• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 488 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

OPA388: Rail-to-rail amp output gets stuck at upper rail.

Glenn Zelniker
Prodigy 20 points
Part Number: OPA388
Other Parts Discussed in Thread: TINA-TI

I'm using the OPA388 in a current-measurement design. I use the '388 to amplify (Gain = 10,000) the voltage across a shunt resistor, with the expectation that 200uA into the shunt generates 2V at the output of the '388. It works as expected with one big problem: if the shunt current is a few tens of mA, it drives the '388 to the output rail (4V) so hard that when the shunt current drops to zero, the output of the amp won't "let go" of the rail, staying high until the output is temporarily connected to GND (with, say, the tips of a current meter). I've tried connected the output to GND with a 10K and with a 1K resistor but this doesn't solve the problem. 

Would a smaller resistor to GND help? I haven't seen an amp behave like this in the past. 

  • 0
    TI__Guru 63990 points

    OPA388 includes internal back-to-back parasitic diodes that get forward biased when the inputs are pulled apart as is the case when the output is pushed into its rail - see below.  This will create offset voltage shift that in high gain like 10,000 will keep the output stuck at the rail even after the shunt current is turn off - the op amp is effectively in the lock-up/latch-up state, which may be removed by cycling the supply or shorting the output to ground.

    You do not show the values of the resistors use but if you use a low value R2 like 10ohm, this will allow lock-up to occur.

    I would advice you to lower the gain to 100 by increasing the R2 and increasing Rshunt by 100x that would keep the input and output ranges unchanged - see below.  I believe this would allow output to recover om its own from over-driven output condition.

    If you continue having problem, please provide details what kind of power supply you use (LDO?) and modifying the attached schematic to match your application.

    OPA388 shunt application.TSC

  • 0 in reply to Marek Lis
    Prodigy 20 points

    Thanks for the suggestions. I've attached a picture of my circuit. The power supply is the raw output of a LiPo battery. I'm not sure how much I can increase the shunt resistance but I'll try simulating. By the way, which simulation packages is associated with the .TSC file above? Is this a TI-flavored SPICE?

  • 0 in reply to Glenn Zelniker
    TI__Guru 63990 points

    Yes, .TSC file is Tina-TI.  For free download, please use following link: https://www.ti.com/tool/TINA-TI

    Having said that, with 2ohm shunt in gain of 10,000, 200uA current would result in the output collapsing on its positive rail - see below.

    OPA388 shunt application 2.TSC

  • 0 in reply to Marek Lis
    Prodigy 20 points

    Thank you so much for this insight. I scaled the shunt up by 100 (to 200ohms) and R2 and R3 to 4K7. I'll do some de-soldering and re-soldering and we'll see how it does. Sadly, I had this thing built with 0402 resistors, but I've got re-work tools.

    By the way, the application is incredibly high dynamic range. It's for production-testing a device that draws a lot of current and then goes into a low-power sleep mode. It's got a cellular (LTE) radio that draws lots of current at power-up (more than 100mA for a few seconds) and then the whole device drops to a few mA and then finally a few tens of uA after several seconds. This application tests to make sure the device goes into sleep mode. Do you think the 100mA draw would be enough to collapse the amp onto its positive rail? The shunt resistor is now 200 ohms and the amp's gain is 100.

  • 0 in reply to Glenn Zelniker
    TI__Guru 63990 points

    With 100mA load the drop across 200ohm shunt would be 20V, which could result in 3.155mA input current as shown below (this is within max rated limit of 10mA).  However, as a practical matter I do not believe this could occur in your circuit and instead the current would be limited to around 20mA where Vcm=~0V.

  • 0 in reply to Marek Lis
    TI__Guru 63990 points

    It's been a week since last reply, thus I'll close this thread. If you need further assistance, please reply or start a new inquiry.