TPS552882: blown chip when rapidly removing load

Alex Ortiz

Part Number: TPS552882

Tool/software:

Hello. I'm having an issue where this buck-boost converter chip becomes damaged when I have a load on the output that is rapidly removed, like disconnecting a battery that is being charged. The damage to the chip is that the SW2 pins (21 and 25) become shorted to ground (pin24), resulting in the external inductor overheating, and the converter stops working. The load I have attached is no more than 3A. How do you recommend I prevent this from happening? It seems to me to be caused by a load-dump inductive spike at the inductor that exceeds the Vds of the internal low side FET, irreversibly damaging it.

Please see my design using this link: webench.ti.com/.../SDP.cgi

Thank you!

Alex Ortiz

1 month ago

0 Bryce Xun 1 month ago

TI__Mastermind 20750 points

Hi Alex,

Can you share the schematic and layout with us?

If the issue is easy to replicate, can you share the waveforms of SW1, SW2, Vout and inductor current at the time of failure to help find the cause?

BRs,

Bryce

0 Alex Ortiz 1 month ago in reply to Bryce Xun

Prodigy 20 points

Hi,

Thanks for the response. Please see attached schematic and layout. I will send the waveforms early next week.

Schematic: /resized-image/__size/1149x706/__key/communityserver-discussions-components-files/196/1108.Schematic.jpg

Layer 1: /resized-image/__size/503x819/__key/communityserver-discussions-components-files/196/2678.Layer1.png

Layer 2: /resized-image/__size/580x793/__key/communityserver-discussions-components-files/196/2678.Layer2.png

Layer 3: /resized-image/__size/549x793/__key/communityserver-discussions-components-files/196/2678.Layer3.png

Layer 4: /resized-image/__size/526x715/__key/communityserver-discussions-components-files/196/0121.Layer4.png

Please let me know if these are not clear enough.

Regards,

Alex

0 Mulin Yuan 1 month ago in reply to Alex Ortiz

TI__Expert 5720 points

Hi Alex,

I will check this and reply to you soon.

Regards,

Mulin

0 Mulin Yuan 1 month ago in reply to Alex Ortiz

TI__Expert 5720 points

Hi Alex,

Thank you for your waiting. I am sorry that the schematic is not clear to me. Can you send a PDF version? Can you mark capacitance and resistance beside each part? Also I cannot tell part number. It seems that layout is also poor. Anyway I need the clear schematic first.

Regards,

Mulin

0 Alex Ortiz 1 month ago in reply to Mulin Yuan

Prodigy 20 points

Hello,

Sorry about that, here are better quality images of the schematic/layout.

6472.Schematic.pdf 6472.Top.pdf Inner2.pdf 5824.Bottom.pdf

0 Mulin Yuan 1 month ago in reply to Alex Ortiz

TI__Expert 5720 points

Hi Alex,

Thank you for your waiting. What is the input voltage range and maximum output current?

About schematic:

1. Recommend to add two 10uF ceramic capacitors and one 0.1uF 0401 capacitor in input voltage rail . Put the 0.1uF capacitor close to buck FET loop.

2. Recommend to use same MOSFET in buck leg in case of asymmetrical turn on/off causing shoot through. Change R33 and R34 to 0.

3. Recommend to add RC snubber in SW1 and SW2. 2.2nF+2ohm(1206).

4.Recommend to add a 0402 0.1uF capacitor in Vout pin and put it close to IC.

5. I am not sure if you use LM74700 correctly here. Can you submit a new thread about this part to seek advise from LM74700 expert? It seem that a load switch rather than Reverse Battery Protection Ideal Diode Controller should be used here.

6.Please provide part number of C25-C30 for loop calculation.

About your layout, I have to say it is really poor. Do you have a chance to design a new version? Please refer to our guideline here.8512.TPS55288 Layout Guideline (Rev B).pdf

Key points:

1. Power loop trace: Vin, Vout, SW2. Current trace is too thin. This will cause huge parasitic parameters when large current is through.

2.Input Output loop. Put small capacitor to minimize loop. Refer to guideline.

3.Drive loop.

4.Two buck MOSFET.

5.Thermal vias in Vin Vout GND planes.

Anyway please refer to our guideline and design a new version.

Regards,

Mulin

0 Alex Ortiz 1 month ago in reply to Mulin Yuan

Prodigy 20 points

Thanks for the feedback. My input range is 9-15V, max current output is 3A. I will try to repeat the failure (it has only happened once) and get back to you about a cause. The chip works as expected other than that failure.

How should I put the snubber on the SW1 and SW2 circuits? Is it one snubber per switching node? How is that circuit done exactly?

Also, I'm using the TPS552882 as a battery charger to charge at 3A maximum, at 13.6V, which is roughly the trickle charge level of a lead acid battery. Is that an appropriate use of this chip? I see it's used for USB battery charging and has active current limiting which is why I chose it.

The ideal diode controller is just there to block negative current from the battery, but I'm not sure it's needed so I may remove it after some parasitic current testing.

0 Mulin Yuan 1 month ago in reply to Alex Ortiz

TI__Expert 5720 points

Hi Alex,

I will reply to you tomorrow.

Regards,

Mulin

0 Mulin Yuan 1 month ago in reply to Alex Ortiz

TI__Expert 5720 points

Hi Alex,

About RC snubber, please refer to EVM board. Recommend to use 1206 package resistor.

TPS552882 can be used as battery charger, but it is not a common application. Do you have a EVM board? I think you can use a TPS552882 EVM first to see if it is suitable for you application. If it works well , then you can modify your design and layout.

TI also have charger IC, which has more precise CC-CV function. You can also look for some suitable device.

Regards,

Mulin

Power management

Power management forum

TPS552882: blown chip when rapidly removing load