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.

TINA/Spice/OPA547: OPA547 Shutdown Question (and simulation)

Part Number: OPA547
Other Parts Discussed in Thread: TINA-TI, OPA548, , OPA140

Tool/software: TINA-TI or Spice Models

Hello!  I'm trying to translate a 5V DAC voltage to a -5 to 45V signal.  The datasheet shows that this part's output can be disabled when placed in shutdown mode (either thermal or using the E/S pin).  When I simulate this, it appears that the part is actually outputting the negative rail (in my case, -12V, which is an issue) rather than floating.  Is this just an issue with simulation not knowing how to simulate a floating/tristated output or does the part actually output the negative rail when in shutdown mode?

Below is the circuit I'm using:

Let me know if any additional information is needed.  Thanks!

  • Hi Jeremy,

    The OPA547/OPA548 output will pull current from a ground referenced load when the device is placed in shutdown mode. The higher the resistance the load presents, the closer to V- the output pin will pull down due to the increased voltage drop. If the load resistance is low, a few ohms or tens of ohms, the output will be closer to 0 V when the amplifier is disabled compared to when the load is a higher resistance. This occurs because of the output stage design.

    The images below show the OPA548 on the bench operating and enabled, then disabled with a 1 kilohm load and lastly with a 100 ohm load. Observe how the output level moves from a negative voltage, to something closer to 0 V as the load is lowered from 1 kilohm to 100 ohms.

    I don't know how closely the OPA547/OPA548 model mimics the shutdown output behavior of the actual device, but the fact that it exhibits the behavior is a useful indicator.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • Jeremy,

    Thanks for the schematic!

    Regards, Thomas
    Precision Amplifiers Applications Engineering
  • Hmm, unfortunately, that's going to be a problem for us.  I'm looking to set V- to GND instead.  I'm poking around at the current limit using an external DAC but the math in the simulation doesn't appear to be working the way I'd expect (really hoping this isn't related to the load as well).  

    I calculated the voltage on the pin using a DAC and it appears that a 50 mA limitation should be about 4.434V (Vlimit = 0 + 4.75 - (31600 * 0.050 /5000).  I tried simulating it but the full voltage appears to be coming through even with 60 mA of current being pulled.  

    Is this output dependent on the load as well or is there another reason the calculations aren't working out as expected?

    Schematic:

    7455.DAC Translator - TI Help.TSC

  • Hi Jeremy,

    If you change V- from -5 V to 0 V, you have to make sure you observe the OPA547/OPA548 linear input common-mode voltage range (Vcm), and output voltage swing range on the low end. If violated, the circuit won't perform as needed.

    The current limit accuracy is specified under only condition in the OPA547/OPA548 Electrical Characteristics table. The max error under that condition is about +/-10 %, but under other conditions it may be higher. I tested the OPA548 model with the current limit set to various from about 30 mA to 100 mA and found that the current where the current limit kicked in was off by 10 mA, and sometimes more. The current limit function is not intended to be a precise current limit, but more so a general limit used to protect a load or the op amp.

    It is possible for the OPA547/OPA548 output voltage to drop below 0 V in a simulation because of the ideal current source connected to the output. You can see this in my version of TINA simulation circuit shown below. In a real circuit that wouldn't be the case unless the current source could support the reversal.

    One issue I noted with your TINA circuit is the very high resistance values selected for Rg and Rf. The OPA547/OPA548 input bias current is on the order of 100 to 500 nA. That can contribute hundreds of millivolts to the output referred voltage offset. That error can be significant and cause the output voltage to be different than the expected value. I lowered the resistors by a factor of 10x in my TINA simulation and error is much reduced.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

    OPA548_DAC_drive_01.TSC

  • Hi Jeremy,

    Did the information we provided resolve your OPA547/OPA548 application questions? If so, please close this E2E thread.

    Regards, Thomas
    Precision Amplifiers Applications Engineering
  • Thomas,

    Thanks for the reply!

