• State
  • Locked Locked
  • Replies 9 replies
  • Subscribers 65 subscribers
  • Views 421 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.

LP8725 resets?

Bruce McKenney
Intellectual 280 points
Other Parts Discussed in Thread: LP8725

We built a (somewhat) simple proto board to test the LP8725. It contains an MCU (MSP430), a flash chip, an RF chip, and a few others.

Observed behavior is that the LP8725 appears to reset frequently.

Details are:

  1. Device is LP8725TLE (baseline, not one of the A-D variants). All registers are at default.
  2. CONFIG=DEFSEL=DVS=0 (GND) [sub-PMU]
  3. EN=VIN1=VINB1=VINB2=Vin (roughly 3.5V) [Always enabled when powered.]
  4. BUCK2 (FB2) feeds VCC to most of the board. BUCK1 (FB1) is fed back into VIN3 to supply LDO4. VIN2 is not connected.
  5. SDA, SCL, and RESET_N are pulled up to VCC (FB2) using recommended values. RESET_N comes out to a test point but is not otherwise used.
  6. B2_EN is not connected (it reads as constant 0V).
  7. The MCU code has been shaved down to only: de-select the RF and NAND chips, turn on an LED, and then spin (LPM0). Startup shouldn't take longer than a few 100s of usec.
  8. Vin (primary power) is supplied by a fairly strong battery, measured at 3.45V. There is maybe 100mV of ripple.
  9. My DMM reports 4mA being drawn. This is higher than I would expect, though this might be averaging of draw over the frequent chip startups. Still, it is well within the capacity of the battery, and I see no Vin "events" correlated with the resets.

Symptom: Watching RESET_N, I see it repeatedly cycle: de-assert (high) for anywhere from 10-200ms, then re-assert (low) for 35-70ms. While RESET_N is high, BUCK1 (FB1) is at 1.0V and BUCK2 (FB2) at 1.8V (as expected), and they drop to 0V in synch with RESET_N (as expected).

At 4mA draw, thermal shutdown seems improbable. Are we violating some rule somewhere? Are there any other things we could be doing to this chip to trigger this behavior?

  • 0
    TI__Expert 3340 points

    Hi , I don't see any violation with your setup now, but one thing you could try is to pull up SDA, SCL and RESET_N to VIN1 instead buck2. how many boards did you test?

  • 0 in reply to Sheng Jin
    Intellectual 280 points

    > how many boards did you test?

    Just one so far; I'll try some others.

    Are there easy/obvious ways to fry this device?

  • 0 in reply to Sheng Jin
    Intellectual 280 points

    I tried another board, and now see a different symptom:

    1. Vin (still) at 3.64V, but with a -0.5V downspike every 400us
    2. BUCK1 pulses to 0.9V, then drops off, every 400us
    3. BUCK2 [Vcc to the rest of the board] pulses to 0.5V every 400us
    4. LDO1 pulses to 3V, then down to 1V, every 400us
    5. LDO4 = LDO5 = 0V (BUCK1 is fed back into VIN3)
    6. RESET_N = B2_EN = 0V [RESET_N apparently never rises from 0]
    7. DMM shows a very steady 7.4mA draw at Vin

    Disconnecting the load (Vcc to the rest of the board) doesn't much change the symptoms, except that every 10 seconds or so BUCK2 rises to 1.8V for 1-2ms, then back to the 400us cycle.

    There does seem to be a theme here. Is 400us a significant time period?

  • 0 in reply to Bruce McKenney
    Intellectual 280 points

    Just to make sure, I tried it with a benchtop power supply (rather than a battery), and the symptoms are identical.

  • 0 in reply to Bruce McKenney
    Prodigy 80 points

    reset n will jump to 1.8v , (pull up to VCC), (buck2), when i run the input at 2.3v, above that, or below that it returns to zero

    buck1 is a steady 1v, , at 2.3v in , above 2.3v in it oscillates, at 2.9v it has pulses at 468 hz rate, but only 1V peaks

    above 2.9v in the buck1 drops off in amplitude, to .5V peak at 3.4khz

    returning to the 1V peak at 400hz when input is 4.4V

    buck2 (vcc)(reset_n) will only jump to 1.8v when input is at 2.3v

    i can try to take the 3 pull up resistors to VIN, instead of VCC(buck2)

  • 0 in reply to randy leet
    Prodigy 80 points

    with the 3 pull up resistors tied to VIN

    at 2.2v-2.3v in buck1 jumps to a steady 1V, reset_n jumps to VIN

    at 2.6V in, buck1 starts to oscillate at 2.7khz, and reset_n drops to 0V

    buck1 will continue to oscillate with 1V peaks, until VIN goes up to 3.1V

    then it oscillates at .5V peaks at 3.4khz

    at 4V in, BUCK1 jumps to a steady 1V, and RESET_N jumps to VIN(4V)

    above 4V, buck1 stays at 1V and reset_N follows VIN

    buck2 is a steady 1.8v when vin=2.3V, it oscillates avove 2.3v, and goes to zero at 3.3vin

    at vin =4.1V, buck1 goes to steady 1V, and buck2 goes to 1.8V -steady

    LDO4 operates in similar fashion, about 1V steady when VIN = 2.3 and then again at 4V and above

    in between these 2 voltages, there is oscillation , or 0V

    LDO5 looks the same as LDO4

    LDO1 jumps to ~=VIN at 2.3V in

    Ldo1 follows the oscillations of buck1, from 2.3v in to 4v in

    at VIN = 4.2V, LDO1 jumps to steady 2.4V

    so the problem is running this between 2.3v and 4v

  • 0 in reply to randy leet
    Prodigy 80 points

    adding capacitance to buck1 just changes the oscillation a little   ???

    it still has steady voltage only at 2.3v, no longer steady ay 4v

    inductor i used for this circuit (1uH) has a 1.5A saturation current, this should be ok for this circuit ???

  • 0 in reply to Sheng Jin
    Prodigy 80 points

    see me notes in the forum, related to this problem

  • 0 in reply to randy leet
    Prodigy 80 points

    turns out that the input capacitor was too far away from the chip

    i added a 4.7uf cap close to the VIN to the chip, and it now works ok