lot for your detailed answer. You know I took apart a solar light fixture because I figured that with such little voltage from the panel there has to be something similar to what I need. There is a battery that probably does what you are saying but unfortunately I couldn't make it work with the peltier. Probably because of my very small power input.

Could I make the cut of for the motor dependent somewhat on time? Ideally the motor will rotate a rod 180° and wait x time and rotate again. I dont mind the time interval as long as we can make it somewhat constant.
You could turn the motor on and off as you like
Could this be achieved automatically? I mean without me turning it on and off.

#### atferrari

##### Well-Known Member
Hola spec,

What is the electrolytic for?

#### spec

##### Well-Known Member
Hi Spec,
I just updated my profile so you can see I'm from the middle of nowhere I'm currently going to school in Arlington, Texas.

Thanks a lot for your detailed answer. You know I took apart a solar light fixture because I figured that with such little voltage from the panel there has to be something similar to what I need. There is a battery that probably does what you are saying but unfortunately I couldn't make it work with the peltier. Probably because of my very small power input.

Could I make the cut of for the motor dependent somewhat on time? Ideally the motor will rotate a rod 180° and wait x time and rotate again. I dont mind the time interval as long as we can make it somewhat constant.

Could this be achieved automatically? I mean without me turning it on and off.
Hy Matienzo,

Thanks for your location info- it helps us a lot to know where you are: access to suppliers for components, mains voltage etc. Plus it is just plain interesting.

There are many reasons why the solar light fixture may not have worked but in theory you thoughts are correct.

Once the battery charging system is working, you can automatically turn the motor on and off whenever you want; just let us know the details of your requirements.

If you really want to get fancy you can control the whole system by a micro controller unit (MCU). Then the the sky would be the limit in terms of automatic functions. There are plenty of MCU experts on ETO.

It would help us if you could describe what you are trying to do.

spec

Last edited:

#### spec

##### Well-Known Member
Hola spec,

What is the electrolytic for?
Hola atferrari,

I can understand why you ask about the capacitor across the Peltier device because, in theory, it would seem to be redundant. The reason though is to smooth out the current demand from the Peltier device and to present the battery charger switch mode power supply with a low impedance voltage source akin to a battery. You should really use Maximum Power Point (MPP) techniques to extract the most energy from the Peltier device but in this case perhaps we can get away without it.

spec

#### atferrari

##### Well-Known Member
Hola atferrari,

I can understand why you ask about the capacitor across the Peltier device because, in theory, it would seem to be redundant. The reason though is to smooth out the current demand from the Peltier device and to present the battery charger switch mode power supply with a low impedance voltage source akin to a battery. You should really use Maximum Power Point (MPP) techniques to extract the most energy from the Peltier device but in this case perhaps we can get away without it.

spec
MPP - It seems that I should to do some reading now. Gracias.

#### Matienzo

##### Member
I figure a drawing will be helpful.
Here is the smallest and cheapest dc motor I could find that would move with 1.5v and 130mA under load.
Here is the TEG. I tested it under the conditions it will be used and it generates 1V and 130 mA.

So It would be ideal for the motor to rotate constantly but I could live with a constant interval. The heat source is used only when the milk must be stirred.

Is this specific enough?

#### Attachments

• 174.9 KB Views: 77

#### dr pepper

##### Well-Known Member
Welcome Matienzo, nice to have someone more mechanical amongst us.
This is as simple as it gets, might be of use to you:
http://www.r3uk.com/index.php/tech-tips/43-electrical-tomfoolery/203-the-joule-thief-flasher
Its a circuit to flash a led, charge is stored until the voltage is high enough then its discharged through an led to produce a flash, if you replace the led with a motor so long as load current isnt that high it might work, you'd need to replace the led with a 1n4148 diode, and put it in the circuit the opposite way round, this is to protect the transistor from spikes made by the motor.
You'd only get a short burst of motion for a 100ms or so, but if the peltier makes 100ma then you might get that to be fairly frequent.

#### Matienzo

##### Member
I was looking into joule thiefs just before I posted this thread. I only saw them being used with LED's so I put that idea aside. Also, I like spec suggestion on using ic to have more control over the motor. I'll definitely look into joule thiefs.

I contemplated a double major in EE but the math is a bit abstract for my brain.

