1
\$\begingroup\$

I am supposed to design a synchronous buck converter (to be driven by micro controller control signals for switching purpose). The input Voltage range is 10V-15V, and output voltage range is 4V-5V but with high current i.e around 200 A (supposing i have a power supply capable of providing me such high current). Also efficiency of this buck converter is not an issue in my case (it can be low). Whereas the purpose of buck converter design is to route power in both directions (from i/p to o/p (buck) and from o/p to i/p (boost)).

I have read online that it is better to use transistor package modules instead of using discrete transistors and diodes. I also went through MOSFETs, BJTs and IGBTs.

For handling a high current at the output of my synchronous buck converter, i need the switching device (MOSFET, BJT or IGBT) to be of that currnet rating. Now what i have found so far is that MOSFETs (or MOSFET/buck modules with MOSFETs) with such high current ratings are not available. For eg Which led me to search on IGBT modules, which are available for high current ratings. for eg This Link gives a number of different IGBT modules which are available.

Now my question is, that am i right assuming the fact that i cannot find such MOSFET modules with high current ratings and i will have to look for IGBT modules ? If i am, then what kind of IGBT module would be suitable for me to design a synchronous buck converter ? While going through These different modules, i found This type to be the most appropriate, looking at its transistors arrangement inside the module. If i am right, would i be able to design a synchronous buck converter (capable of both current polarities) using one of these module ?

And if i am wrong, can anyone recommend me MOSFET modules for buck converters with such high current ratings, because i couldn't find one.

Thankyou.

\$\endgroup\$
  • \$\begingroup\$ Cost target? Space target? Bottom-line performance target? Any forced air cooling? \$\endgroup\$ – Andy aka May 15 '15 at 16:10
  • \$\begingroup\$ For now, no cost and space target, as in these factors doesnt matter, though while going through IGBT modules, i found out that they are pretty expensive, but for now i am finding the possibilities of performing my task without considering the space and cost. As far as forced air cooling is concerned, i would really like if i could skip it, but i guess it wont be skipable and would have to be considered in my design keeping in mind such high currents. Bottom line performance target = somewhere around 60%-70% \$\endgroup\$ – yiipmann May 15 '15 at 16:22
  • \$\begingroup\$ If your circuit is 66% efficient and it's delivering 1 kW to the load, then it will need to dissipate 0.5 kW. That will require some substantial heatsinks and/or active cooling. Better efficiency is likely to lead you to a cheaper solution. \$\endgroup\$ – The Photon May 15 '15 at 16:49
  • \$\begingroup\$ Why do you need the circuit to operate in verse i.e. "capable of both current polarities", "route power in both directions" etc.. Can you be clear about this? Also, what is the load(s) and if multiple loads have you considered different tactics? \$\endgroup\$ – Andy aka May 15 '15 at 17:36
  • \$\begingroup\$ It is supposed to be a battery charger, which will first charge the battery from grid. and after that, grid terminal will be replaced by a power dissipating resistor which will help drain out the battery. Main purpose is to monitor battery behavior during charging and discharging. Battery is LiIon-> approx 4.4V, whereas current could go as high as 200A for a 20Ah battery, being charged/discharged at 10C for e.g. So during charging cycle the load is the battery and during discharging cycle battery is the supply and resistor is the load. \$\endgroup\$ – yiipmann May 15 '15 at 18:27
2
\$\begingroup\$

IGBTs have high voltage drop and are not suitable for low voltage switching. They only exist because reliable high current high voltage MOSFETs are hard to make.

You should use high current MOSFETs rated at 30~40V. Maximum current is often package limited, but several FETs can be wired in parallel to increase current handling. For example the IRFP7430PbF is rated at 404A, but package limited to 195A. Rdson is 1.3 milliohms, so two in parallel should drop 0.13V at 200A (= 13W per FET). In comparison an IGBT module could drop 2V or higher, resulting in 400+ Watts of heat to get rid of!

The advantage of a module is less wiring and simpler installation. One disadvantage is that it's toast if even one transistor dies. Also an IGBT module may be much more expensive than a few discrete FETs.

\$\endgroup\$
  • \$\begingroup\$ Ok. What if i lower my specifications, say a synchronous buck converter giving an output of 5V, 40A. And i charge my battery with this synchronous buck like this one . Will i be able to discharge my battery using the same path, and by replacing a high wattage Resistor at the VIN pin of the IR3553 IC instead of my power supply connection(which was used to charge the battery) ? I mean its a synchronous buck converter so it is supposed to work in a bi-directional (for current) way. So will this device be able to work in that way ? \$\endgroup\$ – yiipmann May 18 '15 at 15:18
  • \$\begingroup\$ any comments please ? \$\endgroup\$ – yiipmann May 19 '15 at 15:25

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.