• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 63 subscribers
  • Views 680 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.

TPS63002 overcurrent input and undervoltage output

Lionel SABATIER
Prodigy 60 points
Other Parts Discussed in Thread: TPS63000, TPS61070

Dear all,

On one of our design, we use a TPS63002DRC. This one is suplly with 2 AA batteries in series. The ouput is configurated to 5V @ 100 max and 15ma in average.

The problem is for any reason, the TPs, doesn't want to regulate the output. This one is at 1.8V and the current on the batterie is over 450mA. We have replaced the tps by an external power supply. The current is at is normal value.

When we replace the TPS by a new one, everything come back to a normal working. The older one seems to be broken.

Is some one could have some reason for that point?

Thanks by advance for the answers.

 

RDS

 

 

 

  • 0
    TI__Guru**** 150620 points

    Could you post your schematic and board layout?  Also, I recommend testing the EVM under your conditions and see if it works well.

    With 2 AA cells, their impedance may be too much for the switching action of the TPS63000.  I would try adding more input capacitance (try 100uF of ceramic) and retesting.  It would also help to provide a waveform of Vin, L1, L2, and Vout when the IC is drawing lots of current from the batteries.

    Lastly, why are you using the TPS63000 for this boost application?  We have many othe rboost converters that would offer a smaller, cheaper solution.  For example, the TPS61070.

  • 0 in reply to Chris Glaser
    Prodigy 60 points

    Hello Chris,

    Please find in attachment the power schematic of the design. The rest of schematic contain a Pic16F6490 powered at 5V and driving a 3 digits LCD, and 8 amplifier 2 SA571. These last components are powered at 15V. This power is only activate when we need to work with them. This is to save power. The 15V is power on at a minimum of 3s after the 5v and stay on during 3S.

    Schematic :

    8883.DV_10_015_D_SCH_ESSILOR_ULTRASON-P1.pdf

    PCB :

    8838.PCB2.rar

     

    In the 20110218-TPS6300-break-down file there are some waveforms as you request above at different state as described below :

    Channels :

    Channel 1 (yellow) : Vin

    Channel 2 (green) : Before L

    Channel 3 (blue) : After L

    Channel 4 (pink) : Vout

     

    File « Alim-2.8v – Carte PB.jpg » :

    It’s a board with problem

    Test is made with a 2.8V power supply (about 1.6A when the board is started)

    Trace shows the state when the chip is powered.

     

    File « Alim-2.8v – Carte OK – Low comsuption » :

    It’s a functionnal board

    Trace shows the state when the chip consume about 20mA

     

    File « Alim-2.8v – Carte OK – High comsuption » :

    It’s a functionnal board

    Trace shows the state when the chip consume about 130mA

     

    Files « Alim-2.8v – Carte OK – Low to High 1 » and « Alim-2.8v – Carte OK – Low to High 2 » :

    It’s a functionnal board

    Shows the state during switch from low to high comsumption

     

    7077.20110218 TPS6300 break down.rar

     

    Thanks again

     

    Rds

    Lio

  • 0 in reply to Lionel SABATIER
    TI__Guru**** 150620 points

    Thank you for supplying this information.

    Is this a 2 layer board?  Could you post the other layers?

    For a boost converter (which is the mode that the TPS63000 is operating in with your conditions), the output cap is the most important component to place.  It should be as close as possible to the VOUT and PGND pins of the IC to keep the loop area from the output cap, back to the VOUT pin, out the PGND pin, and back to the ground of the output cap as small as possible.  If I look at your output caps right now, that loop is nonexistent, which is why I assume you have another layer.  The best way to place and route the output caps is as is done on the EVM: http://focus.ti.com/lit/ug/slvu156/slvu156.pdf

    What is the issue with the board that has a problem picture?  I thought the board worked when powered from an external supply and did not work when powered from the battery?  Could you send the same waveforms that you did when you run from the battery?  Also, please take the Vin waveform with a low inductance robe right across the input cap, C2.  This will help eliminate some of the noise pick up on the measurement.

    Based on the impedance between the battery (I assume this is placed across TP20 and TP21?) and the TPS63000, I would again recommend more input capacitance to help the battery supply the start-up current required.

  • 0 in reply to Chris Glaser
    Prodigy 60 points

    Thank you for your answer !

     

    You are right,  our board contains 2 layers. But there is no other cap on the other layer ! See the attached file for the bottom.

    3833.Cadstar - Bottom signal2.pdf

     

    You are also right the ground used for Vout capacitors is not close to the chip. Can this be a reason for our issue ?

     

    With batteries and when the chip consume more than 1.5 Amp, the voltage goes down very quickly, so that’s why we have used an external power supply.

     

    On the board with a problem, the chip is non stop hashing even if we don’t consume current (we have tested with the rest of the board disconnected).

    Yes, the battery is connected to TP20 and TP21.

     

    We will test with replacing C2 with a bigger value (100uF) .

     

    Conserning our issue, we also have seen it even with an external power supply.

     

    By the way at startup, the PS pin is tri-stated at powerup (there is no pull-up or pull-down). Can this could be an issue ?

     

    Thanks.

     

    rds

  • 0 in reply to Lionel SABATIER
    TI__Guru**** 150620 points

    It sounds like you may have 2 issues--not enough input C and poor output cap placement.  The poor output cap placement can damage the IC while the insufficient input capacitance causes the IC to not work with the batteries but work better with the lab supply.

    To fix the output cap issue, I would recommend scraping away the soldermask above the IC on the Vout and PGND planes and adding the largest capacitance capacitor that you can at that point.

    A quicker way to a solution is to order the EVM and test it by itself with both the batteries and the lab supply.  Then, add your front end filter network and see if that works both with and without the batteries.  If that works, then the problem is only your layout.  The idea is to debug the circuit piece by piece--only looking at one piece at a time.

    I would recommend a 1 Meg pull down on the PS pin to keep it in a known state.

  • 0 in reply to Chris Glaser
    Prodigy 60 points

    Hello Chris,

    We have changed the input capacitors and replace C1 and C2 from 10uF to 47uF, we also have changed the 15v input capacitor from 22uF to 100uF.

     

    With this settings we have :

    1.01A peak current at start

    1.06A peak at 15v start

     

    The original board had :

    1.23A peak at start

    1.27A peak at 15v start

     

    We hope that will solve the problem. One board is under tests and will be cycling testing during the day.

    See you later

    Rds