2018-01-28 08:48 PM
I've just managed to blow up 2 TDA7498E D Class amp chips when testing the first run of 25 production boards. This is the first time we have managed to damage any chips in over a year of prototype testing.
We are using the chips in mono mode and have the left and right outputs tied together as per the datasheet (Mono mode is selected by connecting the mode pin 31 to VDSS pin 26).
The damaged boards have the four VCC (+35V supply pins) blown off the chip (see attached photo). The chip blows when audio is first output during one of our factory tests (you get a small bang and a puff of smoke).
I can get the same result by putting the chip into stereo mode by floating the Mode pin and playing audio (this causes the left and right amplifier outputs to short each other out). I also noticed that when the chip is in Standby the VDDS supply is turned off (e.g. at 0V).
What I think may be causing the problem is taking the chip out of standby with audio present on its audio input as follows:
1.
Put chip into standby mode (VDDS supply is off and at 0V, so Mode input is in ‘Stereo’ state).2. Apply external audio signal to the INPA pins
3. Take the chip out of standby
I assume as the mode pin is initially at 0V the chip powers up in stereo mode and then switches to mono mode when VDDS supply >2.5V. If the audio amplifier starts playing before the chips switches to Mono mode the output stages short out. I also assume if you power up the chip with audio present it will also blow.
Could STM confirm that powering up the chip or taking the chip out of standby with audio present on its input (with the chip configured for mono operation) is causing the chips to blow, or what is the likely cause.
What the recommended power up and standby sequence is when using the chip in mono mode (e.g. Mute chip for x seconds before powering up the chip or taking it out of standby etc).
Regards,
Hugh
JCI NZ
2023-05-18 06:02 AM
Attached a picture of my product with same pins burnt, using TDA7498E in PBTL. I have about 6 chips with same condition taken out from burn-in products from our pilot run.
I think ST must provide a clear explanation. Really hard to believe that you cannot aplly signal during turn-on in PBTL.