Electronics > Projects, Designs, and Technical Stuff

Limiting inrush current on supercaps

(1/2) > >>

philpem:
Hi guys,

I'm designing a circuit which uses a 5.5V supercapacitor (charged to ~5.45V) to allow an MCU to capture and store a few minutes of data when power is lost. My initial tests worked fine on the bench supply, but on a prototype the inrush current is overloading the power supply and causing it to enter a limiting fold-back state.

After the supercap there's a high-efficiency dual-mode switchmode converter (it can adaptively switch between buck and boost modes depending on input voltage).

The obvious solution is to add a series resistor - I calculated 120 Ohms to limit inrush current to ~50mA. The catch is, when the voltage starts to drop, the converter enters boost mode and the voltage drop over said resistor in a discharge condition reaches a point where the SMPS output becomes noisy. Eventually the voltage drop is so bad that the SMPS shuts down even though the supercapacitor voltage is still above the SMPS shutdown voltage.

My next thought was to bypass the resistor with a Schottky diode -- but this didn't make much difference.

How can I limit the inrush current to the capacitor while maintaining a low-resistance path to the SMPS and from then on to the load?

I was thinking one of the MOSFET-based "reverse polarity protection" circuits might be a good starting point -- though I wonder if there's a better way?

Thanks,
Phil.

madires:
Maybe something like the circuit below. You could also replace D2 with a wire.

T3sl4co1l:
Active current limit.

Ideally, from another switchmode controller, if you're real tight on efficiency.  Otherwise, something simple like a PNP current source can provide inrush limiting while supplying enough current to allow the subsequent dual-mode controller to behave itself.

There are two kinds of SMPS behaviors, I suppose; those that limit current softly (continuous, stable operation in current limiting mode) and those that don't (bang-bang limiting).

The latter are often the offline flyback/foward converters, which are self powered; short the output, they stop being self powered, and they go putt-putt-putt all day.

Onboard DC-DC type controllers are usually the "stable" kind, but always check the datasheet to make sure.  That dual-mode converter could have the same instability, for the same reason?

Tim

Kjelt:
I use this for backupbattery support for the RTC in my project.
Make R3 somewhat lower if you want faster charge. In my case i don't need a fast charge since the pcb is 24/7 on.
I thought and learned that it is good practice to put a series resistor for a goldcap since otherwise the too high inrushcurrent would also damage the goldcap.

Tinkerer:
Dont know the strength of the inrush, but would a thermistor be out of the question?

Navigation

[0] Message Index

[#] Next page