LMP2012: Current Sense Circuit fails in low temperature

yan zhang

Part Number: LMP2012
Other Parts Discussed in Thread: OPA365, OPA387, OPA328, OPA392

Hi Everyone.

I am using a LMP2012 for a current sense circuit as following

This circuit inherits from a old OPA365 version, and works very well.

Unfortunately, when I changed OPA365 to LMP2012 and kept all the resistors and capacitors untouched, this circuit fails.

The output of the LMP2012 was always 3.3v nomatter what the real current is in LOAD.

Firstly, I guess the rail-to-rail performance of LMP2012 is not as good as OPA365, and change the attenuation resistor （R2 and R4）from 10k to 4.7k to reduce the input voltage of LMP2012

This new circuit works in room temperature, but when the temperature descends below 0 centigrade，the output of LMP2012 goes up again and finally stops at saturated point of 3.3V at -10 centigrade.

So my question is which parameter difference betweet LMP2012 and OPA365 cause this problem and how to solve it if I have to use LMP2012 instead of OPA365.

Thank you in advance.

Tim

7 months ago

+1 Michael Hartshorne 7 months ago

TI__Genius 11925 points

Hi Tim,

As you correctly diagnosed, the LMP2012 is not rail-to-rail input. This can be seen in the spec. below:

This spec. shows that the input common mode voltage must sit 1.8 V lower than the supply voltage. For 3.3 V supply then the input common mode must be below 1.5 V.

Even reducing the gain in your diff. amp., as you have done, doesn't lower the input common mode voltage enough; it can be simply calculated by the resistor divider from the 3.3 V rail to GND: Vp = 3.3*(4.7/(1+4.7) = 2.72 V.

This part would not be the best choice for high-side current sensing; I would look at a device like OPA392 (10 MHz), OPA328 (40 MHz, zero crossover), or OPA387 (zero drift); these would be our latest devices that would work well here depending on your frequency response and accuracy needs.

Regards,
Mike

0 yan zhang 7 months ago in reply to Michael Hartshorne

Prodigy 10 points

Hi Mike.

Thank you very much for your answer.

I have searched the Amplifiers forum with your key words "common mode" instead of my "rail-to-rail", and found this issue has been discussed 5 years ago.

TINA/Spice/LMP2012QML-SP: Input common mode range when powered by 3.3V and Pspice model. - Amplifiers forum - Amplifiers - TI E2E support forums

I will modify the circuit to use a two-stage LMP2012 to achieve the "difference" & "amplify" respectively.

In my next version I will try the OPA392 or OPA328 from your suggestion.

Thanks