• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 730 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.

LMZ14203: Jumpy output voltage

John Kanuchok
Prodigy 20 points
Part Number: LMZ14203
Other Parts Discussed in Thread: TLV62130

Questions for TI on LMZ14203TZ.docx

  • 0
    TI__Guru 50120 points

    Hi John,

    Looking at your layout I have a few suggestions. If possible can you do the following:

    1. Cut a trace on the output and put a sense resistor to measure the output current
    2. Flip the C66 capacitor so that ground is directly connected to the LMZ14203's DGND pad like C67
    3. Can you confirm that your load is constant current loading of 600mA? 
    4. Can you reconfirm the added resistor is 1kOhm? 

    From the waveform you've attached, it looks like the LMZ14203 is slightly in DCM mode. Using Equation 11 from the LMZ14203 datasheet, the current that borders the DCM/CCM boundary for your application is approximately 568mA. In DCM mode, you would expect higher output voltage ripple because of the lowered switching frequency as a result of minimizing switching lose. However, adding a 1k resistor would only result in and additional 1mA to your load which shouldn't have greatly affected your readings. 

    Regards,

    Jimmy 

     

  • 0 in reply to Jimmy Hua
    Prodigy 20 points
    Jimmy,
    Thanks for the quick response. Here we go:
    1. I cut the trace on the +5V output and using a 0.25 ohm shunt (resistance measured with good ohmmeter) found that the load is quite steady at 425mA. It fluctuates a bit from 415mA to 430 mA.
    2. I flipped C66 and so saw no improvement whatsoever. I left the cap in the relocated position for the rest of the tests.
    3. The load current is 425mA as mentioned in #1 above.
    4. I confirmed that the resistor is 1K ohm using two different ohmmeters. That's only 5mA!
    More new information: After adding several test wires to measure various voltages on downstream regulators that are all powered by the LMZ14203 on one pcb) which are located on a separate pc board (1.0V, 1.1V, 1.8V, 2.5V, two 3.3V (all are TLV62130) and after having the unit under test powered up a good part of the day we saw that the 5V output stabilized itself with NO ADDITIONAL LOAD AT ALL! The 1K wasn't needed! I then shut the test unit off for about an hour and found the instability returning once again. Then adding the 1K resistor was found to stabilize the 5v once again. To clarify: the 5V LMZ14203 is located on one pcb and all the other regulators are on two other pcbs. The 1K resistor fixed the jumping voltage whether placed it on the 5V board or on the other board. For clarity the 5V on the first board feeds all the child regulators on another board. The 5V is carried to the other board via a 2 inch ribbon cable using 4 pins for +5V and 4 for 5V Rtn (GND). Any help would be appreciated.
  • 0 in reply to John Kanuchok
    TI__Guru 50120 points

    Hi John,

    Can you provide me with the switching frequency waveform for both before and after the 1k resistor is added? Theoretically adding the resistor shouldn't push the output current more than 5mA which is still below the calculated boundary DCM/CCM current. 

    Also can you add another output capacitor with the same value to see if that helps regulate the output voltage? In parallel, we will get an EVM and populate it with your BOM to check the output voltage.

    Regards,

    Jimmy 

  • 0 in reply to Jimmy Hua
    TI__Guru 50120 points

    Hi John,

    First I want to start by saying that in DCM mode, it is normal to see a bump on the output voltage. This is because in DCM mode the system is trying to maximize conversion efficiency lose as the frequency begins to drop which results in high output voltage ripple. 

    Secondly looking at the your first waveform without the 1kOhm, it looks like the load may be pulsing because there are two light load values every second during DCM. Every 4 seconds the load is increased and the system exits DCM and enters CCM and the output voltage improves. This can be seen when the voltage drops from 5.23Vdc to 5.08Vdc. 

    I wanted to see if this theory was correct so I took a LMZ14203TZ EVM and modified it to match exactly your BOM. I had an electronic load that stepped the load current between 50mA to 150mA at 1Hz, 50% duty cycle. Then I had an external 5ohm resistor in parallel with the electronic load that I manually inserted in and out every 4 seconds to simulate a sudden 1A loading to ensure the device exits DCM mode. It looks like I can duplicate the waveform from your document with the light load pulse and heavy load method. 

    Below is my waveform:

    I am still unclear how a 1kOhm resistor could fix it since like I've posted before, 5mA added onto 425mA will not push the device out of DCM mode.

    Regards,

    Jimmy