• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 98 subscribers
  • Views 527 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.

TMS320C5535: What would cause "drop outs" on the 1.8 volt supply of a TMS320C5535 eZDSP card?

Tom Makovicka
Intellectual 400 points
Part Number: TMS320C5535
Other Parts Discussed in Thread: TLV320AIC3204, TPS62260

I'm seeing random "drop outs" of the 1.8 volt supply every 100 to 300 ms that last for a duration of roughly 20 to 30 ms where the voltage drops to less than .1 volts.  This is on one of TI's TMS320C5535 eZDSP cards.  The 3.3 volt supply works fine which is derived from the same 5 volt supply from the USB connector.  Also, the software running on the DSP works fine on other eZDSP cards.  And, on this card, the DSP is able to boot up from the micro SD card, re-encrypt the "requesting to be bound" code on the microSD card, and write it back to the card, all before these 1.8 volt "pulses" start to occur.  Once that is done, all four LEDs in the middle of the board stay on and the green LED on the XF pin "flickers".

This same software worked fine on this board for at least for a couple weeks of operation, and also works fine on other 5535 eZDSP boards.  The 1.8 volt supply on the eZDSP card connects to the LDOI pins ( B14, C14, and B10 ) of the DSP and also the DVDD and AVDD pins on the TLV320AIC3204 codec so it potentially could be either part that "went bad".

Any information or help in debugging would greatly be appreciated!

Tom

  • 0
    TI__Guru**** 161600 points

    Hi Tom,

    As far as I understand, these drops are happening on only one board, right? Did this board worked fine for some time or was this behaviour present from the beginning? In both cases you're facing some hw issue either with the pcb design (mistake in the manufacturing of the PCB), or some of the power parts.

    I'm seeing random "drop outs" of the 1.8 volt supply every 100 to 300 ms that last for a duration of roughly 20 to 30 ms where the voltage drops to less than .1 volts.

    Anyway, check if these drops in the range of .1 volts are be within the recommended operating conditions described in the device datasheet. If yes, then your device should function well.

    As I see from the schematics, the 1.8V supply VCC_1V8 is produced from the U1 (TPS62260DDC) dc-dc step-down converter. Is possible to try and change this part and see if the behaviour of the board is the same?

    Best Regards,
    Yordan

  • 0 in reply to Yordan Kovachev
    TI__Guru**** 161600 points

    Also you could try and replace the capacitors on the VCC_1V8 supply, which are part of the dc-dc converter schematic, these are: C21, C123 & C124: 

    The other thing that comes to mind if changing the dc-dc & capacitors don't work is a mistake in manufacturing the power rails of the pcb.

    Best Regards,
    Yordan

  • 0 in reply to Yordan Kovachev
    Intellectual 400 points

    I probed around with my scope a bit more and realized the "switching waveform" out of the regulaor ( U1, pin 5 ) does "drop out" to zero volts and stops switching.  I don't have another TPS62260, but did notice that due to unpopulated "jumper resistors", the 1.8 volt supply is only connected to the LDIO pins of the DSP on the eZDSP board.  And from the 5535 data sheet is sounds like these inputs can accept anything from 1.8 to 3.6 volts.  So, for this board, could I simple remove the R118 resistor ( as shown above ) to disconnect the 1.8 volt supply and jumper over from the 3.3 volt supply?  Or is there some other reason that might be a problem?

  • 0 in reply to Tom Makovicka
    Intellectual 400 points
    I removed R118, disconnecting the 1.8 volt supply from the DSP and "jumper wired" the 3.3 volt supply to the "DSP side of the R118 pads. This fixed the DSP from not working, but the U1 voltage regulator and it's inductor got hot to the touch. Using a multimeter, the resistance from pin 5 of U1 to ground measured around 40 ohms, while a "good board" measured over 4k ohms. Not knowing if the capacitors C21 or C123 were shorted, I removed them but the resistance stayed at 40 ohms and U1 and it's inductor still got hot when power was applied. So I finally removed the U1 regulator and the impedance of the pin 5 pad went to above 4k ohms and the inductor didn't get hot when the board is powered. So apparently, the U1 regulator had some kind of short internally to pin 5.

    Thanks for you help!

    Tom