#### Tony Stewart

##### Well-Known Member
• Hello.
150 mW is small and in any case load impedance must be matched to source generator for max power transfer during charging. Switching is done with high-side or low-side MOSFET switches. storage can be battery or Ultracap, each has pro/cons.

This is sub-optimal , what you described. Rather than chose parts, specify your all input & output power, Seeback device and electrical output parameters of load or device desired.

Please avoid implementation ideas until above is complete.

Thanks
Tony
• Hello there,
I'm a beginner with electronics so I apologize in advance for my basic questions.

I'm working on a project that uses a Peltier device (TEG) as the power source. The conditions under which it is used allow me to harvest about 1V and 130mA. The motor I want to power on a cycle - off and on with intervals of about 30' or a minute- works using a minimum of 1.5V and ~100mA under the load.

I think that if I could store the electrical energy in something like a capacitor and then release it with the desired voltage and current it could work.

My questions are:
How do I tell the capacitor to release the energy once it is full?
How do I transform the stored energy into a useful, constant output? 1.5v ~100mA

The main constrain: the only power I can use is the power coming from the peltier.

I'm a ME and my knowledge about circuits is rather rustic so I would appreciate if you guys/gals keep your explanations or instructions as basic as you can Any resources, references, suggestions, ideas or schematics are welcome.

Paulo

#### Matienzo

##### Member
Hi Tony, I'm not sure I followed what you said. Do you mean describe the problem and constrains before I choose the parts?
Could you rephrase a bit what you said?

#### spec

##### Well-Known Member
MPP - It seems that I should to do some reading now. Gracias.
No probs atferrari

Maximum power point sounds complicated but it is not. All it means is taking the optimum current from a device: solar cell, Peltier device, wind turbine etc so that that V*I (power) is maximum. The most basic algorithm is to keep increasing the current drain until the voltage starts to drop. Update 2016_07_08. This simple MPP approach only applies to PV cells. Peltier devices have a constantly falling output voltage with increasing current drain.

I do reading all the time

The more you know, the more you realise how little you know.

spec

Last edited:

#### Matienzo

##### Member
Spec I found this IC to use as a step up. Do you think it is adequate? I'm not sure how I would connect the battery and such after the IC but I'm looking into that

#### ronsimpson

##### Well-Known Member
Your Peltier device only outputs 1 volt. This is so low it is hard to boost up. Many boost systems struggle down there. Is there a chance you can use two to get 2 volts?

1V 130mA ..... can you find a 2V 65mA? A 2V Peltier could charge a large cap. A simple circuit could turn on the motor when the voltage gets to 2V and off when the voltage drops to 1V. Just a thought.

#### Tony Stewart

##### Well-Known Member
Everyone ends up chasing skinny rats into tiny holes until you specify all the parameters of your system.

Input power, desired converted power , efficiency, output power

For example motor rated power,, actual load vs rpm,,,, coil resistance and rated voltage all affect discharge rate. Start surge currents are often 800%* rated current

Same for Seeback device,,
Thermal,power input vs V & I output. e.g. Normally Solar cells are optimally loaded around 80% open circuit V while Seeback are around 50% of no load V= V open cct=Voc

ok? only specify p/n with specs of MUST HAVE components. ok?.
I trust you understand this important engineering design method.

#### dr pepper

##### Well-Known Member
I found a couple of sites that might be of interest, one is a joule thief microcontroller supply, I was thinking if you want to use a smps boost chip you could power it with the 5v regulated joule thief from the 1v, you'd need to use an ic that can drive an external transistor and a bipolar transistor at that as fets need around 10v to switch well, or you could use a logic level fet.

Heres a vid of someone using a peltier module with a joule thief type circuit.

How about buying a cheap 1w led torch, one that uses a single 1.5v battery, the controller ic will most likely run at 1v, if you robbed the guts out of one of these it might well be all you need to power the motor as white leds run around 3.5v, dont forget the back emf diode.

Last edited:

#### ronsimpson

##### Well-Known Member
How efficient is the Joule Thief? 50%??

#### Tony Stewart

##### Well-Known Member
How efficient is the Seeback generator? 1-5%.
Peltier Device inputs electrical power to device to cool.

Similar but different solder temperatures of attachments ec.