TAS5630 OVER-TEMPERATURE PROTECTION FAILURE

Did further testing , the same weird ringing occurs before thermal shut down ...

Videos showing the ringing on the output before the chip goes into thermal protection and the rising Thd figures as the temperature increases..

http://sendvid.com/5k87i4k0

 

http://sendvid.com/r4u2v03z

any pointer on what is going wrong would be very help full .

regards

ardhendu

Hello Ardhendu,

Unfortunately I am having issues viewing the videos.
Can you clarify a few things for me:

Are you using the EVM or your own board design?
Do you see the ringing every time the device approaches OT protection?
Are you experiencing the 12V short only after cycling the device into OT protection multiple times? (Is this short on one of GVDD pins?)After the short issue, does the device recover?

One thing that can help is to add schottky didoes on the output switching nodes (PWM_x pins) to ground. These help clamp the ground voltage bounce on the PCB at high output currents to the threshold voltage of the didoes. These should be surface mount diodes soldered as close as possible to the outputs and soldered directly to the ground plane on the PCB.

Best Regards,
Matt

hi matt,
i am using my own PCB design ,yep every time the device reaches OT i can see ringing .
Once the 12V shorts out to gnd the chip is damaged it doesnt recover . I have kept space in the PCB for schottky diodes , need to populate them and check .

regards
Ardhendu

Hello Ardhendu,

Thanks for sharing the video. I think the schottky diodes will help with this by clamping those noise spikes you see.
To clarify, are you using the TAS5630 or TAS5630B? Which package type, DKD or PHD?

Please send schematics and PDFs of the layout for review.
Are you using 2oz copper?

Regards,
Matt

i am using TAS5630B PHD , i aint using 2oz copper but all the high current tracks have been reinforced with solder mask removal so that add on solder can be used to reduce DCR of the tracks.

regards
Ardhendu

Hi Ardhendu,

I reviewed your layout. It’s a little difficult to tell exactly where all of the components go but it looks like you have a lot of long horizontal traces on the bottom side Cu. These will break up the ground plane significantly and cause weak grounding across the board. For example LC capacitors C32 and C88 should have a very strong ground path back to the ground pins on the amplifier. With the current layout, PVDD breaks up the bottom and top grounding quite badly. Vias cannot be placed to interconnect bottom and top ground for most of the high power section of the board.

I recommended improving PVDD traces so you can add significantly more visa between top and bottom grounds layers around the perimeter of PVDD traces (and any other long traces that break the ground plane). This will allow ground currents to flow between layers freely. There must always be a large ground path back to the amplifier on these high power devices.

I this grounding is causing a differential voltage to develop across the ground plane of the PCB (and the die on the TAS5630) at high output power, causing the issues you are experiencing. Schottky didoes should help keep the part happy by clamping these ringing voltage levels, but the grounding should be improved for best reliability.

Best Regards,
Matt

i matt ,
I used schottky diodes(SS26) on the output as u suggested, it even increased the efficiency a tad bit , this time the amplifier ran for longer duration of time without going into oscillation , but say after 10min the oscillation appears and heats up the schottky diodes significantly ..and eventually blows them up .

regards
ardhendu