TPS7A8001 - Current Limit Operation

Hi Ryuji,

As per section 7.3.1 in the datasheet, the device has what is known as a "brick-wall" current limit. This type of current limit sources a constant current until the short is removed or the device enters thermal shutdown. Figure 13 in the datasheet shows the current limit value are multiple output voltages.

Current limit is an active loop in the circuit and it does have a finite response time - typically on most LDOs this is between 10 - 100 us, so the 60 us response time seems correct.

The Tj needs to be kept less than 125C - the device is specified for operation from -40C to 125C as per the EC table. Higher junction temperature decrease the life-span of the device.

The output conditions are not acceptable - the device is only rated for 1 A of current, not 1.8 A. For a 1.8 A load current use a device like the TPS7A8300.

The app note below explains reliability of semiconducting devices:

www.ti.com/.../spraby3.pdf

Regards,
David

Hello David san,

The 1.8A condition is only inrush(60us).
This is not continuous current.  Continuous current is 0.6A.
Is this acceptable?

Regarding current limit type.
I think that  there are three types of current limit as attached image.
Is "brick-wall" current limit B in the attached image?

Also, are there any plan to release the high current version of TPS7A8001 which has pin compatibility with TPS7A8001?

Best Regards,
Ryuji Asaka

Hello David san,

As I checked the detail to the customer, the 1.8A 60us is not inrush current.
This current is occurred when they accessed digital interface.

I understood that this current is not recommended range.
I proposed that they should change the device.

However, could you please let me know that the device was damaged on this condition?
Vin=3V, Vout=2.5V, Iout=1.8A 60us. Since they are using liquid cooling, the device Tj is not violated on the datasheet specification.

Best Regards,
Ryuji Asaka

Hi Ryuji-san,

The problem is that they may see long term electromigration on the die due to this, depending on how often it happens. These types of failures happen quite a while after the boards go are sent to their customers, and engineers can get an estimate using TI's FIT rate here: http://www.ti.com/lsds/ti/quality/reliability/calculators.page

As they are using liquid cooling the risk is less as Tj << 125C, and if the event happens with a low duty cycle (a few times every day) the risk is also less.

Regards,
-David

Hi Ryuji-san,

One other note - it is outside of the specified operating range for the device so my recommendation is still to use a higher current device where you are in the specified range.

-David

Hello David,

Thank you for the reply.

In the datasheet, absolute maximum ratings of Iout is internally limited and Tj is 150C.
From this specification, the absolute maximum ratings of Iout is limited by power dissipation and junction temperature.
Is my understanding correct?
On the condition of the absolute maximum ratings, there is the risk of electromigration?

Also, the 1.8A is occured many time on every day....
I understood that the recommendation is using higher current device.
However they couldn't change the device due to their board is fixed already.

Best Regards,
Ryuji Asaka