On my other question I asked about how could I regulate a TEG like this:

enter image description here enter image description here

for a 12 V, 10 W load I have. My first idea was to use some resistors, a zener diode and a linear voltage regulator. But people said that a buck converter would be better.

The problem was that I could not find a buck converter that could handle the 96 V max input from the TEG, at least where I live.

After thinking for a while, I remembered that this TEG curve I showed is actually for 20 peltier modules connected in series so I could instead connect two strings of 10 in parallel for a resulting IxV curve like this:

enter image description here

That should mean something like this:

enter image description here

And now I can use an adjustable LM2576HV module to regulate the 12 V and allow at least the 10 W from the TEG.

The only detail I saw is that maybe the voltage from the TEG won’t be exactly the same and some reverse current might occur, my idea was to add simple blocking diodes to the output of each string to avoid reverse current, so my final circuit looks like this:

enter image description here

What do you say? Any room for improvement? for me it seems pretty ok.

Thanks in advance!

  • \$\begingroup\$ Note that when you use the CircuitLab button on the editor toolbar and "Save and Insert" on the editor an editable schematic is saved in your post. That makes it easy for us to copy and edit in our answers. You don't need a CircuitLab account, no screengrabs, no image uploads. \$\endgroup\$
    – Transistor
    Apr 10, 2021 at 20:18
  • \$\begingroup\$ could I regulate a TEG You're not asking about "regulating a TEG", you want to regulate / convert the output of a TEG. But let's take one step back. Are you suggesting that you have TEGs that generate 12 V, 10 W? I do know that 12 V, 10 W Peltier elements exist but that 10 W is the (maximum) electrical power that goes into the element and then the element will transport heat. I do hope you're not thinking that if you apply a temperature difference, that you will get 12V, 10 W out of such a Peltier element because you will not.... \$\endgroup\$ Apr 10, 2021 at 20:21
  • \$\begingroup\$ The heat to electrical power efficiency of Peltier elements is extremely poor. You will NOT get 12 V, 10 W. You will get about 5% to 8% of that, see: en.wikipedia.org/wiki/Thermoelectric_generator \$\endgroup\$ Apr 10, 2021 at 20:23
  • \$\begingroup\$ @Bimpelrekkie if the I/V plot is correct, there should be a hair over 16W available out of the TEG. Presumably it's big, and Stirling and/or steam engines should be considered, but leaving that aside -- one ought to be able to extract 16W out of it. \$\endgroup\$
    – TimWescott
    Apr 10, 2021 at 20:46
  • \$\begingroup\$ Is the heat source steady, or does it vary? You can't regulate the TEG, but if it has a steady heat input you could regulate the output. OTOH, if the heat input varies, then you need to accept that the maximum power will vary, and you can only extract maximum power with a maximum power point tracker. Trying to pull a steady amount of power out of it when the input power varies means that at times your system will just stall. \$\endgroup\$
    – TimWescott
    Apr 10, 2021 at 21:02

1 Answer 1


Put a high-power fixed-voltage shunt in parallel with the TEG. I'd set it to let the voltage rise to 50 or 55V -- that should protect your LM2576HV, and still give you headroom above the TEG's maximum power point of 48V. Then let the LM2576HV do its thing. If you use the right value for your stack of Zeners, the shunt should stop conducting before the voltage gets down to 50V. You could even put an LED in series with the Zeners as a rough indication that you're almost sucking the TEG dry.

Mind your volts and watts -- 96V can be fatal, and all the bits in the circuit below, except for R1, will dissipate considerable power. So don't burn yourself, or kill yourself. If you're not already accustomed to proper safety measures, learn. This is house-wiring voltages, so there should be resources from DIY sites if nothing else.

Note that your shunt will have to dissipate up to 15W -- I'd design the thing for 20, in high heat.


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Thanks for the answer Tim I suppose you are considering all modules in series, right? Don't you think using two strings in parallel would be better? So voltage would never go above 49 V anyway. \$\endgroup\$
    – Fábio
    Apr 11, 2021 at 0:09
  • \$\begingroup\$ Yes, this is for all modules in series. You could probably just parallel two, but I'm not sure of all of the implications of that. \$\endgroup\$
    – TimWescott
    Apr 11, 2021 at 1:01
  • \$\begingroup\$ Well, guess we'll never know if we nerver try out. I'll proceed with the parallel strings and the blocking diodes. I'll let you know if something goes wrong (hope it does not). Thanks again for teh attention! \$\endgroup\$
    – Fábio
    Apr 11, 2021 at 2:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.