TLV9051: Current Measurement with TLV9051IDBVR

Hi Shibin,

the output voltage of TLV9051 must stay in its linear operation range:

Otherwise the output hangs in saturation and doesn't work properly.

You could modify your circuit to run as a differential amplifier while adding a pseudo ground to the REF pin of differential amplifier. Or you can add a small negative supply voltage by the help of LM7705.e.g..

Kai 

Hello Kai,
The input of the TLV9051 is from [(V–) – 0.1] to [(V+) + 0.1]. So the op-amp must stay in its linear operating range.

What is the minimum value of the shunt resistor I can go without any distortion for the signal?.

Hello Shibin, 

As kai mentioned, the output range must be considered. A single supply amplifier will not be able to output down to 0V. Both of Kai's suggestions would work well to ensure suitable operation. 

Please see this TI Design for resources regarding a similar sensing application: 0-1A, Single-Supply, Low-Side, Current Sensing Solution Reference Design

I figure the 100mOhm resistor is reasonably sized for the sensing application. You may be able to reduce this value a bit, but eventually the Vos will become a significant portion of the signal you are seeking to measure. 

I am less familiar with zero-drift amplifiers, but it is safe to say the OPA2333 would outperform the TLV9051 in many specifications. 

Is the zero drift amplifier being considered as an alternative to software calibration? 

Are there any accuracy requirements for this current sense application? 

Best,

Jacob

Hello Jacob,
Thank you for your reply.
Is the zero drift amplifier being considered as an alternative to software calibration? 
     Yes, if a zero-drift amplifier is needed, I can move forward with that.
Are there any accuracy requirements for this current sense application?
      We are looking for the best accuracy we can achieve.
When I googled for solutions, one thing I can find is to use a current sense amplifier like INA180B3IDBVR.
Will INA180B3IDBVRdo my job better?.
Looking for your advice.

Hi Shibin,

kai klaas69 said:

You could modify your circuit to run as a differential amplifier while adding a pseudo ground to the REF pin of differential amplifier.

Now I have found the link to the pseudo ground method:

tidu675.pdf

Shibin Varghese said:

When I googled for solutions, one thing I can find is to use a current sense amplifier like INA180B3IDBVR.
Will INA180B3IDBVRdo my job better?

These specialized current sense amplifiers are optimized for high side current measurements by providing an extremely high common mode rejection. In a low side current measurement, on the other hand, a good OPAmp is equally well or even better suited.

Kai

Hi Jacob,
Thank you for your reply.
The opamp OPA333AIDR seems like a good option for my application.
But in the Applications of the IC, the current measurement is not included.

Even though it has the exact specifications as that of OPA2330AIDRBT

In my application, the current measurement is unidirectional. So I believe a noninverting amplifier like the following will do my job without U1B op amp.

The OPA333AIDR  is a Rail to Rail opamp.
using another IC like LM7705 will increase my BoM size and more real estate on the PCB.
The output swing of the OPA333AIDR  is a maximum of 70mV from all rails.
I believe this won't be an issue for my system.
Because the current to be measured will be either 0mA, 250mA, 500mA, 750mA, or 1A.
For a 100mE  shunt resistor, the voltage across the shunt resistor for 250mA current will be 100mE*250mA = 25mV.
The Common-mode voltage range of the  OPA333AIDR  is (V–)–0.1 to (V+)+0.1.
The output voltage will be 250mV with a gain of 10V/V.
Please correct me if I am wrong.
If my shunt resistor is set to 10mE, the voltage will be 2.5mV for 250mA current flow.
Setting the gain to 100, I will get the same output voltage.
Will this configuration cause any issues?.
The input offset voltage of the OPA333AIDR is only 10uV.
Also, could you please explain the other parameter that affects when the shunt resistor is made too small?.

Hi Kai,
Thank you for your reply.
Could you please explain what you meant by adding a pseudo ground to the REF pin of the differential amplifier?

Hi Shibin,

Shibin Varghese said:

Could you please explain what you meant by adding a pseudo ground to the REF pin of the differential amplifier?

Just what you have shown in the last figure.

Don't overlook that the precision of this circuit is also affected by the pseudo ground generation with R5 and R6. Not only the tolerances of these two resistors affect the accuracy of pseudo ground but also the buffer OPAmp U1B and the supply voltage Vcc itself. And then you have the tolerances of R1, R2, R3 and R4. Ok, Vcc may be replaced by a precision voltage reference, or the error of Vcc could be eliminated by the help of ratiometric measurement, if Vcc is used as reference voltage for the ADC as well.

A circuit with the additional LM7705, on the other hand, would be way more precise, because the pseudo ground generation can be totally omitted and only two gain setting resistors affect the accuracy with their tolerances. So, when Jacob wrote

Jacob Nogaj said:

If I were tasked with designing the circuit, I would pick the LM7705, a low offset amplifier, and low tolerance resistors. Depending on your accuracy requirements, you could likely skip the need for software calibration all together.

then he was quite right.

But how to proceed also depends on the ADC. If it can accept input voltages down to 0V, I would go for the LM7705 solution. In the other case I would go for the differential amplifier with the pseudo ground.

Kai

Hi Shibin, 

Thank you for the additional details.

I hope our suggestions are getting you a bit closer to a solution.

Correct, you should get 250mV output using both methods described above (excluding any small error sources of course).

OPA333 is a good choice if you don't require the wide bandwidth of other precision amplifiers. 

You can use the 10mOhm shunt if you would like. Note, it is imperative that you run the sense traces properly when using a sense resistor of such low resistance. Here is a blog post regarding design considerations for low-side current measurement: https://e2e.ti.com/blogs_/b/analogwire/posts/how-to-lay-out-a-pcb-for-high-performance-low-side-current-sensing-designs

Shunt resistors are typically selected to balance power dissipation with output voltage. Naturally, a larger output voltage from the shunt allows for amplifiers with a larger typical offset voltage to be used. 

In your case, there would be few disadvantages that come to mind when using a precision amplifier with a small shunt resistance. Again however, even a relatively short sense trace can add to this total resistance, so exercise caution in routing. 

Please let me know if you have any questions. 

Best,

Jacob