OPA2990: Using An Op Amp for High-Side Current Sensing , is it ok for us to have negative output?

This op amp (like other rail-to-rail op amps) can handle all voltages between the voltages of its own power supply.

Please show a schematic of your desired circuit, including all bus and power supply voltages.

Hello,

current sense ref diagram.pdf

6170.current sense ref diagram.pdf

0574.current sense ref diagram.pdf

Hello,

Hello,

Hi Maitry,

why not running a TINA-TI simulation?

Kai

Hello,

thanks for your feedback. we will be using op-amp with dual supply +7 V and -7 V, not with the single supply. 

We will try to simulation in TINA simulator. 

Also, Can you please suggest any other suitable op-amp for the specified requirements? then, it will be very helpful

Hello,

I have gone through the given video for current sense implementation https://training.ti.com/ti-precision-labs-current-sense-amplifiers-current-sensing-different-types-amplifiers?context=1139747-1139745-1138708-1139729-1138709.

Thank you for suggesting to go through the video, It clears a lot of concepts for current sensing. Based on that, I have below query.

For OPA2990 , dual supply range is +/- 1.35 V to +/- 20 V. If we use opa2990 with dual supply + 7 V at V+ and -7 V at V- . 

VCM range will be =( V- )-0.2 TO (V+ )+ 0.2 V , right. 

So, VCM range for dual supply + 7 V at V+ and -7 V at V-  will be -7.02 to +7.2 V . 

And as per our requirement for -5 V Vbus and 1.4 uA current. 

Vshunt will be 1.4 uV.

VCM = (-5 - 5.0000014)/2 = -5.0000007 V ( which is inside the range of VCM ) 

Please verify. and suggest the above understanding is ok or not? 

If not, then, I think ,we have to search for difference amp or current sense amplifiers , in which we can get VCM range > Vs, right?

Please guide asap.

waiting for your positive response.

Hello Maitry,

Of course you can measure a difference voltage that has a common mode of -5V when supply is +/-7V. Use a difference amplifier with perfectly matching resistors. Normally they wouldn't have to be perfect.

To accurately measure 1.4uV difference riding on -5V does require near perfection. Your unwanted signal is -5V (noise) and the wanted signal is 1.4uV. So your signal to noise ratio is -141dB 

For the op amp, it needs to be nearly perfect too. For 1.4uV signal, the input offset of the amplifier should be less than 0.01uV, OPA2990 isn't close at all.   

Can you make the shunt voltage vastly larger with a larger sense resistance?

Hi Maitry,

you could do it this way:

maitry_opa2990.TSC

Kai

Hello,

Yes, It is ok for us to use larger sense resistor.

Hello,

Thank you for the guidance and detailed simulation given.

But, we have a size constaints for our application, that's why we are trying to acheive this current sensing using single differential amplifier. Is it possible to have current sense of 1.4 uA with -5 V supply using single difference amplifier , if we go for the difference amplifier that has very lower input offset voltage and if we increase the value of sense resistor as suggested by Ron.

Kindly reply ASAp.

Hi Maitry,

Using a difference amplifier configuration with a -5V common mode voltage and trying to measure a 1.4uV signal will require you to use resistors with approximately 0.0001% tolerance based on Ron's calculations. This is because the CMRR of a difference amplifier is dominated by the resistor tolerance. Using a higher precision device will help with offset and other DC error sources but the difference amplifier configuration is a limiting factor due to resistor tolerance.

Thank you,

Tim Claycomb

Hello,

Thank you for your quick response.

What do you suggest to use for this requirement ? if possible , kindly suggest suitable part number.

Maity,

Is the 1.4uA conventional current flow really from -5V to load. If the load was resistive the current would flow the other way.

Do you want to keep the voltage of the load a constant -5V , meaning correcting for the voltage loss across sense resistance?

Is ground and 7V the only supplies available for use?

Maity,

With a +7V and -7V supply and the current flow is the other way, I have two solutions. 

This one measures current in the one direction and outputs zero in the other direction (no negative output).

