This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

INA282: common mode voltage question

Part Number: INA282
Other Parts Discussed in Thread: INA283

Hello, I am using the INA282 in a new design. I inserted a schematic showing how I am using this part in the design. The bus supply is 15.7 VDC. The nominal power draw to the "Trillium_Power+" node is 0.650W and the maximum power consumption is 5W. I am trying to calculate the total error and am a bit confused about what the common mode voltage is for my application. Based on the information in the datasheet I think the common mode voltage is 15.7V but I am not sure if this is correct. 

Thank you for your help and time. 

-Sean

  • Hi Sean,

    as the maximum voltage drop across the shunt is only about 48mV, your common mode voltage is very close to 15.7V.

    Kai

  • Sean,

    15.7V is correct here.

    Common mode voltage is defined as the average voltage seen between the IN+ and IN- pins, or (IN+ + IN-)/2. By your current diagram, IN+ will always see 15.7V. As Kai points out, in the worst case (5W load), the voltage created across the shunt would be approximately 48mV, so the voltage seen at IN- will not drop by very much, and the common mode would remain around 15.7V. An exact answer in this case would be (15.7 + 15.652)/2 = 15.676 which is still approximately 15.7V. 

  • Hi Kai, 

    Thank you for clarifying the common mode voltage in my application for the INA282 part.

    I was going through the process of calculating the total error for the part based on the formulas in the datasheet and noticed that the 'RVRR' spec gets lower as the INA28x part's gain increases.

    Based on the equation for the 'Vos_ref' error... it would seem to indicate that the error due to the offset voltage would decrease as higher gain parts are used?

    If this is true is it best to reduce 'Rs' and use a higher gain INA28x part in an application in order to reduce the error due to the offset voltage of the part?

    The task of selecting the 'Rs' and the INA28x part's gain seem to be driven also by making sure that the output voltage is within an acceptable range given the range of current to be measured.

    For my application the current through 'Rs' will nominally be 41mA but can rise to 191mA continuous and can even momentarily be as high as 318mA. My ADC can tolerate 0-2.5V input range.

    So one option would be to use the INA283 part for a gain of 200 and an 'Rs' of 39mOhms. This would keep the output voltage within 0-2.5V for the current range. 

    Alternatively I could use the INA282 part with a gain of 50 and an 'Rs' of 150mOhms for the approximately the same Vout range. But this option would seem to have a greater 'Vos_ref' error due to the higher 'RVRR' spec for the INA282 v.s. the INA282. So I think I should select the higher gain INA283 part with the 39mOhm 'Rs' resistance but would like to make sure I am thinking of all this correctly. 

    I appreciate you help and input. 

    cheers,

    -Sean 

  • Hi Carolus, 

    Thank you for clarifying how the common mode voltage is calculated. I agree with you that my application's common mode voltage would be ~15.7V. I will use this value for the total 'error' calculations I am attempting to do for this application. 

    cheers,

    -Sean

  • Sean,

    I'm in agreement. Keep in mind that the RVRR component of error is the error introduced from the reference of the device, and does not have bearing on the standard offset voltage. As you can see by the equation and examples in the DS, a reference of 2.048V reduces this component to zero, while a ground reference for both pins worst case creates an additional 92.2uV of offset voltage,, so this error source will dominate the further you move the reference from mid-supply. Are you planning to use the part unidirectionally, or bi-directionally?

    You are correct that your target should be to ensure that your dynamic range created by the shunt satisfies the input range of your ADC, and optimization of the offset voltage will help with accuracy at lower measurements, where the offset typically begins to dominate. Normally the trade-off here is BW, but for both the '282 and '283, the BW is 10kHz, so I would agree that the INA283 is probably the correct choice here, provided you can find a 39mOhm resistor. 

  • Hi Carolus, 

    I can see what you mean regarding the offset voltage error being dominated by the external reference voltage applied to the 'ref1' and 'ref2' pins. For my application I chose to tie both reference pins to ground...which as you mentioned is the worst case scenario for the offset error calculation. I will be using the INA28x current sensor in the uni-directional mode. The current flowing to the "Trillium_Power+" node is DC and will only flow in one direction. For uni-directional operation it seemed as though it was best to connect both ref pins to ground to make full use of the output voltage range. But as the data sheet points out... the offset voltage error (induced by the external reference) is at worst case in this configuration. My ADC input is an MSP430 which I believe is 0-2.5V. So I don't know if there was a better configuration for the INA28x that I could have used to minimize the offset error given my uni-directional mode of operation?

    You mentioned that the RVRR portion of the error does not have any bearing on the stand offset voltage. By "standard offset voltage" do you mean the offset voltage of the internal difference amplifier of the INA28x part which I believe is worst case 70uV and due to the part's internal 2.048V reference?

    Thank you for clarifying the best strategy for setting the 'Rs' value and INA28x gain for an application. I have been able to find a 39mOhm resistor so I am going to proceed with using that and the INA283 part for my application. 

    Thank you again for your help and time. 

    -Sean

  • Sean,

    If you are planning to use the device unidirectionally, what you have described is ideal to maximize the FSR of the output. I was asking purely to uderstand where in the error calculation you fell, but it sounds like you have a firm grasp here.

    Regarding the offset voltage, yes, the "standard" offset I was referring to was the worst case 70uV that will be present in all parts.

    Let me know how implementation goes, and if you have any other questions. Thanks for using E2E!