I'm trying to build a Bi-Directional buck-boost converter. The input should be between 9-12 and output should be 12. Can any one recommend a good topology to follow. I also have a few conceptual questions. If anyone can recommend good topologies to follow along with a good chip I would really appreciate that.

  1. Are all buckboost converters bi-directional?
  2. Do i still need transistors for the switches? Or will the chip do the switching for me.
  • 1
    \$\begingroup\$ Could you explain what you mean by bidirectional? Buck-boost implies the circuit block can switch voltage up or down as required, to get the desired output voltage. Is this what you refer to as "bidirectional"? \$\endgroup\$ – Anindo Ghosh Nov 25 '12 at 4:38
  • 2
    \$\begingroup\$ What I meant is that the current can flow into and out of the source. This buck-boost will be used to control current from the battery to the grid, charging and discharging the battery. \$\endgroup\$ – smokey Nov 25 '12 at 4:55
  • 1
    \$\begingroup\$ +1 for someone named smokey asking an electronics question. \$\endgroup\$ – kenny Nov 25 '12 at 13:19

Are all buckboost converters bi-directional?

A non-synchronous switching converter uses a diode for one of the switches. Such a converter typically transfers power in only one direction. (Is there a better term for this than "non-synchronous switching converter"?)

A synchronous switching converter replaces each of those diodes with an actively controlled FET. Such converters are inherently bi-directional and usually more efficient than non-synchronous converters.

Do i still need transistors for the switches? Or will the chip do the switching for me.

If you only need 1 amp or so of input or output current, there are several switching regulator ICs available that have internal transistors that do all the switching for you. For example, you might look at the LM2587T-12 datasheet figure 13, which shows how to use that chip to produce 12 V regulated output from input anywhere in the range 8 to 16 V. (Alas, it is non-synchronous, transferring energy only in one direction).

If you need higher currents, you're pretty much forced to use external discrete FET transistors for the switches. The typical circuit is shown on the first page of the ADP1873 datasheet, which shows a synchronous (and therefore bidirectional) 10 amp switching converter.

recommend good topologies

Some tips have been collected at http://opencircuits.com/Switching_regulator .


enter image description here

Simple uni-directional Buck-Boost without resonant caps using "Supercaps".

enter image description here

Bi-directional Buck-Boost using resonant caps on each switch. (More efficient but more complex dead-time commutation control)

Imagine a regenerative braking car in this arrangement with the battery voltage on the one side and motor/generator on the other side, and the demand to accelerate and brake the motor in an intelligent phase control controller (Not shown for simplicity). (There may be many parallel MOSFETS to reduce the Ron value)

...or a dual source power supply which can be charger on left and battery on right. These are some of the patented and licensed waveforms used to regulate the resonant switching commutation of the Buck-Boost high efficiency regulators.

enter image description hereenter image description here enter image description here enter image description here

You can read the patent for theory of operation. This is just intended to wet your appetite of the complexity of a bi-directional Buck Boost controller switch. Voltage and current sensing is an important aspect to optimize commutation duration when the switch voltage drop is minimal. Resonant switch Capacitors & Diodes during free-wheeling play an important part in the commutation to make it loss free.

These charts show the On state of each transistor with the thick bar, not the voltage or current. Each phase is carefuly controlled according to state of V1-V2 and power demand. Optimal range is typically 1:1 to < 2:1 for voltage ratio of the bigger side.

  • \$\begingroup\$ If assumes idealized behavior for the components (including perfect switching with no dead time), and a frequency which is sufficiently high relative to voltages and inductance, the circuit shown in the schematic will behave as a four-quadrant power converter, such that the ratio of voltages will be proportional to the ratio of duty cycles. In some ways, it's easier to understand the generalized circuit than buck-only or boost-only configurations. \$\endgroup\$ – supercat Feb 27 '15 at 22:47
  • 1
    \$\begingroup\$ The patent mentioned in this answer, US5734258A has expired today! \$\endgroup\$ – user60561 Jun 30 '19 at 17:23
  • \$\begingroup\$ Note the same circuit topology can be used to store energy for active Li Ion string cell balancers with higher mismatch that limits passive balancers to low power during CV mode with reduced bypass current. FWIW @user60561 \$\endgroup\$ – Tony Stewart EE75 Jun 30 '19 at 17:32
  • \$\begingroup\$ @SunnyskyguyEE75 I don't quite understand what you're saying. Are you describing using one instance of the circuit for each cell, or something else? \$\endgroup\$ – user60561 Jun 30 '19 at 21:55
  • \$\begingroup\$ Yes cascaded 1 per cell \$\endgroup\$ – Tony Stewart EE75 Jul 1 '19 at 0:16

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.