• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 289 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

TPS552892: Efficiency does not meet specifications(5Vin/ 5vout)

TPS552892: Performance at lower Vin and higher load

Irene Hu
Prodigy 20 points
Part Number: TPS552892

Tool/software:

Hi there,

I'm trying to use the TPS552892 in my application with specifications Vout = 12V, Vin = 8-24V, and Iout = 2A.  I have copied many of the values from the evaluation module TPS552892EVM-111, but I am still running into some issues with lower Vin at a higher output current. 

I am using a programmable DC electronic load, which should be resistive.  With Vin = 8V and currents above 750mA or so, the output turns into a triangular wave.  With higher Vin, even still in boost mode, it is able to handle more load (I haven't characterized this relationship too carefully).  I played a little with output capacitance and it helped some stability issues at even lower load currents, but did not help at 750mA.

I then tried using the programmable load with the evaluation module.  The EVM does better than my board, but still seems to fail its own specs.  With Vin > 6V it should be able to handle 3A.  However, when I set Vin = 8V and Iout = 1.5A, the output produces a triangular wave similar to what I saw with my own board.  I measured the following waveforms on the scope:

Ch1(yellow) = Vout, Ch2(blue) = SW1, Ch3(red) = SW2

Using Vin=8V, Iout=1.2A:

Using Vin=8V, Iout=1.5A:

(Note the timescale is very different in the two images).  Changing Vin and Iout create different versions of these two waveforms.  The patterns make sense in the working conditions, but I'm not sure what is happening when it's failing at the higher loads.  At 1.5A it's still able to hit 12V for a little bit of time, but as Iout increases it stops hitting 12V and turns into a triangular wave.  I thought it might be a pattern that's familiar to somebody who has more experience with switching regulators.

I'm hoping to use this as a clue to troubleshoot my own design, since it seems to be failing in the same way (just at a lower load than the EVM).  But since the EVM should work to its rated spec, I'm not sure if there is something else wrong with my setup.  

Thanks!

  • 0
    TI__Mastermind 27900 points

    Hi Irene,

    Thanks for reaching out.

    Have you checked Vin and EN waveforms? From the second waveform, the device is continuously restarting and shutdown. Please check if the current limit of power supply is triggered, and increase the current limit value of DC supply.

    BRs,

    Bryce

  • 0 in reply to Bryce Xun
    Prodigy 20 points

    Hi Bryce,

    Thanks for the response.  Somewhat embarrassingly, that was indeed the problem with my EVM setup.  However, it's not the problem with my own board which is also failing with a triangular Vout.

    My own board doesn't have test points for SW1 and SW2 (the two sides of the inductor), but I tried to probe them anyways via the capacitors' pads.  This takes some fidgeting and somewhere in the process, something seems to break.  After "breaking", the regulator outputs a lower Vout (and draws some non-insignificant current) even with open load.  I've broken two regulators on two boards this way and I'm not sure how or why.  It's possible I shorted something while probing.

    I'm having trouble getting consistent behavior out of my regulators/boards, and they keep breaking before I can look for patterns.  I'll keep trying when more regulators arrive in the mail.  In the meantime, any chance you recognize these waveform patterns? 

    Ch1 = Vout, Ch2 = SW1, Ch3 = SW2, Ch4 = Vin. 

     

    This is happening at low load this time (Iout = 250 mA) as this incarnation of the regulator/circuit seems unable to handle low loads.

    Thanks!

    Irene

  • 0 in reply to Irene Hu
    TI__Mastermind 27900 points

    Hi Irene,

    Thanks for feedbacks.

    1. Do you mean the EVM can operate as expected now?

    2. Can you share the schematic and layout of your board regarding to this part? 

    3. From the latest first picture, Vin is stable at about 8V, can you also probe EN voltage on a new board? 

    4. About the latest 2nd picture, SW1 waveform is abnormal, I am doubting if the device is already failed. Please replace with a new device and test again.

    BRs,

    Bryce 

  • 0 in reply to Bryce Xun
    Prodigy 20 points

    Hi Bruce,

    1. Yes, the EVM seems to operate as expected now.  Thank you!

    2. Here are the schematic and layout (4 layers).  Power for the TPS552892 (Lamp_Pwr_Unreg) is coming through a high side switch (TPS27S100ARRKR) set with a current limit of 3A which has an additional 10uF on its output (feeding input to FB2, which itself is left over from an older design and we are thinking of removing it).  Main power that goes through the high side switch to TPS552892 has about 100uF located nearby, so I had considered all of this to help with input capacitance, which might not be valid.

    The layout is sort of difficult to see, but we did try to keep everything close and follow the layout guidelines.  I did not separate analog and digital grounds for the regulator, though.

    Re: 3 and 4, I installed a new regulator and observed the same behavior, including a similar waveform on SW1.  I also realized that I was probing the wrong Vin last time since I probed system power; instead, I should probe what comes through the high side switch.  This new Vin looks much noisier, perhaps additional capacitance would help.  But would this cause the strange behavior on SW1?

    Using Vin=8V, target load 500mA, I observed the following waveforms.  Ch1=Vout, Ch2=SW1, Ch3=SW2, Ch4=Vin (pin 2 of J11 in the schematic).  I also probed EN separately but it is voltage divided off Vin and looked stable in comparison.

    The parts are manually soldered and it is possible there are some assembly errors, but at this point I have observed the behavior twice on two separate boards.  Please let me know if you have any ideas... thanks!

    Irene

  • 0 in reply to Irene Hu
    TI__Mastermind 27900 points

    Hi Irene,

    From the PCB screenshot, the layout is very poor especially for the power pins (Vin, Vout, GND). The crossed connection of copper plane would introduce much parasitic parameters into switching loop. And the waveforms below also shows Vin has large distortion on that. With such poor layout, the device is hard to operate as normally. 

    You can try to add more ceramic capacitors very close to Vin and GND pins. But it is strongly recommended to revise the layout as the layout guideline suggested.

    BRs,

    Bryce