• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 417 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.

TPS62420: Vout accuracy and Def_1/ADJ2 leakage currents

Mike Twieg
Intellectual 960 points
Part Number: TPS62420


My question is basically the same as this one: https://e2e.ti.com/support/legacy_forums/automotive/f/26/t/550678?tisearch=e2e-sitesearch&keymatch=tps62420%2520def_1%2520bias

I'm using the TPS62420 to create 3.3V and 1.2V for a F28377D MCU. On the few assembled boards received so far, I'm seeing larger than expected variation on the output voltages (1.15-1.2V on the 1.2V rail, 3.33-3.39V on the 3.3V rail). Feedback resistors for  1.2V are 200k/200k, for 3.3V they're 402k/86.6k. All resistors are specified 1%. The variation on the 1.2V rail is especially unacceptable. I measure the voltage at ADJ2 to be 0.599V for all boards, so I don't think the reference voltage in the TPS62420 is the issue. This leaves leakage current on ADJ2 as the primary suspect. These measurements are independent of the load current on the outputs (the part seems to be otherwise operating normally in pwm mode).

The leakage for DEF_1 is specified at 0.01-1uA (no specification is given for ADJ2). As Marion pointed out in the above thread, this implies a huge source of Vout error when using suggested resistor values over 100K. Murthy suggests we can just ignore this for some reason. Not sure if this is also applies to ADJ2. But the only other explanation is that my CM is shafting me by using poorer tolerance resistors, which I find unlikely.

If the 0.01-1uA range given in the datasheet isn't correct, what should we assume? Certainly if I were to use 2M instead of 200k, I would see a change in output.

  • 0
    TI__Genius 14876 points

    Hello Mike,

    Murthy in this thread states that because of our ATE limitations/accuracy, max 1uA is defined in the datasheet. But the typical value does still remain 1uA. All other bias current are defined this way as well.

    I recommend you to have a look at this App Note which discuss this very matter, it could be very helpful for you:

    http://www.ti.com/lit/an/slyt469/slyt469.pdf

    Then if efficiency at light loads is not a great concern of your design then you should consider reducing the value of your resistor bridge (R1+R2).

    If you have any other questions, please ask!

    Thank you.

    Best Regards,

    Dorian 

  • 0 in reply to Dorian Brillet de Cande
    Intellectual 960 points

    If 1uA were the typical leakage current, then using a feedback resistor of 200k would give a typical output error of 200mV. Fortunately it's not quite so high in practice, based on observations. But it's still significant, apparently.

    SLYT469 suggests a max total resistance of 1.65Mohm, and that's assuming a max Ifb of just 100nA. If Ifb were 1uA, then the recommended divider resistance should go down by a factor of ten, in which case all the recommended values in the tps62420 datasheet are too high. So why the blatant contradictions?

    I'd at least like an answer as to whether the leakage current for the adj2 pin is the same as the def_1 pin.

    Regards,

    Mike

  • 0 in reply to Mike Twieg
    TI__Genius 14876 points

    Hello Mike,

    Losses in buck converters can come from different sources, High/low side Rds-on MOSFET, DC resistance of the inductor you are using, bad schematic, not choosing LOW-ESR capacitors, a bad layout, etc...

    Fact is here, the fix for your issue is quite simple, if you see that the voltage you observe at ADJ_2 is the one you expect and with the theoretical calculation you made you don't think losses due to leakage is a real problem. You should think of readjusting your regulation point by changing the value of your resistor bridge or maybe also adapt with the selectable output voltage for converter 2 (Table 6). 

    Looking at the numbers at ATE, leakage current for the adj2 pin is equivalent to the leakage definition of def_1 pin.

    Thank you very much!

    Best Regards,

    Dorian

  • 0 in reply to Dorian Brillet de Cande
    TI__Genius 14876 points

    Hello Mike,

    Do you have further questions on this part? Were you able to achieve a regulation point which fits your application specifications better?

    Thank you very much!

    Best Regards,

    Dorian

  • 0 in reply to Dorian Brillet de Cande
    TI__Genius 14876 points

    Hello Mike,

    I haven't heard back from you.

    I will now close this thread. If you want to discuss further on this topic, you can reply below or open a new thread.

    Thank you very much!

    Best Regards,

    Dorian