TPS63802: Excess current draw at 3.3v output

Steven Pisciotta

Part Number: TPS63802

I have a battery powered application which outputs 3.3v. My design uses the recommended component values: Cin (4.7uF), L (470nH), R1 (511k), R2 (91k) and Rpg (100k). My layout matches the datasheet. My initial testing of the board, I found that it was drawing over 1A and Vout was just over 5V. I isolated the circuit with only a 500ohm load resistor and received the same results. I changed to the design (R1 to approx 450k) which decreased the output voltage to around 3.0v. This solved the problem.

Do you have any ideas on why this is so sensitive? I am worried that the circuit is very close to being out of control....

Also, when I probed the FB voltage while outputting 3.0v, I see 0.575V and the current increases by about 70mA. Seems like my circuit is not very stable. Thoughts?

Thanks,

Steve

over 5 years ago

0 Jasper Li over 5 years ago

TI__Mastermind 46905 points

Hi Steven,

did you observe similar behavior in several boards? make sure it is not assembling issue.

0 Steven Pisciotta over 5 years ago in reply to Jasper Li

Intellectual 270 points

Yes, all my boards have the same problem. I noticed that your dev kit has MODE going thru a 1M resisitor to GND. I have them tied directly I have it tired directly to GND. Could that be the problem?

0 Jasper Li over 5 years ago in reply to Steven Pisciotta

TI__Mastermind 46905 points

Hi Steven,

it is OK to connect the MODE to ground.

did you see the same issue if you change the resistor back to 511K from 450K?

0 Jasper Li over 5 years ago in reply to Jasper Li

TI__Mastermind 46905 points

hi Steven,

I closed the post assuming that the issue is solved, if not, just reply below

0 Steven Pisciotta over 5 years ago in reply to Jasper Li

Intellectual 270 points

We had better results with by decreasing R1 to 75k and R2 to 422k. That seemed to work, but i found that by slightly increasing the load on the 3.3v the current draw to increased progressively. It seems like the control loop is very narrow and small increases in the load make it unstable.

0 Jasper Li over 5 years ago in reply to Steven Pisciotta

TI__Mastermind 46905 points

Hi Steven,

could you share the schematic and layout? please also apply an EVM for reference. the device is not sensitive to feedback resistor or loading.

0 Steven Pisciotta over 5 years ago in reply to Jasper Li

Intellectual 270 points

0 Steven Pisciotta over 5 years ago in reply to Steven Pisciotta

Intellectual 270 points

I pasted images of the schematic and layout. Hopefully you can ready them.

0 Jasper Li over 5 years ago in reply to Steven Pisciotta

TI__Mastermind 46905 points

hi Steven,

i can't see the attachment.

0 Steven Pisciotta over 5 years ago in reply to Jasper Li

Intellectual 270 points

5482.schematic.pdf

0 Steven Pisciotta over 5 years ago in reply to Steven Pisciotta

Intellectual 270 points

6560.Layout.pdf

0 Steven Pisciotta over 5 years ago in reply to Steven Pisciotta

Intellectual 270 points

1581.Layout.pdf

0 Steven Pisciotta over 5 years ago in reply to Steven Pisciotta

Intellectual 270 points

5483.Layout.pdf

0 Jasper Li over 5 years ago in reply to Steven Pisciotta

TI__Mastermind 46905 points

Hi Steven,

the layout shows that the L2 short to VOUT, could you help to check?

0 Steven Pisciotta over 5 years ago in reply to Jasper Li

Intellectual 270 points

There is a gap between VOUT and L2. That line you see is the outline of the copper pour area, but when the copper is poured makes sure there is a gap between those two areas. They are not shorted. See attachment, which shows the L2_Vout.pdfboundaries...

0 Jasper Li over 5 years ago in reply to Steven Pisciotta

TI__Mastermind 46905 points

1. could you try add one more output capacitor

2. did you connect the ground of the FB resistor to the ground panel through the vias?

3. I would also suggest you apply an EVM for reference.

Power management

Power management forum

TPS63802: Excess current draw at 3.3v output