Post Go back to editing

ADP2301 outputs 2.5V instead of 5V

Hi 

I am trying to make a DCDC converter using the ADP2301. The design is more or less taken from the design tool ADP230x_BuckDesigner_Release.xlsm. 

Down below is a picture of my circuit: 

I was hoping to get 5V as output, but instead i am seeing 2-2.5V with a lot of ripple. (see picture below) 

Here is some more scope pictures for you to understand what is going on. 

At BST pin: 

Enable pin:

Feedback pin: 

And here is the PCB layout. Underneath is a solid GND plane. 

_______________________________________________________________

Can anybody help me out here? I am a bit confused about the feedback pin. From the datasheet the regulation voltage is 0.8V which it is designed to be at 5V. Why am i seeing almost 2V when my output is 2.5V? 

I have no load at the output, does that have an impact? 

Best,

Mathias

  • Hello, Mathias.

    Good day. Can you please provide the input/output specifications (Vin range, Vout, Iout)?

  • Hello again, Mathias.

    Thanks for providing the i/o specs. Upon checking your schematic and layout, it looks like there is no problem. Can you please double-check the feedback resistors you installed? The FB pin voltage should be 0.8V at any condition, even at no load.

    Also, please probe the SW pin and see if there is switching which should be similar to what is shown on figure 26 of the datasheet. The amplitude of the switching waveform should be equal to your input voltage, while the frequency should be 1.4MHz (since you are using ADP2301) and the duty cycle should depend on your input voltage (duty=Vout/Vin).

  • Vin range: 12-16V

    Vout: 5V

    Iout: 1A

    Best,

    Mathias

  • Hi Migs, 

    Thank you for your time! Much appreciated. 

    This is what I see at pin 6 (SW), which is far from what I expect. 

    Here is the same pin AC coupled: 

    which shows that it is running at 630kHz and not 1.4MHz. 

    I will try and resolder the Feedback resistors. 

    Thanks! 

    Mathias

  • Okay, so I reseated those feedback resistors, which give me some better results. 

    This is the output, which looks perfect. 

    However it only stays like this for approx. 2 seconds, then it begins to oscillate, as seen here: 

    Is this because of poor PCB layout? It switches back and forth between a clean 5V output and that nasty oscillation. 

    And BTW, i only have a 10µF output capacitor not 27µF as it is on the schematic, but I wouldn't think that is the problem..

    Best,

    Mathias

    ______________________________________________________________________________________________________

    Added: 

    I decided to resolder all of it. And now I find that under light load (<250mA) i get a nice and stable 5V output. 

    However when I increase the load to over approx. 250 mA the voltage at the output decreases. 

    In this case I am using a 100ohm resistor as load, which gives a stable output voltage at 1V.

    This is the SW pin: 

    Why can't the regulator source enough current? 

  • Hello again, Mathias.

    Thanks again for providing some details. At what input voltage are you doing tests? Please be noted that at Vin=12V, only 1.1V is present at the EN pin, which means that the regulator is disabled. To enable the regulator, Vin must be within 13V to 16V; you may do the tests within the said Vin range.

    Also, please check the GND and power traces. Try to reinforce the traces and then see if there is improvement. Encircled on your schematic are the power traces.

  • Hello again, Mathias.

    Thanks for providing some info. Can you try installing a new regulator, if you still have any? Where did you purchase the units?

  • Hi Migs, 

    I am feeding in 13,8V which gives 1.22V at the EN pin, so that would not be the problem.

    The feedback pin is at 174mV which of course is due to the low output voltage.

    I think that the power traces and GND is "big" enough, compared to your reference layout. However I will try and reinforce them, if it could help.

    Best,

    Mathias 

  • Yes, I do have some spare ones. I can try and change it. 

    Mouser :-) 

    ________________________________________________

    Added: 

    The exact same problem I am seeing is this: ADP2302 do not want to be loaded 

    But his solution by turning on and off the supply rather than putting the load in and out of the circuit doesn't have any effect. 

    I guess it could be my lacking soldering skills, causing a short somewhere. 

    Measuring the resistance between the input GND and the output GND is exactly 0.0ohm as expected. 

    ________________________________________________________

    Added 2: 

    I soldered a new DCDC converter (see picture below). This has the exact same behaviour as before. At no load or light load (100 ohm) works just perfect. If I then insert a 10 ohm resistor it only outputs just over 1V and draws 100mA (of course)

    Also the frequency at SW is divided, so when outputting 1V the frequency is about 300kHz or 1/5 of 1.4MHz. 

    When trying to power a Pi the voltage is 2.5V and the frequency at SW is 700kHz. 

    Hmm

    _____________________________________________________________________________________

    Added 3:

    After fiddling around, I got it to output a stable 4V into a 10ohm load (400 mA), so i am heading towards a solution. 

    The input voltage in at this test is 14V.

    The SW is now looking good (see below). The yellow is SW and the blue is the output, which is 4V as you can see. 

    The duty cycle seems reasonable at about 40%, but why is there only 4V at the output.. 

    The solution is near :-) 

  • Hello again, Mathias.

    Thanks for your updates. For now can you please try using a 2A-rated diode? This one may do.