This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Triple output PSU using 2xTPS54331 with TPS61200 as "UPS"

Other Parts Discussed in Thread: TPS54331, TPS61200

Hi.


I am building a PSU board for an application with 3 different Vouts.

My input will be either 12vdc or 24vdc in an industrial environment (aka noisy).

I use:

- one TPS54331 to feed ~3.8volts at 600~900mA

- one TPS54331 to feed ~5ish volts to two devices (one router @ 500mA and one RaspberryPi @ 300mA) and charge up two supercaps in series (2x 50F/2.7V = 25F/5.4V).

-and finally one TPS61200 to power Pi from the supercaps for some time if 12/24v power fails.

So far only the 3.8v part works as supposed to.

Here is the schematic and layout of the 5V circuit.

I have the following problem(s) with the rest of my circuit. When I power the circuit the router powers up immediately, after the supercaps charge to (around) 1V the RaspberryPi boots and the caps keep charging until they reach 1.6~1.8V and then they stop. Also then, the Pi starts to reboot due to insufficient power(current?). If I disconnect the Pi for 2~3 seconds and the supercaps charge above 2V and reconnect the Pi they keep charging but the Pi also keeps rebooting and after a random amount of time the router hangs too (probably from bad power source).

If I power only the Pi it works without a problem.

To test the "UPS" function I remove the 12/24v supply and the Pi hangs from insufficient power(current) and starts rebooting. The supercaps discharge slowly but I measure at the TPS61200 output 3.6V~4.8V (the Pi usually hangs at 3.8V from the tests I have done with a variable DC/DC).

I have built with some cheap chinese DC/DC's and the same caps 10 working units.

So to the Questions...:

-Is my setup doable, at all, with these IC's? or I have to search for another solution?

-Is TPS61200 limiting the current output for some reason? (thermal protection due bad PCB layout?)

-Is so important the PCB layout (I haven't really cared for it in the handmade test PCBs) that may be the cause of my problems?

-Does someone have any suggestions to make about my design?

-Why the inductors buzz so much?

G.Ph.

Edited to change the images to the correct ones.

  • Can you post the rest of your schematic? Do you have any waveforms showing the improper operation?
  • Hi John,

    Thank you for your interest.

    Here is the rest of the schematic (the 3.8v step down) and layout. That test PCB works OK (after soldering a diode across PH and GND which I forgot when i designed it)

    Unfortunately I don't have an oscilloscope. I take measurements with 3 or 4 multimeters...

  • ... the sound of silence?
  • holiday weekend in the US...I'll take a look in a bit.
  • Where to start? I suppose first off we should talk about PCB layout. Yes, for dc/dc converters, layout is critical. I am not surprised you are having difficulty. You should look at the layout of the TPS54331 EVM shown here:

    www.ti.com/.../slvu247.pdf

    As you will see there is a lot of attention paid to proper grounding. Also the VIN capacitors must be located close to the IC and tightly coupled to the VIN and GND pins with wide etch fill areas.

    In the schematic I notice that compensation is the same for both. That is usually not the case.

    It will take a very long time to charge the super caps. i would suggest that you disconnect them and the other load circuitry. then get your two TPS54331 circuits fully debugged along with that other TPS61200 converter. then add in the super caps only. see how long it takes to charge them up fully. Is the 5V converter still working properly? You may see frequency fold back or other unusual behavior. Be sure to take waveforms. Once we get that up and running you can worry about other parts of your system.
  • I didn't know you had a public holiday. My respects for your Memorial Day.

    I will read the user guide and report/ask you back any questions.
    What waveforms do you need except Vout/time?

    The 3.8V dc/dc though worked fine.
    I don't have the waveforms of it but the tablet, that is powered from, works with no problems.

    For the caps charging time is not a problem but I am too considering a dedicated dc/dc for them to manage better the power (maybe with a PS mode TPS54331).

    Thank you very much for your time John.
    George Photopoulos