MT2492 buck converter does not output correct voltage

[user749887]

Hello.

I am designing a PCB on which I will use a buck converter to stabilize voltage from 12V to 3.3V that will be used by a low-current microcontroller.

I am a new to PCB design and I've never used SMD buck converters, so I decided to buy the buck converter I intend to use and test how it works on a breadboard.
Since the chip comes in SOT23-6 package, I used SOT-to-DIP converter board.

I intend to use MT2492 buck converter (datasheet). It seems to follow my needs (12V input voltage, 3.3V output voltage). I will use no more than 200 mA of current even though 2A is allowed.

The problem
When I connected it as recommended by the datasheet I found that the output voltage is not what I expect it to be (3.3V). Usually it is higher, but it also can be lower.

If I understand correctly, the voltage at the Vfb (voltage feedback pin) must be equal to the regulator feedback reference voltage. In the case of MT2492 it must be equal to 0.6V. But it's not equal to 0.6V. Sometimes the voltage at the Vfb is 0.4V, sometimes it is 2V.

It is also dependent on the input voltage, while it should not be (or should it?).
I used two different power supplies: 5V and 12V. For the 12V power supply the output voltage also was higher, while all the other components were the same, including the voltage-dividing resistors.

Also, the output voltage seems to be stable, not noisy. However, wrong.

Connection
Here is the exact schematic of the connection of the buck converter on breadboard:

The resistor selection for the output voltage of 3.3V:

The inductor inductance selection:
I selected the inductor ripple current to be 30% of the maximum current (200 mA), so 60 mA. As per datasheet, the switching frequency is 0.6 MHz.

I could not find the inductor with the exact inductance in my local store, but I could find 10uH, 47uH, 100uH, 220uH.
I tried all of them. None of them fixed the problem.

For both input and output capacitors I used 22uF ceramic capacitors.
I could not find 22uF ceramic capacitors in package with legs, I could only find in SMD package. So I soldered copper wires (with isolation) to the terminals of the capacitors and then inserted the wires in the breadboard.

Notes
To ensure that the buck converter is not broken I bought and tried two more, and none of them worked as I expect.

Question
Can I always expect it to stabilize voltage to 0.6V at the Vfb pin?
If so, why the buck converter may not stabilize the voltage?
What should I try?

Thanks in advance!

[edybala]

Hello my friend.. I've selected the same chip for may application, but I have not started its development.

A few considerations:

• Using a 600KHz oscillator on a breadboard might bring some issues on your tests.
• I could not see any load attached to your circuit. As far as I know you might need a minimum current on your circuit to at least trigger the regulator to start (not sure what that is).
• If you maximum current is only 200mA, I would suggest to abandon this strategy and use a regular LDO component (US$ 0.07)
  AZ1117H-3.3TRG1 | Diodes Incorporated | Linear Voltage Regulators (LDO) | JLCPCB

[markbng]

Using such a step down converter on a breadboard is a big "no no". Precise inductance (breadboards have metal strips) and decoupling capacitors are very important. I guess that if you make a PCB of it that all will work as it should. Use also small capacitors parallel on C1 and C3. 10NF will do just fine. These capacitors filter the higher noise frequencies (harmonics of the square wave).
DC DC converters have 1 disadvantage. Like I mentioned...Noise. But I will do the same. Use it also to make 5V out of the 12VDC power supply, but the 3V3 power comes from the 3V3 LDO (connected to this 5V). You made the right choice to not use a LDO for making 3V3 out of the 12V. P = U*I = (12-3.3) * 0.2 = 8.7 * 0.2 = 1.74W. Ok, 200mA is the peak value perhaps, so it won't be disipating 1.74W on average, but I don't like these numbers also. Imagine this schematic in a closed plastic box in the summer. The LDO will heat up (a lot!). And heat can also cause other problems on the PCB. accurate sensor readings for example. The step down converter makes your circuit also more power efficient, so a battery will last longer.
You can also order cheap pre-assembled step down converters that has the same pin out as a regular 78xx regulator. Search on aliexpress for step down regulator. That is also my recommendation for most users. Buy an assembled step down regulator to save power.