VLOAD is 5V constant.  VOUT = ILOAD * R4 * R3 / R1  ; VOUT >=0

Neg low I sense3.TSC

This one measures current in both directions, output can be positive or negative.

VLOAD is 5V constant.  VOUT = ILOAD * R4 * R2 / R1 ; VOUT + or -    R3=R1,R5=R2

can change to +/-7V , and OPA2990

Neg low I sense4.TSC

Hello,

It is not necessary to use +7 and -7 V as a supply for op-amp.

For load current flow, In detail, below is our requirement.\

we want to sense the gate current of HPA. Gate voltage variation : -5 to -2 V

Gate current : 1.4 uA in normal condition and in unexpected condition: max up to 10 mA.

This gate voltage supply is connected to the gate pin of HPA through current sense resistor and op-amp. 

so, if we select any of the high precision op-amp from TI and use sense resistor in such a way( that the difference voltage is higher than the offset voltage) , and with use of R1, R2, R3 and R4 (we get a slight gain). It is ok for us to have negative output, then after we will use voltage divider to shift the same to the ADC pin range. 

But, our main concern is to implement all the above with any single high precision op-amp because of the size constraint. I request you to guide for the same. Is it possible to implement above using any high precision op-amp(single)?

waiting for your positive response.

Hi Maitry,

why didn't you tell earlier that up to 10mA can flow?

Kai

Maity,

I was suggesting better solutions. I see that you would prefer a simple solution.

Of course a single op amp could do this. Before suggesting a simple solution, I need to know with certainty which way the load current flows. 

I see the voltage varies -2V to -5V and it can provide 10mA but the HPA should only draw 1.4uA; therefore we can use this voltage to power the single op amp. I wonder what makes to -2V to -5V (does it not have a negative power supply?)

How fast does the output voltage change?  (max slew rate or Vpp and frequency)

Does the controller that reads the ADC also know the value of the output voltage?  (if it knows both, the CPU could use math on both values to get the right answer)

With the answers to all questions, I could suggest a simple (small) circuit 

Maitry,

VBUS is -5V? 1.4µA shall be measured? What error is allowed? What is the load? What is the allowed maximum voltage drop across the shunt? What maximum shunt is allowed?

Kai

Hello Ron,

Thank you so much for your prompt reply.

i will again explain detailed application.

I have power amplifier as laod which will draw 1.4uA current and in case of any problem occure such as shorting it will draw 10mA current so i have to countiunue sense current for knowing device condition.

herewith attcahed block diagram for the same.

input voltage of my power amplifier is -5V in intial stage and in working condition voltage is -2V, in both draw current is almost same in uA. 

Input voltage -5V to -2V negative supply and it switching is in micro second.

yes, Controller ADC will know the value of output voltage.

Please suggest simple small circuit.

Thanks in Advance.

Hello,

Actually, in normal condition, the current will be ramain in uA range. but in unexpected situation such as shorting, the current can goes upto 10 mA. 

Sorry for misunderstanding.

Hello, 

VBUS is -5 V to -2 V. 

Load is power amplifier gate.

max allowed voltage drop is: 0.5 V

Hello,

yes 50 ohm shunt is ok .

Yes, -5V  to -2V will go in gate pin of power amplifier via current sense resistor.

only one time we are switch from -5V to -2V, after that it will reamin -2V. 

we will read output voltage of OPAMP through ADC pin.

As our main current is in uA, and input voltage through shunt resistor is -5V .

Thanks 

Hello,

Please provide response. As I am stuck at middle of my project.

Thanks in advance.

Hi Maitry,

this way?

maitry_ina849.TSC

Kai

Maity,

Here is an option that sends the current information to the ADC on two single ended channels or one difference ADC channel. In case of output being under heavy current demand, the 50k resistor will keep that current quite low (<100uA). Vloss at 1.4uA is 70mV

The ADC voltages can range from (1V to 7V) 

Neg Low I sense6.TSC

Hello,

Thank you for your reply and guidance. I will go through it and give you the update.

Hello,

Thank you for the guidance.

Hi Maitry,

does it work now?

Kai