• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 312 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.

LMZ12001: Higher ripple voltage measured using Webench design circuit

Jonathan Geronga
Expert 7925 points
Part Number: LMZ12001
Other Parts Discussed in Thread: LM3102

Hi Team,

Looking for your assistance with the situation of our customer's design using the Webench circuit below:


In the datasheet, there is low voltage ripple which is about 8mV
In their measurement using Oscilloscope, the ripple was 200mV
100mV fluctuating and 100mV spike


Please let me know how to resolve the issue.

Thanks,

Jonathan

  • 0
    TI__Expert 7490 points

    Hi,

    Jean will feedback to you soon.

    Thanks,

    Lishuang

  • 0
    TI__Intellectual 1125 points

     

    A few comments.

    The ripple shown is very low frequency, that looks like a no load condition, which is not really what is called "ripple" in data sheet. I assume the 8mV refers to when the operation is in CCM.

    You need to look at the load level applied at Vout.

    You may also need to check what is the Vin voltage for that simulation.

    Regards

    Jean

  • 0 in reply to Jean Picard
    Prodigy 170 points

    Hi Jean,

    I want to be clear for your mention about the ripple and switching frequency mode. In this picture, it is shown that there is no load. Moreover, switching frequency is about 450Hz and the Vp-p ripple about 200mV. Input voltage is 12V and output voltage is 5V.

    The picture which is below belongs to load. The switching frequency is roundly 8kHz and Vp-p riple is about 400mV. I also put the zoomed picture of the signal.  

    We thought that TI components give the result which is giving the result on the datasheet. We designed our PCB board with LMZ12001 component, and the result is awful. I do not know whether webench design is wrong or not. However, we cannot continue with this result. If you say that you cannot stand behind what you refer on the datasheet, we can accept and search other firms that will give better result than you give. We have been waiting a week. Both switching frequency and ripple do not match with the information from datasheet.

  • 0 in reply to Ozancan Ugur
    TI__Intellectual 1125 points

     

    I notice the EN pin is driven from a divider by 10, from the 12V input.

    The EN threshold is ~1.1V, I suspect this is the issue.

    You need to change the divider ratio to give much more margin, for example a divider by 5.

    Jean

  • 0 in reply to Jean Picard
    Prodigy 170 points

    We already change divider ratio as 3 and other dividers. But, it will not change any specifications, still give same outputs. In the datasheet, the EN threshold is written as 1.18V. 

  • 0 in reply to Ozancan Ugur
    TI__Intellectual 1125 points

     

    Could you tell us more about the output capacitor which shows 100uF on your schematic?

    What is the type? is it ceramic?

    Is it X7R or something else?

    What is its voltage rating?

    wanted to make sure about since with ceramic cap the DC bias voltage cases high reduction of capacitance depending on %of voltage rating.

    Jean

  • 0 in reply to Jean Picard
    TI__Intellectual 1125 points

     

    I was also wondering if you could resend your webench file, I tried it and it did not work or point to the correct device.

    Jean

  • 0 in reply to Jean Picard
    TI__Intellectual 1125 points

     

    Here are some comments/explanations.

    First, the high frequency/spikes are noise pickup from probe, need to probe directly across a Vout capacitor for reliable measurement.

    Low frequency ripple (<100mV) in no load condition is expected, although the level of it depends on things like amount of true capacitance and parasitics.

    As explained earlier, ceramic capacitor exhibits significant capacitance reduction (-70% is common) as its DC voltage gets closer to its rating. Quality of capacitor and type (ex: X7R vs Y5V) also adds to it, as well as temperature.

    The IC used in the LMZ12001 is the LM3102 (see below the link). See figure 16 of this data sheet which shows an example of ripple at light load.

    Basically, this type of COT device  defines an ON time based on Ron resistor and Vin as explained in LMZ12001 data sheet. In no load condition, that means that each time it turns on, it will charge Vout cap  and also will build a specific amount of energy in the inductor (10uH). All the energy is then transferred into Cout and this will result in some voltage increase at the output. Since it is very light load, the switching will restart only once Vout has decreased enough.

    So, at the end, this Vout ripple will depend on the amount of true capacitance at the output, since the inductor value is fixed. Of course, it will depend on Vin level as it impacts the energy involved during each cycle.

    To improve the situation, you will need to look at increasing the amount of output capacitance until it becomes satisfactory. That could also mean to increase the voltage rating of these capacitors as it will reduce the capacitance reduction due to bias voltage.

    Jean