Having fun with the FP6276A converter

SOP-8L (EP) | 2020-07-08, 01:58:00

In my last post, I talked about the FP6276 step-up converter without properly introducing it.

According to the datasheet, it's a "High Efficiency Synchronous current mode boost DC-DC converter with PWM/PSM control".

Of particular importance is the current mode parameter. I wanted to try one of these to see how they produce work.

A connected load, the part of the circuit requiring energy, can be reduced to a single ideal resistor. Voltage-mode regulators will monitor the output through the feedback resistors in order to maintain the voltage on the load. Increasing the current consumption would incur a voltage drop since I = V/R, so the regulator will increase the voltage. A current-mode regulator will monitor the output to maintain its voltage by regulating the current drawn, I x R (load reduced to single resistor) = V.

This has some advantages since you could isolate the regulator and the regulated circuit from the input by using a coupled inductor, removing the ground pour under the inductor and expanding it to the pcb's limits, so any failures would leave the input unaffected.

From the example circuit, we learn that we're going to need at least two input and output capacitors, and three-resistors.

This circuit in particular would output 5V with a 4.4 current limit. It's specifically design to regulate a 5V line.

Remember this a current-mode regulator. This means the gate between the input and output is going to open/close, leaving the inductor to deal with the entire current passing through it, generating a lot of heat, so make sure you'll dimension an appropriate circuit for an appropriate inductor, 3-6A inductor.

I like to use some easy-solderable small parts in order to reduce the size of the board and the CDRH3D16 inductors I use are completely unfit for this job, since the 2.2 uA inductor is rated for 1.2A at ambient temperature, so I've set my current limit at 0.75A with an 330k resistor (the datasheet specifies at most 300k for 0.8 limit and I wasn't sure at first if it was going to work).

Also take into account that the maximum operating temperature is 85°C and that it heats 83°C/W so that leaves you with 2°C for the ambient temperature to generate 1W, unless actively dissipating the heat. No but it's soldered on the board with an exposed pad and that should spread and dissipate the extra! Which means increased thermal resistance.

Beside the 330k current limit resistor, I've used 560k and 75 for the output resistors. This lowers the leakage current, which matters in battery powered circuits left idle and the output voltage is 5.09V without load, nearing 5 under load.

The other components are two 47uF capacitors, a ferrite and a current limit resistor for the usb port.

Do not use that kind of an inductor for a current-mode regulator unless you know what you're doing.

Testing load response

We can test the output using a variable load that pulls a reasonable amount of power.

I suggest using the Wifi_Scan.ino for ESP8266 or ESP32, since scanning for networks consumes a lot and the delays in-between scans drop some of power drawn.

You can insert a digitalWrite(LED_BUILTIN, LOW); right when the loop begins and a HIGH at the end.

The output voltage differs in only 2mV from the no-load voltage. It swings another mV when it stops scanning, going up to 5.089V.

That voltage is for a 123mA load, which is quite within the operating limits, 0.615W for an operating temperature of 51.05ºC + the ambient 25ºC, ~76ºC (max op temp is 85). And we're well within the 1.2A limit of the inductor.

You can download the board design from my github.


Categories: regulator, design

Leave Comment