enter image description hereThis query is for a college project. The Source can give 0.3 V to 0.8 V. Need to boost the voltage to 12 V, so 6 x NiCd batteries can be charged through LM317 CC charger. The converters available are from Pololu.

  • Can the boost converter be used in series as shown?
  • If not can somebody suggest an alternative?


Hi ,

Sorry for the confusion I have created. First time I have used the forum, and it is learning time. Now I will put the project in details and you people help me out to take it to the logical end.

The Energy source for the TEG is pure sunlight, and there are eight TEG’s in series. This TEG Hot Side are mounted on to an Aluminum Plate 3mm thick (300mm x 140mm). The Cold side is also fixed to an aluminum sheet (120mm x 85mm x 3mm) and also two heat sinks are used to help in the cooling. I have used Thermal paste in between all mating points. There are no other energy source and the idea is to harvest the sun energy and charge 6 * Ni –Cd Batteries (1.2v, 2500mAH), which in turn drives an Arduino and two servo Motor. Also The Hot side plate is enclosed in an Acrylic (plexi) Box, to enhance the temperature of the plate. (Green House Idea). The sides of the box are fixed with reflectors. (Will attach photos for clarity) The cold side is outside the box. (At ambient temperature)

The Setup sits on a dual axis tracker, so that the hot side plate is always perpendicular to the sun.

I have not measured the temperature on the heat sinks as of now (no access to the measuring instrument). I am planning to fix the Dallas DS18B20 on both the hot side and cold side to take reading using Arduino.

The voltage I am getting is approx 800mV from the series connected 8 no’s TEG. (Also I have a same setup where I have used TEC 12706, 8 no’s in series and I am getting approx 700mV).

The specification of the TEG & TEC is way higher from what I am getting. The temperature difference I am assuming might be around 15 – 20 C. will check ASAP and update. Can you people guide me, if I am doing something wrong? enter image description here

  • 2
    \$\begingroup\$ Can the boost converter be used in series as shown? Shown where? Boost converters have basically 3 connections: input, output and common ground. They're not just "series connected". You will have an extreme challenge to boost even 0.8 V (let alone 0.3 V) to 12 V and do constant current charging using an LM317, the LM317 will probably consume most of the energy. To me your project sounds like a challenge even for an EE Master student. If you know little about electronics it looks like just currents and voltages. Reality is more complex though. \$\endgroup\$ – Bimpelrekkie Jun 3 '19 at 10:39
  • \$\begingroup\$ Hints there: electronics.stackexchange.com/questions/439761/…. But this is certainly much harder than you imagine. What is this source, by the way? Are you sure it can provide enough power to actually charge anything? \$\endgroup\$ – dim Jun 3 '19 at 10:40
  • \$\begingroup\$ I had added the block diagram, but was not showing up, so i have attached them now for clarity. sorry for the confusion \$\endgroup\$ – SNEHA Jun 3 '19 at 10:50
  • 1
    \$\begingroup\$ HELP US TO HELP YOU !!! | Put ALL available info in your question., Spreading it through the comments makes it hard for everyone. If you have 9 TEGs in series SAY SO. If you get 0.8V with 8 in series say so. If that's foir ONE teg SAY so. | You MUST tell us your delta temperature on each TEG. And if energy input is limited in some way. | Without a proper picture of what you are attempting you cannot expect the best possible answers. \$\endgroup\$ – Russell McMahon Jun 3 '19 at 12:50
  • \$\begingroup\$ There are a number of different boost converters designed explicitly for TEG systems, like analog.com/en/products/ltc3108.html, which can in theory run from as little as 20mv. You may want to investigate this sort of boost converter \$\endgroup\$ – BeB00 Jun 4 '19 at 7:42

0.3-0.8 is really low voltage but, on the other hand, no other info are provided therefore i assume that we can draw as many current as we wish.

Forget to do something at 0.3V but starting from 0.5V we can harvest something. The U1V10F5 can operate from 0.5V, with a max current of 1.2A. The output voltage is 5V, and the efficiency is between 70%-90%, we can assume 70%.

1.2A @ 0.5V are 0.6W in input, considering the efficiency we have 0.42W at the output.

From 5V we have to increase the voltage again, the U3V50F12 will increase the voltage up to 12V with the highest efficiency i've found so far in the pololu catalog: 85%-95%. We assume again the worst case, it won't be far from that in reality.

0.42W x 85% = 0.375W 0.375W @ 12 = 0.03A (approx)

I don't get why you want to use a linear CC charger with only 30mA of available current, for a NiCd battery pack, which is one of the most resilient cell chemistry.

Edit: now that you attached your diagram this answer doesn't add much except few considerations on the current. You are considering only the voltage. To have 250mA at the LM317 input you'd need, assuming 80% efficiency in the LM (which is not really conservative, imho), more or less 9A @ 0.5V. Are you sure that your power supply is anywhere near this spec? (not to mention the unavailability of pololu converters that can handle this much current)

  • 1
    \$\begingroup\$ ps: even if the LM317 is able to stay operative, you'd need half of eternity to charge an electric-drill-sized NiCd battery pack with this device. \$\endgroup\$ – theGarz Jun 3 '19 at 11:04
  • \$\begingroup\$ Thanks for the info. \$\endgroup\$ – SNEHA Jun 3 '19 at 11:35
  • 1
    \$\begingroup\$ Thanks for the info. The source is TEG SP 1848-27145, 8 nos in series, which is giving 0.8v and 250mA approx. Now using the first boost module U1V10F5, and efficiency of 70% can i charge the ni-cd battery directly, without second stage booster and LM 317. i can get an output of 5v @ 170mA, which should charge the battery in 14 hours. the battery will not be fully discharged and only running an Arduino and an servo and the time for charging is not an issue. i calculated the charging voltage for 6 batteries at 15% of the rated voltage and so the idea of second stage booster. \$\endgroup\$ – SNEHA Jun 3 '19 at 11:44
  • \$\begingroup\$ Mmm... no. 70% of 250mA is 175mA, ok, but you still have to transform 0.8V in 5V. Considering the 70% efficiency, you'll have ony 28mA @ 5V. Conservation of energy (and therefore power) is stil to be considered. :D Moreover, how can you charge 6 NiCd with 5V? They can be considered more or less charged at 7.5V, not 5. \$\endgroup\$ – theGarz Jun 3 '19 at 11:59
  • 1
    \$\begingroup\$ @SNEHA I don't know where you got your specifications but, fortunately :-), they seem to be wrong. Your specs are for ONE device, not 8 in series UNLESS you are running it from body warmth or similar. This webpage and many like it give outputs of 0.97V at 225 mA for 20 C delta temperature. 8 TMGs would give about 8V at 225 mA. | You need to tell us the energy source and expected delta temperature. | Even if the spec you gave is for one TEG you'd 50 mAget about 6V at 250 mA. \$\endgroup\$ – Russell McMahon Jun 3 '19 at 12:48

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.