TLV61046A: Parts blowing

Hi Matt:

Sorry, too busy today. please let me reply tomorrow. Thank you!

Hi Matt:

May you share the waveform and layout with me? What's the load condition?

For waveform, please measure the Vin, SW, Vout with and without loading. Thank you!

The layout is attached to initial email.  The load condition is a small piezo buzzer rated at 10mA at 12 volts.  Thanks I will try to get waveforms later but please look at schematic.

Matt

Hi Matt:

It's better to do the analysis after getting the waveform. 

May I know where is the layout? I didn't find it..

Minqiu,

Here is the layout:.....Matt.  Look below mistake directly below

https://e2e.ti.com/

Hi Matt:

There is still something wrong with the attachment. And wait for the waveform same time. 

Hello Minqiu,

The schematic is enclosed in the message.  I can see it.  There must be something wrong on your end or the software.

Matt

Hi Matt:

Yes, I can see the schematic. But I can't find the layout. 

Hello Minqiu,

My apologies, I thought you meant the schematic instead of the PCB layout.  How do you want the layout presented?

Matt

Hello Minqiu,

Here is the layout:

Hi Matt:

According to the layout, we can find the Cout has a long routing to connect the GND and Vout. It will bring large parasitic parameters, and cause high spike during switching.

The line to the L1 is very week, you can enhance it. 

And where is the R4? The FB should short to Vin reliable. 

Besides, the waveform is also helpful for the analysis. 

Hello Minqiu,

Cout (C5) connects to a ground plane so routing to ground is very short.  I agree the line to L1 is thin, should be beefed up.  R4 is on the bottom of the board, I can attach part of the layout  so you can see it.  I will try to get you some waveforms ASAP.   Matt

Hello Minqiu,

Here are some waveforms:  Vin with Load

Vout with Load

SW with Load

SW without load

Vin without load

Vout without load

Hi Matt:

I just found the C5 is 16V rating. If the Vout is 12V, the effective value is very small actually. You can add more to ensure the effective value. 

According to the waveform, I found the Vout is unstable with loading. Is the loading fixed during that time? Or always do the transient?

And what's the loading? 10mA as you mentioned?  You can do a test with resistance to act 10mA loading. This can isolate the issue with converter or loading. 

Hello Minqiu,

Yea, I noticed that also regarding C5.  We should update rated voltage as 16 volts is not a lot of headroom.  Also you state more might be better for that?  By unstable do you mean the one volt oscillation on top of Vout  waveform at load?  That load was fixed while making that waveform picture.  It is a piezo buzzer with a rating of 10mA at 12 volts.  In actual operation of the product, the output would be turned on and off (via the enable line) rapidly (1-5 Hz) for a more enunciated beep..  I will check it out with a resistor drawing ~10mA and see what output looks like.  Thanks.

Regards,

Matt

Hi Matt:

Thanks for update. 
I also recommend you measure the waveforms in one image if possible. Including Vout, Vin, SW, better with Icoil.

Hi Minqiu,

Don't know if problem has been solved as we have had no more parts blowing issues.  My scope is out for calibration, so am unable to get pictures right now, but I will say that with a resistive load at 10 mA the Vout signal is flat without the slow oscillation riding on top.  I am currently working with Allen Wachter, your TI analog field applications engineer to see if he can help to resolve the issue.  Thanks for your help,

Regards,

Matt