LM5085: Short circuit issue

Julian Chen

Expert 2550 points

Part Number: LM5085

Hi E2E,

I use WEBENCH to design, and choose LM5085 to design.

Vin=56V, VOUT=50V, IOUT=0.5A ; The circuit as below.

When I test short circuit protection (Vout touch GND), The IC is burned out every time.

We find out the VCC pin is damage of burned out IC, the impedance just only 8.8 ohm~200 ohm. (OK IC is 4M~8M)

I try to modify the RADJ to 200R (current limit is 800mA), but IC still be burned out again.

How to modify the circuit to comply SCP?

How to choose D14 IF value?

over 5 years ago

0 Samuel Jaffe over 5 years ago

TI__Mastermind 25590 points

Julian,

50V short to GND is a harsh transient. Please observe the voltage on the pins of the IC during this event to see if any voltage is violating abs. max in the datasheet. Especially check VIN, VCC, ISEN, and ADJ. Also check if the diode and FET are still working after this event.

-Sam

0 Julian Chen over 5 years ago in reply to Samuel Jaffe

Expert 2550 points

Hi Sam,

I measure VCC and find overshoot voltage to damage it. Is it normally?
Could we add Zener diode from VCC to GND for clamp it?

0 Samuel Jaffe over 5 years ago in reply to Julian Chen

TI__Mastermind 25590 points

Julian,

Yes you could add a zener but we should find out what is causing the overshoot to begin with. Is this overshoot at startup?

Please send a scope shot showing VCC and VIN during the overshoot.

Thanks,
-Sam

0 Julian Chen over 5 years ago in reply to Samuel Jaffe

Expert 2550 points

Hi Sam,

After I push customer to measure the waveform which is short circuit instantly.

BLUE-VIN to GND
Yellow-VCC to GND

I think that is not overshoot to damage IC

0 Julian Chen over 5 years ago in reply to Julian Chen

Expert 2550 points

Hi Sam,

I update the status.

Our customer only had 1 borad can comply SCP.
When SCP happend, PMOS Gate will goes high in a moment, so Vin can't drop with VOUT.

They use same PMOS at all. So what SPEC of PMOS shoule we attention?

0 Samuel Jaffe over 5 years ago in reply to Julian Chen

TI__Mastermind 25590 points

Julian,

That waveform shows VCC reaching about -10V. This can cause damage to the IC. You may need more input capacitance and you may need a new layout. Please add 10uF and test again. Also share the layout so I can take a look.

Thanks,
-Sam

0 Julian Chen over 5 years ago in reply to Samuel Jaffe

Expert 2550 points

Hi Sam,

Thanks for your support.

Could you kindly let me know your e-mail address?

Thanks!

0 Samuel Jaffe over 5 years ago in reply to Julian Chen

TI__Mastermind 25590 points

Julian,

I've received your email. We will carry on our discussion there, then circle back here with our results.

-Sam

0 David Baba over 5 years ago in reply to Julian Chen

TI__Mastermind 33050 points

Hello Julian,

A fast negative voltage transient on Vin (<8V) will cause VCC to go negative because the Voltage at VCC is being regulated from Vin and not from Gnd of the Device. As such the the VCC capacitor has the regulated VCC voltage across it as Vin goes negative. This is highlighted in the Datasheet and you must place a schottky from Gnd of the device and VCC. Please refer to figure 24 of the datasheet. There you will find the details I am referring to there.

Alternatively can can prevent the negative voltage transient at Vin by improving the transient response of Vin supply, or increasing the input capacitance?

Hope this helps?

David.

0 Julian Chen over 5 years ago in reply to David Baba

Expert 2550 points

Hi David,

Thanks for you reply.

Customer measure the waveform to us and find out PGATE don’t go high when short circuit.

How can I check that why PGATE don’t goes high to protect the circuit?

(However, PGATE voltage to -2xV is unusual, I will check it again.)

0 Julian Chen over 5 years ago in reply to David Baba

Expert 2550 points

Hi David and Sam,

Thanks for your support!

We Add Cin can resolve it!

Thanks!

Power management

Power management forum

LM5085: Short circuit issue