0
\$\begingroup\$

I am trying to develop the first stage circuitry of a kinetic energy harvester from that is made using magnets and a coil. I attach a photo of the harvester taken during development.

Photo of the Electromagnetic Harvester

My intention is to efficiently convert the voltage coming out of the coil to a DC voltage so that I can power up a TI energy harvester IC ( hopefully BQ25505 or similar) at a later state to charge a Lithium Battery.

The very first step assigned to me is to develop a good rectifier bridge. I would like to ask from the community if I can use the following IC that has a set of schottkeys integrated by TI for my purpose.

IC : http://www.ti.com/lit/ds/symlink/uc2610.pdf Internal Schematic of IC enter image description here

I have gone through the IC and it claims to give a ~0.3V drop for 1mA current.

enter image description hereI have measured my harvester to not exceed 10mA of generation. It is also one of my requirements to simulate my results soon and thats why I have chosen such an IC.

My first question is Can I use this IC for my low voltage dropout rectification perpose?

My second question is that i can notice that there are two rectifier bridges. Therefore, can i provide the same AC output of the harvester coil to both bridges to with the hope of obtaining a lower voltage drop?

Finally, what are the disadvantages of following this type of an approach? Is there a better way I can rectify the generated AC energy waveform?

\$\endgroup\$
  • 3
    \$\begingroup\$ "0.1V drop for 1mA" but at 100 mA the drop is 0.5 V, that doesn't sound that good to me. I have the PRLL5817 here which does 0.35 V at 100 mA. Also 1mA is not a lot, even at 50 V (Vmax of those diodes) that is 50 mW. Energy harvesting is a hot topic, have you done any calculations on the amount of energy you expect to harvest ? Many people don't and are then surprised that it takes a month to charge their battery. \$\endgroup\$ – Bimpelrekkie Mar 1 '16 at 14:56
  • 1
    \$\begingroup\$ Although Schottky devices do indeed have a low forward voltage, you give up reverse leakage current for it, which you will need to take into account. The reverse leakage current is highly temperature dependent (in an exponential fashion). This is not to say that this is a bad idea; just something to be aware of when you do not get all the energy you believed you should. \$\endgroup\$ – Peter Smith Mar 1 '16 at 15:07
  • \$\begingroup\$ @FakeMoustache Thank you so much for introducing me to a better schottkey diode. I will purchase a set of PRLL5817 diodes and test out for performance. I think the I,V characteristic seems better than my choice. Thank you for all warnings. they are very helpfull. I expect to harvest in the range of mw. My voltage will be about 5-6 Volts and current (hopefully) will be 5-10mA. if there is a lower drop schotkey you know, please state here. I have spent much time searching one. I endded up with that I stated in my question. \$\endgroup\$ – Denis Mar 1 '16 at 15:11
  • \$\begingroup\$ @PeterSmith: or none at all since it can go up to 1mA in worst case with that ti diode pack ^^ \$\endgroup\$ – PlasmaHH Mar 1 '16 at 15:14
  • \$\begingroup\$ You can search yourself for a lower drop diode at Farnell or Digikey (component suppliers). When you search for a diode you can sort on almost any parameter. It does not get much better than 320 mV @ 100 mA though. Note that the PRLL5817 is an SMD component ! It has no leads (wires) !! \$\endgroup\$ – Bimpelrekkie Mar 1 '16 at 15:22
2
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. Centre-tapped energy harvester.

If you double the coil windings you can halve the diode losses as shown in Figure 1. Only one diode will be conducting at any time compared with two on a full-wave bridge.

Since each winding conducts only half the time the wire gauge can be reduced for the same total output power.

It won't hurt to parallel the diodes and may result in ever-so-slightly lower voltage drop.

\$\endgroup\$
  • \$\begingroup\$ Thank you for this approach. I believe this is one of the simplst ways of donig it with minimal loss. And becasue only one diode is biased at a time this looks excellent. What are the disadvantages of this over the 4 bridge full rectifier? Do you think I am missing any half waves of energy using this approach? \$\endgroup\$ – Denis Mar 3 '16 at 5:56
  • \$\begingroup\$ Im just thinking if the voltage output will be reduced. \$\endgroup\$ – Denis Mar 3 '16 at 6:23
  • \$\begingroup\$ No. Each half coil will work on its half-cycle as though the other half wasn't there. You may have a sneaking suspicion that there's an opportunity loss because only half the coil is working at any time but we have to remember that the more energy we draw out the more energy we must absorb from the mechanical components and that energy has to come from the mechanical system. Nothing is free! \$\endgroup\$ – Transistor Mar 3 '16 at 10:04
1
\$\begingroup\$

Both of your assertions seem to be accurate. However, if you are still unsure, try it out. There is no replacement for hands on experiementation.

As for other ways to rectify, I believe this is the most efficient way to provide rectification in this circuit.

\$\endgroup\$
  • \$\begingroup\$ Truly appreciate your answer. How about active rectification using MOSFETs? Would that give me a better efficiency? I was searching about it but realised that an active rectifier pushes current from storage back to source as MOSFETs allow flow of current backwards. Should you recommend active rectification a better choice? I do not mind cost at this point. I require efficiency. Any specific best methods you know to properly apply active rectification? Truly appreciate your time. \$\endgroup\$ – Denis Mar 1 '16 at 15:51
  • 1
    \$\begingroup\$ LT makes some "Ideal Diode Bridges", which are FET controllers, but they have minimum voltage and current requirements, which implies to me that the action will be somewhat lossy. I think the idea there is more to avoid heat losses at high temperatures by leveraging the low RDSon of FETs. A quality Schottky diode will probably provide better performance at very low currents like you are talking about. In summary, my impression is that FET based rectification is for very high current applications to reduce heat losses, but at your power levels, a diode will probably be best. \$\endgroup\$ – Brendan Simpson Mar 1 '16 at 16:01
  • 1
    \$\begingroup\$ If i may ask, why are you avoiding SMD? You can solder 0805 or larger packages onto a protoboard pretty easily, and if they have the specs you need, then I would go for it. SMA/SMB package are even bigger and easier to solder than an 0805. Otherwise I don't have any specific suggestions. Mouser and Digikey have pretty good parametric search engines, so they will probably lead you to the best solution. \$\endgroup\$ – Brendan Simpson Mar 1 '16 at 16:08
  • \$\begingroup\$ I am sorry if this is not a valid reason but I am a student and I am suppressed by time. i have two days to get this to work. So, may be i can live with having a one 0.1V more drop and still finish the initial design. Afterwards, I will definitely take chose the best IC whether it is SMD or not. I completely agree with you. \$\endgroup\$ – Denis Mar 1 '16 at 17:09
  • 1
    \$\begingroup\$ If you're pressed for time, you wouldn't be able to order anything in fast enough (well, unless you want to pay for overnight shipping), so probably just go with what you have, and it will likely work OK for now. You can always optimize later. \$\endgroup\$ – Brendan Simpson Mar 1 '16 at 17:14

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.