LMP8481: LMP8481 temperature drift on 1 in a 100

Bram de Haas

Part Number: LMP8481

I have a design where i use the LMP8481 to sense a DC current. I made a 1000 boards and 10 of them have an issue where the output voltage at no current drifts as I increase the temperature with a heat gun.

I have check if it was the reference voltage that was drifting but that seemed to be stable on both the good boards and the bad board.

Also when i replace the LMP8481 the problem seems to resolve.

Is this expected or predicted behavior?

over 5 years ago

0 Javier Contreras4 over 5 years ago

TI__Mastermind 29830 points

Hello Bram de Haas,

Thanks for using the fourm to get your question answered. The specification is +/- 900µV VOS over temperature. Is the value worse than 900µV over temperature?

0 Bram de Haas over 5 years ago in reply to Javier Contreras4

Prodigy 30 points

Hi Javier,

Thank you for the quick reply.

I see a 10mV drift when I apply heat to chip, but its only on 1 in 100 of the chips.

Kind Regards,
Bram

0 kai klaas69 over 5 years ago in reply to Bram de Haas

Guru 140200 points

Hi Bram,

can you give some more details and show us your schematic?

What shunt do you use? What is the gain? What is the common mode input voltage? Just some more details, please.

Kai

0 Javier Contreras4 over 5 years ago in reply to Bram de Haas

TI__Mastermind 29830 points

Hello Bram,

I assume that you mean this is at the output drifting. I have also have a question of your setup especially the gain that you are using.

0 Bram de Haas over 5 years ago in reply to Javier Contreras4

Prodigy 30 points

Hi,

correct, the output is drifting up. We use a 20x gain, but the drift is measured even with no current flowing through the sense resistor .

-Bram

0 Javier Contreras4 over 5 years ago in reply to Bram de Haas

TI__Mastermind 29830 points

Hello Bram,

I just wanted to show how much can be expected in the worst case for VOS based on the specifications. This will not be all parts but could be at the tail ends of the distribution.

An example of a worst case for VOS is having a -265µV VOS at room. Over temperature change to +900µV of VOS.

From these two worst case VOS numbers of -265µV to 900µV we get a delta of 1.165mV for VOS from room to the full temperature range. The delta of VOS could cause a change of the output of 1.165mV *20.2 = 23.53mV. 20.2 Gain comes from gain over temperature.

VOS error is added to the input so even when Vsense = 0V the VOS is added at the input and Gained up by the amplifier.

You also could see other contributor to VOS error due to PSRR and CMRR. These can be calculated from the difference in your application vs the conditions in the datasheet. Which is VCC and VCM = 48V. Training links below go over more in detail these error contributors.

https://training.ti.com/getting-started-current-sense-amplifiers-session-9-understanding-common-mode-voltage-error?cu=456802

https://training.ti.com/getting-started-current-sense-amplifiers-session-11-understanding-power-supply-rejection-error?cu=456802

0 kai klaas69 over 5 years ago in reply to Bram de Haas

Guru 140200 points

Hi Bram,

I agree with Javier: 900µV x 20 = 18mV. So, you should not be surprised to see a shift of 10mV...

Kai

Amplifiers

Amplifiers forum

LMP8481: LMP8481 temperature drift on 1 in a 100