First revision of the DCM MPPT design now up

Hot off the presses! Download!

The new tracker is vastly different from the old one. The biggest difference is that the trackers now run in discontinuous mode boost. That means the inductor current hits zero on every cycle. This has the advantage of being much more efficient with extreme boost ratios by eliminating diode reverse recovery loss. The inductor is also substantially smaller, so the trackers are much lighter and more power dense.

Other new features include more LEDs, a hardware based overvoltage lockout on the output, a 16×2 LCD screen, and integrated board-level precharge. The other “big deal” is that these boards are now powered off of the solar input rather than the CAN bus. They’ll operate over the entire range from 5v to 60v on the input, but I’ve chosen input capacitors on this model that top out at 35v.

As usual, you’re free to use the design for your own purposes, but please contact me for permission if you intend on using them commercially. I’d also like to thank Robert Pilawa out at MIT for his contributions to the inductor selection process. They have a really friendly and very talented solar car team that he is a part of. Now we just need to figure out how to beat them.

  • Avatar
    Samuel Lenius
    Posted at 15:41h, 02 February


    This is Sam Lenius from the, we met at NASC2k8 in TX. I noticed that you abandoned the IR1167 synchronous rectifier chip in favor of a plain diode. In our design, we also abandoned the IR1167 in favor of the microcontroller directly controlling synchronous rectification. Our experience was that the chip would work well on the bench and then mysteriously fry once on the car, leaving the MOSFET dead short. With the new topology, we haven’t had any issues yet.

    I’m curious if your experience was the same as ours.

    I’m also curious if you’ve done any efficiency characterization with this design.

  • Avatar
    Sasha Zbrozek
    Posted at 14:22h, 04 February

    Well, the hardware just arrived, so if all goes well I’ll be able to do an efficiency characterization in a week or so. The DCM converter isn’t a good case for synchronous rectification because you never know when the output current will cross zero.

    Also, the diode has much faster switching time and no propagation delay, so everything works out better.