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.

TPS61181 hitting overvoltage immediately on startup

Other Parts Discussed in Thread: TPS61181, TPS61180

I'm experiencing a problem using the TPS61181 WLED driver, driving an LCD LED backlight (specced as 6 strings of 18V 18mA).


When the driver is enabled it illuminates the backlight very briefly, then cuts out. EN needs to be toggled before it comes on again, at which point it repeats the same effect.

Here's an oscilloscope capture. CH1 is the "VO" node driving the backlight common anode and CH2 is an "IFBx" node connected to a cathode. Other IFB nodes looked similar.

The output voltage appears to ramp smoothly and the backlight forward voltage is around 18V as expected, but the voltage seems to keep rising until the overvoltage cutout kicks in.


Here's the relevant schematic. Very similar to the sample in the datasheet, apart from omitting the optional "fault" FET:

VBAT is 15.5V. EN & DCTRL are both driven low when VBAT is first powered, and several seconds later are driven to 3.3V simultaneously.

For the momentary period that the screen is on, before overvoltage protection kicks in, the backlight is even and constant and appears to have the correct brightness.

Applying PWM input to DCTRL results in the backlight dimming as expected, and it stays on for proportionally longer before cutting out. Oscilloscope traces show that the ramp of VO is less rapid.

Grounds are connected to a single internal ground plane (this is a four layer board). I can supply a picture of the layout if required.

Do you have any suggestions for further diagnosing or solving this problem?

Thank you,

Angus

  • Short update: I'm still trying to figure this one out.

    I'd bought my previous TPS61181s from a slightly suspect source, so I got some replacements from Mouser just in case the original ones were fake or factory seconds. Same behaviour with the new ICs.

    Have a general suspicion that I've missed something subtle here. Perhaps the "optional" input FET was not really optional, or perhaps powering on VBAT before applying EN is causing a problem.
  • Hi Angus,

    Is it possible you hit the minimum PWM Duty? Did you measure the pulse length at SW?

    Leo.
  • Hi Leo,

    Good suggestion. I don't think so though. :( On paper even the worst case duty cycle should stick above the minimum by a good margin (Vo=18V Vi=15.5 gives me D=14%).

    I also didn't mention this, but I get the same result running with VBAT (Vi) much lower than 15.5 - have run it as low as 9V or so.

    I hadn't actually looked at the SW pulse though, here it is just before cutting out (seems consistent across the entire voltage ramp):

    (CH1=output, CH2=SW)

    Looks closer to max duty cycle than min to me. :(

    Hooking the scope up again I noticed something new, which is that one of the cathode connections (IFB4) stays at 0V and doesn't ramp like the others. I think this is a faulty connection on the flex cable for the display. Regardless, according to the datasheet (page 13) it shouldn't matter for TPS61181 - that string should be removed when overvoltage hits, and from then the remaining strings should work.


    Will fix that string's connection and see if it makes a difference.

  • When it stays on max PWM, it’s possible you reach the ‘sustained over-current condition’. Can you measure the Switch or Inductor current?
  • I fixed the cold solder joint on IFB4 and now everything works! I had a second TPS61181 with the same symptoms and it also had a cold solder joint on a single cathode connection to the panel.

    This seems to contradict the datasheet which says that the TPS61181 should recover from a single open string, not shut down. Perhaps I have too much capacitance on the output, so output voltage didn't drop below the overvoltage threshold fast enough when IFB4 was removed from the feedback loop?

    Thanks for your input Leo, it was due to your prompting that I re-measured the pins on the scope and noticed this problem.

    I'm going to leave this question open to see if anyone has any suggestions about why the TPS61181 displayed the symptoms it did, rather than behaving as specified on the datasheet.

  • Good you have solved the problem, and so it goes many times, also for me, a trigger in another direction can make another fault visible.

    If it isn’t the ‘open string detection’ then I think it could be the ‘overload condition’. To boost the voltage from 15.5V to 39V with the (relative) small inductor needs a high switch current and maybe becomes too high. You could test this to boot with a very low PWM Duty.

  • That's possible I guess, but 39V isn't the designed operating output voltage - it's the TPS61181 overvoltage threshold. Now it's working properly it regulates output around 20V as intended.

  • In your first case you had an open connection then it tries to rise this output pin to the minimal required voltage –in which it doesn’t succeed- and rises VO to max 39V, and this (power-up) boost can be too much for the switch.

  • I think it somehow sounds like you've had TPS61180 mounted instead of TPS61181... it has exactly the described "one-takes-all-down" behaviour...
    BR - Jesper K.