• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 63 subscribers
  • Views 482 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.

TPS61161 failures

David Fansler
Prodigy 70 points
Other Parts Discussed in Thread: TPS61161

I have an ongoing love hate relationship with the TPS61161.  It is the IDEAL LED controller for driving a string of LEDs from 12v, low part count, great current carrying ability - but a fink when it comes to usability. 

I will be the first to admit that part of the problem is that I am hand soldering this chip – but still . . . .

I am designing strips of LED lighting for sailboats to replace incandescent lighting that draws several times more power from the battery.  The board layout around the TPS61161 is from the TPS61161 data sheet.  I had been milling the PC boards, but recently purchased the “Fab In A Box” that allows you to etch a PCB with traces down to 5mil. 

So I spent yesterday laying out a board which I etched late yesterday.  After supper I loaded up the board and gave it power from a power supply (13v) and it worked.  Took it down to show my wife (the Admiral, I am just the captain of our sailboat) the results of the etched PCB running it off of a 7v lithium battery.  Took it out to the shop and drilled some holes in it the board.  I came back in the house with the unit running on the battery.  Hooked it up to the power supply still at 13v and the LEDs flashed and went out.  Now every time I apply power, the LEDs give a brief flash and die.  Looking at the feed to the LEDs from the diode, when the power is applied the voltage spikes to 23 volts and then, what appears to be exponentially decreases to 13v.

This is not a good thing.  And it has happened to me more than just a few times.  On occasions, replacing the TPS61161 has caused operation to resume.  But what is killing the chip? 

I built 5 of these strips for the galley and installed them this spring.  2 died immediately and 3 are still running.  By the way, I am using 80mA LEDs with a .82 ohm sense resistor. 

Any speculation or desire to look at the board would be appreciated.

Kind regards,
David

  • 0
    TI__Guru**** 166626 points

    Sorry you are having problems. The 2x2 QFN package is delicate.  Your board should have thermal relief off of the corner pins, i.e. don't take wide traces to large planes off the corner pins.  Please see the EVM user's guide for pics of the board layout http://www.ti.com/litv/pdf/slvu221c.   There is also a link in the user's guide to an app note that gives more detailed explanation of the problem and guidance on what to do to avoid the problem.  In short, we soldering, you need to either use hot air or an oven to more or less solder all the pins and power pad at the same time OR place the entire board on a hot plate so that both sides of the IC are closer to same temperature when soldering.   Do not touch up the pins after soldering.  Otherwise, the thermal stress on one side of the package causes it to expand, creating fractures in the package and wirebonds.  The wirebond fractures cause intermittent open connections with various pins which results in various failure modes.

  • 0 in reply to Jeff F
    Prodigy 70 points

    Thanks Jeff - you seemed to have confirmed what I thought was probably the problem - soldering.  My layout is almost identical to the link you gave - if anything, my area around the inductor is smaller.

    I do not have or have access to a reflow soldering iron or oven, so I tried the hot plate trick.  I was able to remove the TPS61161, add some flux to everything and place the TPS61161 back on the board and heated it back up.  It was neat seeing all the joints go active about the same time.  I let the unit cool, and applied power - it worked - for a little while.  After it went off, I tried flexing the board (.0625) and the LEDs would come back on, for a while.  I reflowed the IC a second time and still have the same intermitt results.  Not sure if I want to etch a new board or try a different TPS61161 on the exsisting board.  I changed the inductor with no luck.

    Thanks,
    David