• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 63 subscribers
  • Views 266 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LM25145: LM25145 can t achieve correct regulation

Emanuele Peruzzi
Intellectual 510 points
Part Number: LM25145
Other Parts Discussed in Thread: CSD18534Q5A, CSD18514Q5A,

Hello, i have designed a schematich through webench tool for LM25145 regulator, my specs are:

Vin = 24V

Vout = 12V (about 6A current)

actually i have:

RT resistor = 11.2 Kohm

EN = tied to 18V.

Cin = 10uf ceramic

Cout = 4x22uF ceramic

Cbst = 100nf

Rilim = 470ohm.

Rfb1 = 1Kohm.

Rfb2 = 14Kohm.

Inductor = 4.7uH  6.25mohm 15.5A

Q1 = CSD18534Q5A

Q2 = CSD18514Q5A

The problem is that i can t get proper regulation, output voltage is swinging around 12V for some seconds, then it goes to 20-22V. During this initial period, there is a loud "frying" noise. Putting probe at SW pin, i see a wave of 1.47 MHz frequency and amplitude from 5V to 28-29V. Duty cycle is ~ 33%. Voltage at Vcc is alaways 4.8-5V not very stable.

I can t understand why i have such high switching frequency, and why i can t get 12V output. What should i check?

  • 0
    TI__Guru 57555 points

    Hello

    First check you values against the calculator tool found on the product page for this device and check to be sure that

    all components are correctly soldered in place.

    Also, please send a view of your PCB layout.

    I will be assigning this post to someone in our controller group.

    Thanks

  • 0
    TI__Mastermind 22299 points

    Hi Emmanuel,

    Can you send over your filled out LM25145DESIGN-CALC as well as the layout?

    Also the SW node waveform would be very useful.

    I think your schematic might be ok as it starts up to 12V, but there is definitely some instability I think is layout related.

    -Orlando

  • 0 in reply to Orlando Murray
    Intellectual 510 points

    Hi,

    hints really appreciated, i try posting some things, i can t post full gerber, cause it s a part of a bigger project that i can t share, but i will post switching part, which is almost separated from all other stuff.

    This board has 4 layer, top, bottom, layer 2 is ground and layer 3 is power supply.

    I drew schematic with webench tool.

    I desoldered from board all component related to other circuits, so in fact this regulator is completely separated from the remainig part of the board.

    Testing this schematic, i cannot get proper output regulation, so tried filling your design calculator and i noticed that values suggested for the compensation circuits are different from the ones suggested by webench. So i tried values suggested by calculator.

    R5 = Rilim = 470 ohm.

    C15 and C17 are short circuit. C16 = 47nF

    C23 = 100pF

    R12 and R17 are short circuit and R13 is 6.4Kohm.

    I modified also R9 rt resistor to 11Kohm, so switching frequency should be about 950 KHz,

    instead i get this waveform at SW pin, wave going from ~ 6V - 30V and frequency 1.493 MHz, and a loud frying noise...

    if you have any hint, will be very appreciated.

    LM25145_12V_reg.zip

  • 0 in reply to Emanuele Peruzzi
    TI__Mastermind 22299 points

    Emanuele,

    I find it strange your SW node falls to 5V instead of 0V.

    Can you check the source of the low-side FET is actually 0V? Also can you probe the Gate-source voltage of the lowside FET?

    Right now the SW node suggests the Lowside FET is not turning ON completely.

    Let me know,

    -Orlando

  • 0 in reply to Orlando Murray
    Intellectual 510 points

    Hi,

    thanks for suggestion. I recorded some waveforms.

    In these images, yellow trace is SW node, and blue traces are:

    Q1g = high side gate.

    Q2g = low side gate.

    Q1d and Q2s are drain of high side and souce of low side, that seems to be correctly tied to 24V (Vin) and 0V (gnd).

    Voltage on Q2 gate (low side fet) goes from 0 to ~ 6-7V, while voltage on Q1 gate (high side) goes from 0 to ~ 30V. It is normal all this difference??

    Kind Regards

    Emanuele

    gates_waveform.zip

  • 0 in reply to Emanuele Peruzzi
    TI__Mastermind 22299 points

    Emanuele,

    For the high-side MOSFET to turn ON the VGS must be greater than the threshold voltage, since the source voltage goes to VIN, the gate voltage must go to VIN+7.5V so the highside turn-ON is correctly ~30V(achieved with the bootstrap BST circuit).

    The high-side turn-OFF signal looks weird. The high-side pull-down is SW node, so HO should fall to SW voltage to turn off high-side MOSFET.

    We expect SW to fall to 0V and HO fall to 0V with it, but here it looks like SW only falls to 5V and HO falls to 0V, implying negative VGS. 

     

    Your low-side MOSFET signals look good, the source voltage is 0V and the LO gate signal looks correct, so the SW node should be pulled to 0V but it is not.

     

    There is something wrong with the gate driving, can you try replacing your MOSFETs and IC see if the SW node waveform improves? Please try at no-load with SYNC-IN tied to GND.

    -Orlando

  • 0 in reply to Orlando Murray
    Intellectual 510 points

    Ok,

    now i have changed Q1, and sw waveform looks good. also can achieve stable regulation at 12V and also with load. I consider this problem solved.

    Thanks you for hint, and suggestion i really appreciated.

    Kind Regards