TINA/Spice/OPA333: Replacing OPA333 to OPA388

Sivakumar,

GBW of OPA388 is much higher than OPA333, 10MHz vs 350kHz, respectively, which may result in higher integrated noise - use a low-pass filter to limit the bandwidth to the same frequency to see if this causes the issue.  Also, OPA388 has higher max IB possibly leading to higher offset related error.  If you need further assistance, please click on insert or attached botton shown below and include your circuit schematic showing supply voltage and value of all components used.

Hi Marek,

Thanks for the reply. I have simulated the design OPA388 with low pass filter at cutoff frequency as 339KHz and achieved the result as per the attached table and schematic circuit.

The output current is become linear only above the 9mV input, but i need to achieve the 25nA as a minimum. Kindly suggest what else have to be updated.
And the output current is varying with respect to load resistance but i have to Current Source_Schematic.TSCprovide the constant current source as independent of load resistance. Kindly suggest the circuit to make this circuit as constant current source and constant current sink.

Thanks,
Sivakumar

Sivakumar,

The problem you see comes from the output swing limitation of OPA333 and INA326 - they both can rougly get within 10mV of negative rail but to support your  25nA requirement the outputs would have to be within 2.5mV - see below.

A simplest solution to your problem would be to use a negative charge pump (-0.23V), LM7705, to bias your negative rails of OPA333 and INA326 as shown below.

Current Source_Schematic edited.TSC

Hi Marek,

Thanks for the reply. I have simulated the constant current source and constant current sink approach. The following limitation are facing with the design.
1. Maximum Output Current: My requirement is the output current should be independent of load resistance and output current range is 0-50uA. For example i fixed the load resistance is 50Kohm and Rset is 10Kohm.
Maximum load current Iout_max = Vs/(Rset+Rload) = 55uA
The step variation is 2.5mV for 25nA so i can able to achieve the maximum output current is 33uA.
Is it anyway to achieve the increase maximum output current range independent of load resistance?
(Note: The load resistance value is not fixed, its variable resistive load)
2. Accuracy: Need to achieve the 0.1% accuracy. But the simulation results shows more than 5%. May be the actual PCB current output comes with more accuracy compare to simulated result. But i wanted to fine tune the circuits in simulation itself before proceeding PCB.
Is it anyway to improve the accuracy?
3. I have prepared comparison list for selecting the DAC and ADC. By going through this Analog devices parts are more suited i think or else please suggest the TI parts for this application. Selection Criteria is: Communication Interface: I2C, Channels: 4, Resolution: Min 16 bits, Input Channel type: Single Ended

Source.TSC

Sink.TSC

Regards,

Sivakumar

Sivakumar,

With the current values of Rload and Rset resistors and 3.3V supply voltage, the circuit is capable only of maximum output current of 30uA becase Vout_opa node cannot get higher than 3.3V supply - see below.

One way to meet 0 to 50uA output current range would require changing the values of Rset and Rload resistors - see below.

This would translate Vin input range from 0 to 2.5V into 0 to 50uA output current - see below.

Another way would be to leave the circuit unchanged with the exception of increase in the supply voltage to 5.5V - see below.

This would translate Vin input range from 0 to 5V into 0 to 50uA output current - see below.

Hi Marek,

Thanks for the reply and suggestion. I have only 3.3V as a input supply and i don't want to allow switching module comes into the board to up convert the 3.3V because i have precision analog circuits in the board. So planned to use the 3.3V alone for all operations.

When i am changing the Rset value below 10Kohm to increase the maximum current output but the accuracy of the current output is reducing. How to overcome this issue?

Refer the attached sink circuit, there is non linearity in the output current. How to solve this?

Regards,

Sivakumar

7215.sink.TSC

Sivakumar,

The theoratical conversion factor between the input and output is: Vin/Iout=2.5V/25uA=100k but there will always be an error caused by Vos of OPA388 and Vos of INA326.  For that reason, in order to get precise conversion factor you must calibrate the system as shown in the simulation below:

Hi Marek,

If i increase the Rset value to around 100Kohm the output current range is going to reduce. But i have to provide the current range up to 50uA.

Initially i don't want to use switching module (boost converter) in my design but now i am forcing to use this.

Please suggest the best boost converter for this design and how to arrest the losses in terms of resolution and accuracy if am using the boost converter (3.3V to 5V).

Regards,

Sivakumar

Hi Marek,

There is linear 1.47nA offset current in every conversion. How to nullify this offset current in hardware configurations itself.

Regards,

Sivakumar