• State
  • Locked Locked
  • Replies 10 replies
  • Subscribers 63 subscribers
  • Views 459 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.

TPS65130: TPS65130 Vpos ripple issue in Power-Save mode

sgomez
Prodigy 120 points
Part Number: TPS65130

Hello:

The problem has been detected with the last production lot, and it was never seen before. 

The schematic is the one shown below:

And the gerber Top and Bottom layer:

GND is always in an inner layer.

The power consumption is nearly the same for the positive and the negative rail, and is about 20 mA.

In all the circumstances, the waveform in the negative rail is something similar as expected:

… with a AC level of about 50mV, very similar to the one in the figure 47 of the datasheet:

But in nearly the whole of the units for this batch of production, the waveform for the positive rail is not as expected, producing a very high amount of noise:

… with an AC level of about 200mV, far away from the expected in the figure 46 of the datasheet:

A few units are good, with a waveform very similar to the one of the datasheet:

If we replace the IC in a unit with the problem, with one from an older production date, the problem is solved. If we replace the IC with one of the same reel, the problem persists.

We have produced a preproduction lot of ten units, with the same hardware, but older chips, and none have this problem.

I think the problem is the same as the one that can be seen in:

https://e2e.ti.com/support/power-management/f/196/p/671424/2473050?tisearch=e2e-sitesearch&keymatch=tps65130#2473050

Please, could you help me? Con somebody see anything wrong? Could be this batch of parts defective?

Tanks in advance.

  • 0
    TI__Mastermind 18550 points

    Hello,

    Thank you for your post but I am  not able to see schematics, layout or any of the other pictures in you post. Can you please re-attach them?

    Kind Regards,

    Liaqat

     

  • 0 in reply to Liaqat Khan
    Prodigy 120 points

    Yes, I had a problem when upload the images, I don't know why...

    There is a pdf attached with all the images...

    Thanks a lot.TPS65130 Vpos ripple issue.pdf

    Regards

  • 0 in reply to sgomez
    TI__Mastermind 18550 points

    Hello Santiago,

    Thank you for providing the schematics, layout, scope plots and other information. I am a bit tied up with other issues right now but will review and get back to you by no later than Friday this week, if not earlier.

    Kind Regards,

    Liaqat

  • 0 in reply to Liaqat Khan
    Prodigy 120 points

    Hello again Liaqat:

    I have more information about the problem.

    The bad chips have the marking shown below:

    TPS6

    5130

    TI 99W

    Z18K

    I have tested new chips bought at Digikey that are good. Their marking:

    TPS6

    5130

    TI 83W

    Z166

    Unfortunately, I am unable to find any document to decode manufacturing data (year, week and fab...).

    Regards

  • 0 in reply to sgomez
    TI__Mastermind 18550 points

    Hello Santiago,

    Thank you for your patience. I reviewed your schematics and layout in detail and did not notice anything glaring that should cause higher ripple. I also did some testing in the lab with few EVMs to see if I can observe the high ripple that you are seeing but was not successful. Unfortunately, ripple is not a specification of the device and is therefore not measured during production testing. So it is possible that it can vary significantly from device to device in the power save mode where switching occurs in bursts as opposed to continuous switching in PWM mode (PSP/PSN pins tied to GND). I am not sure if it is feasible in your application but, if output voltage ripple is an important consideration in your target application, I would suggest to tie PSP/PSN pins to GND so that device always operates in PWM mode and provide a consistently lower level of output voltage ripple. I can not think of any other suggestions at this time to reduce the ripple.

    Kind Regards,

    Liaqat

  • 0 in reply to Liaqat Khan
    Prodigy 120 points

    Hello Liaqat:

    Thank you for your answer.

    I can understand that ripple of the output of DC/DC converter is not a specified value, but it's external components related: inductor, load and capacitor value can affect the output ripple. When the DC/DC isn't working in Power-Save mode, the output ripple is good for us.

    But when the converter enters in Power-Save mode, the converter stops commuting until the output voltage drops below a unknown value (Datasheet 7.4.1 "the converter only operates when the output voltage falls below an device internally set threshold voltage."). It's true, I can't find this value in the datasheet, but I think it's a very important value for a DC/DC converter. Then, I can assume that figure 46 of the datasheet can show a reference about the behaviour in Power save mode. In fact, the behaviour for the positive rail is always good. But I can't understand how this value could be from a few millivolts to 200 mV, just for the negative rail (not for the positive rail), production to production... I can't understand that this "unspecified" value for this threshold could have this variation: if that is true, Power-Save mode for this converter is not usable... 

    Unfortunately, as can be seen in the images, I can't tie PSP/PSN to GND to disable power-save mode in production, unless these pins would have internal pull downs. Yes, I know: there is no pull-downs.

    Then, the only way I can do is to assume that this chips are defective, and replace them for a new ones, but, please, could you tell me how I can know what chips are the correct ones?

    There is a document that shows the meaning of the top marking of the chips, in order to know manufacturing date and origin?

    Thank you for your help, and best regards.

  • 0 in reply to sgomez
    TI__Mastermind 18550 points

    Hello Santiago,

    As you requested, I will try to find the date code of this device and provide this information if am able to find it.

    Kind Regards,

    Liaqat

  • 0 in reply to Liaqat Khan
    TI__Mastermind 18550 points

    Hello Santiago,

    Here is the manufacturing date information for the two device markings you provided earlier.

    ======

    TPS6

    5130

    TI 99W

    Z18K

    September 2019

    ========

    TPS6

    5130

    TI 83W

    Z166

    March 2018

    ======

    Kind Regards,

    Liaqat

  • 0 in reply to Liaqat Khan
    Prodigy 120 points

    Hello again Liaqat:

    I tried more units with the Top marking shown below:

    TPS6

    5130

    TI 07I

    C5Z2

    I think these ones are the most recent chips I tried. The result was as bad as the first ones.

    Then, I have done some experimentation and tried to change the 10 pf small capacitor in the feedback network (in my circuit, C193). Y tried up to 1nF. The result is not as good as the one for the right chips, but is better.

    But, I don't know other consequences when changing this capacitor with a value outside the recommended in the datasheet.

    Please, could you give me some advice about this capacitor? Or about the whole feedback network? Please, keep in mind that this circuit provides power to an analog circuit, then there is no big transients.

    Thank you so much and regards.

  • 0 in reply to sgomez
    TI__Mastermind 18550 points

    Hello Santiago,

    The feed forward capacitors are there to speed up the control loop in order to improve transient response. It is not recommended to deviate too much from the datasheet recommended calculations. Since increasing this capacitor value seems to improve output ripple in your application, you can probably achieve similar effect by keeping the capacitor value at 10pF but replacing R160 with a zero ohm resistor.

    Kind Regards,

    Liaqat