• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 31 subscribers
  • Views 413 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.

Original question:

TAS5424A: Changes from previous version

TAS5424B-Q1: I2C Crash on B version at high power output

Steve Scarlata
Prodigy 170 points

Part Number: TAS5424B-Q1

Hi,

We are migrating our amplifier design to the TAS5424B from the -A.

The A has been working reliably for over 13 years in thousands of field amplifiers.

On a sample unit with the B, when I play a sine wave on all 4 channels at 1/2 power or greater,  then eventually the chip will Hang.  It continues to play,  but the I2C buss SDA line gets held low by the 5424 and communication is inhibited.

There is no FAULT pin pulled. It isn't overheating, not hitting any warning zones,  all the registers read normally until the crash. No load dumps, no power supply glitches.VDD is 21V.  I reproduced the same fault on 2 other units with the B chips.  I can't reproduce the fault on the TI EVM with the B chip.

I'm using the TI USB2IIC board to communicate with the chips,  taking my own micro out of the loop to eliminate any possible influence.

Any idea what could cause this condition?

  • 0
    TI__Mastermind 20241 points

    hi Steve

    the status is so tricky.

    maybe you can check what is the schematic difference between your board and our EVM.

    also, the waveform at i2c line.

    tks

    jesse

  • 0 in reply to Jesse Ji
    Prodigy 170 points

    Jesse,

    After closer inspection of the TI document  "TAS54x4A to B Comparison r2.pdf" that was provided a while back,  I see on the last page this note:

    The E2 device has the same specifications and datasheet as the TAS54x4C-Q1 device. The E2 device has fixes the I2C NACK issue that was found on the TAS54x4C-Q1 device.
    1.I2C NACK
    At sustained high power the I2C could possibly NACK or overwrite (corrupt) the data in the control registers. This has been observed when the sustained power is above 55W per channel with all 4 channels in play mode. It has also been observed at 70W into 2 ohms, one channel driven.

    This sound like it's my problem.

    The Chips we are using have these markings.   TAS5424BQ1     16ZHV14G4

    The Chip on the EVM and another set that don't show a problem are marked  TAS5424BQ1   2AZDHD4G4.

    Could you please confirm that this defect found it's way into the B version,  and clarify how we can identify and purchase known good ones?

    Thanks,

    Steve

  • 0 in reply to Steve Scarlata
    TI__Mastermind 20241 points

    hi Steve

    about his document "TAS54x4A to B Comparison r2.pdf", our team do not have that at hand.

    does it have any contact information, it might be better to contact with that information.

    our team now do not have very detail information about that.

    tks

    jesse

  • 0 in reply to Jesse Ji
    Prodigy 170 points

    It was sent to me by Gregg Scott last year.  He is one of your guys.  Could you please ask him to jump into this thread?

    Thanks

  • 0 in reply to Steve Scarlata
    TI__Mastermind 49260 points

    Hi Steve

       I checked some remaining material, and find out the B and C version indeed has this I2C problem. It's caused by internal clock architecture affected by the noise from PCB. Would have the chance under high output power using. Only A version could totally avoid this problem.

       Sorry for the trouble, and I would suggest to consider using some newer device, like TPA6304. This old device has been through lots of organization change, and very limited resource left for us.

       I'll also notice Gregg to take a look.

  • 0 in reply to Shadow He
    Prodigy 170 points

    Thanks,  Yes but I'm quite certain TI has fixed this problem,  so I would like to know how to recognize the parts that will actually work.  These chips were used extensively in Automotive and other applications.  We realize we will need to re-design our product but first we need to do a last-time buy to continue shipping product.

    Did you quarantine the defective chips from the market?  Can you send a list of Date codes or part number suffixes that are not defective?

    Steve

  • 0 in reply to Steve Scarlata
    TI__Mastermind 49260 points

    Hi Steve

       After discussion with Gregg, he would try to dig up more data and solve your problem through e-mail. Thank you.