    It looks like the reduction in components didn't actually make any difference.  I think you might be comparing the new value (-0.8V) to the output of the initial circuit that has V- connected to -12V rather than GND (-12V or so).  Let me know if I'm wrong about that or if you were able to see a difference by scaling the components by a factor of 10 (maybe I missed something). 

    I did swap out the ideal op amp with the OPA140 and I didn't see any difference.  I'm not entirely sure where to go from here apart from just sampling some parts and testing the circuit in house (either the simulator or my setup is incorrect or the part will not be able to meet our needs).  I can change the Vsupply to GND rather than -12V but it's going to eliminate some of the functionality of our box (we need the -5V supply). 

    Have you been able to simulate any possible outcome where thermal or manual shutdown doesn't result in a negative output (or know what real world testing we should be doing in the lab to test this) or is that something we can expect?  I can switch in an external -5V but that would require an additional MUX or optoisolater, which just adds (potentially) unnecessary complexity. 

    Thanks, Thomas!

  • Hi Jeremy,

    It looks like the reduction in components didn't actually make any difference. 

    Okay, then the input bias current (Ib) of the OPA547 you are testing must be towards the low end. I mentioned it because it might be an issue with a different production run of OPA547 that just happen to have higher Ib.

    I think you might be comparing the new value (-0.8V) to the output of the initial circuit that has V- connected to -12V rather than GND (-12V or so).  Let me know if I'm wrong about that or if you were able to see a difference by scaling the components by a factor of 10 (maybe I missed something). 

    That -0.8 V would only come about in simulation because the ideal current source can accommodate the polarity switch. The actual circuit with a real load wouldn't be able to support a negative output voltage when a single, positive supply is being applied. Simulation shows that with a resistive load (1 k) connected from the output to ground the minimum output voltage is about +0.75 V and +45 V, for a +48/0V supply arrangement. The output doesn't go negative as expected.

    I did swap out the ideal op amp with the OPA140 and I didn't see any difference.  I'm not entirely sure where to go from here apart from just sampling some parts and testing the circuit in house (either the simulator or my setup is incorrect or the part will not be able to meet our needs).  I can change the Vsupply to GND rather than -12V but it's going to eliminate some of the functionality of our box (we need the -5V supply). 

    Actually, having the a -5 V, or -12 V, negative supply to power V- really makes applying an op amp such as the OPA547 easier. Then the output can truly swing down to 0 V and even somewhat negative, depending on the negative voltag elevel used for V-.

    Have you been able to simulate any possible outcome where thermal or manual shutdown doesn't result in a negative output (or know what real world testing we should be doing in the lab to test this) or is that something we can expect?  I can switch in an external -5V but that would require an additional MUX or optoisolater, which just adds (potentially) unnecessary complexity.

    Unfortunately, there isn't any way to keep the OPA547 from going to a negative voltage when the device goes into shutdown and a negative supply is being used. However, if the load resistance is low that will keep the voltage close to 0 V.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

     

  • Jeremy

    We haven't heard back from you so we assume this resolved your issue. If not, post a reply below, or create a new thread if this one has timed-out.

    Thanks
    Dennis
  • Thomas/Dennis,

    Honestly, it sounds like you answered our question but it doesn't necessarily mean it solves the problem for us.  I don't have any control over the load on the output as this would be going on a production line with a bunch of different parts and the loads are likely to vary from part to part.  If putting the part in shutdown (either thermal or disabling the supply) causes the part to go to a negative voltage, this part can't be a solution for us.  We can't have negative voltages going on parts that may or may not have reverse voltage protection. 

    Sounds like we might need to find a different solution as it doesn't sound like this part will work for our application. 

  • Hi Jeremy,

    I checked our other high-power op amps with regard to your application and disabled state requirements, and don't find one that can be assured not to have some voltage level present at the output when in the disabled state.

    Regards, Thomas
    Precision Amplifiers Applications Engineering