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.

LPV821: Dual supply operation

Part Number: LPV821
Other Parts Discussed in Thread: LM7705, BQ76920


I need a sanity check regarding the LPV821's power supply capabilities.  Because my client's application requires accurately measuring +/- 400A, the sense resistor is 100uOhm to reduce power/heating (inside a sealed enclosure).  The 76920's current sense input span is +/- 200mV, and so to use its full range, I've added a x5 gain stage using 2 821's. Doing so requires the "CHG" amplifier to drive SRN 200mV below ground.

My TI FAE (Antonio Fadhel) suggested the LM7705 precision bias generator, which outputs -0.23V. My intent is to bias the LPV821's VSS to -230mV which should (?) allow it to drive SRN to -200mV.

However, I need a sanity check as to the 821's ability to do so, and this dual supply configuration. I'd appreciate any suggestions/comments/corrections


AFE and Current Sense.pdf


  • Hello Jeffery,

    The LPV821 certainly can be operated from dual supplies providing the linear common-mode and output voltage ranges are observed relative to the V+/V- levels being used. The common-mode voltage range is from V- to V+ so that is easy to understand and observe. The linear output voltage range is best bounded by the open-loop gain, Aol, which is  (V–) + 0.1 V ≤ VO ≤ (V+) – 0.1 V, RL = 100 kΩ to VS / 2. Therefore, it is best to plan for the LPV821 output not to swing any closer than 100 mV from V+ or V-. Using the LM7705 -0.23 V output for V- will certainly assure that is the case.

    I do see from your schematic that you are using the -0.23 V as V- for U11, but not U13. If there is a possibility that U13's output will try to move closer than 100 mV above ground, then it would be best to use the -0.23 V for its V- as well.

    One item I do notice about the design is the inclusion of the 800 kilohm R20 and R21 resistors in series with each LPV821 non-inverting input. Since the input current direction can be plus or minus there is no assurance that will improve the offset-referred voltage offset component due to input bias current. Additionally, the 800 kilohm resistors add significant thermal noise at the op amps inputs. Therefore, I would exclude these resistors from the circuit.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Thanks Thomas;

    Unfortunately, I need to drive the current sense input to +/-200mV. I thought that since these were RRIO, I could do so? If I understand you correctly, you're saying I can/should not?  If that's the case, I'll need a suitable replacement. Can you suggest an alternative Amp? I'd really appreciate it. I'll also check with the FAE.

    I'll delete the 800K R's. My intent was to match the source "impedance" of both inputs, but since Ib is so low, that's irrelevant. 

  • Oh, and forgot to mention that this is a near DC application. A golf-cart BMS which monitors the motor's current via a shunt R.  I've attached an updated schematic just in case.

    THanks4073.AFE and Current Sense.pdf

  • Hi Jeffery,

    You should be able to drive the LPV821 circuit with a +/-200 mV differential input voltage providing the linear input common-mode and output voltage ranges are maintained.

    Can you provide information about the current source and load and their connections to the shunt resistor? I need to understand which LPV821 sees the source and which one sees the load.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Thomas;

    The current source is a 4S25P LiFePO battery, which is used as a "smart battery" in a Golf cart application. It supplies power to the main controller, which I believe is an inverter, which drives the golf cart's AC motor. The current shunt resistor is on the low side, which is why I selected the 821, with the help of TI's FAE.  The nominal discharge current is 100A, with the maximum measurable current of 400A. Therefore, the shunt R would see 10mV nominally, and a maximum of 40mV. The load is the negative input (SRN) of the bq76920 AFE's, sense resistor measurement block. This is a 16-bit, integrating ADC, but the datasheet does not list its input characteristics. Its input range is listed as +/- 200mV, which is why I set the amplifier gain at 5.

    My confusion stems from my failure to understand your reference to "linear input common mode and output voltage ranges". As I mentioned, since this is a RRIO amplifier, I had "assumed" that I could drive its +input to -40mV as long as its VSS supply was at least that low (and is actually -230mV)....And that it's output could be driven to -200mV, or within ~30mV above Vss.

    My "guess" is that the "linear input common mode range" would refer to a AC signal (?), but since this is a near DC signal, I don't understand how that applies?

    If I've missed something, please don't hesitate to correct me

    Thanks for your help!

  • Hi Jefferey,

    I am certainly not intending to make this more difficult to understand. I am just trying to make sure that the input/output conditions presented to LPV821 Op amps assures they operate correctly in the application.

    The LPV821 has a common-mode voltage VCM range of (V–) to (V+), and a linear output voltage range of (V–) + 0.1 V ≤ VO ≤ (V+) – 0.1 V, for RL = 100 kΩ to VS / 2. As long as operation is maintained within these ranges your application of the LPV821 should be fine.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Hi Tom,

    Looking at Figure 30 of the datasheet, a 30mV headroom on V- seems sufficient for a current of up to ~1mA (I assume this is just to make sure the IR drop on the internal FET is low enough). They are driving a high-impedance sense pin so they should be able to meet this. Would the issue then be the non-linearity? If so, what would be the impact of operating in this non-linear region for a DC application? We are just trying to evaluate if it is something that can be acceptable for their application.



  • Hello Antonio,

    Yes, the ability of the LPV821 output stage to swing close to the supply rails is a function of the ON resistance of the output FETs. Typical Characteristics graphs 29 through 32 in the datasheet show a resistive characteristic when sinking/sourcing current that extends over several decades. A point to keep in mind is that the output is being driven into the rail in these cases and therefore the linear output range is being exceeded. Once that occurs the normal Op amp behaviors and performances degrade severely. 

    The way to be sure the Op amp is operating in its linear output operating region is to apply the Open-loop gain (Aol) parameter specified output swing as a useable maximum. The LPV821 Electrical Characteristics table specified for Aol, (V–) + 0.1 V ≤ VO ≤ (V+) – 0.1 V, RL = 100 kΩ to VS / 2. So for this case an output swing of 100 mV from each supply rail should be safe for this condition.

    Regards, Thomas

    Precision Amplifiers Applications Engineering