• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 48 subscribers
  • Views 669 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

OPA810: bias current

OPA810: Abnormal latching input bias current on powerup

Alex Pogossov
Prodigy 150 points
Part Number: OPA810


Tool/software:

Hello,

I encountered an abnormal (unexpected) behaviour of OPA810 op-amp in a voltage regulator circuit (attached).

This is a quite common single-supply architecture, where the zener ZD1 is powered from the regulated output via R4. The op-amp, by virtue of R1R5 feedback, scales the zener voltage (nominally 6.8V) to the required output (nominally 10.8V).

For a proper startup, a pull-up resistor R2 is essential. On powerup, due to R2 and finite output resistance of the op-amp some voltage would develop on the output. This voltage, being amplified by the rail-to-rail op-amp, will cause the output to rise, consequently rising the input on non-inverting input. This process ramps up until the zener starts clipping.

However, with OPA810, the circuit fails to start up. Somehow, probably due to some parasitic latch-up type process, particularly possibly when the output is pulled up, some current of about 1uA got injected in both inverting and non-inverting inputs of the op-amp, and some clamping structure between the inputs is possibly activated.

As a result, as the equivalent resistance in the feedback (about 200K) is larger than in the zener filter (100K), parasitic voltage on the inverting input (about 130mV, shown in red) exceeds voltage on the non-inverting input (about 90mV). As a result the output of the op-amp is permanently held low, and the op-amp would not unlatch, even when the supply Vs is at nominal 12V.

Further tests (not discussed here) showed that the op-amp unlatches once non-inverting input voltage exceeds about 130mV.

Thus I found a workaround -- reduce impedance in the feedback below the impedance of the zener filter. For example with R1=27K and R5=47K the circuit does start up.

But the whole situation is worrying for those who do not know. Perhaps you can mention of it in the datasheet or fix the bug in future.

Regards, Alex

  • 0
    Prodigy 150 points

    Would be interesting, how TI engineers handle that... Will they admit a bug in OPA810 and provide some explanation and notice to the datasheet, or just deny/ignore?

  • 0 in reply to Alex Pogossov
    TI__Intellectual 2065 points

    Hey Alex,

    Thank you for your question. We'll test the circuit in the lab using the EVM and get back to you.

    Regards,
    Aditya Gosavi

  • 0 in reply to Aditya Gosavi
    TI__Intellectual 2065 points

    Hi Alex,

    I tested the circuit in the lab and it was starting up properly, with Vout = 10.7V, even without a pull-up resistor. In your case what is happening is that the op-amp output is trying to go to negative supply (gnd), but due to the output headroom, its getting stuck at 60mV and not 0V. 

    Also as you suggested, changing the R1 and R5 to 27k and 47k is a possible solution, which will keep the Vin+ voltage greated than Vin- (Voltage due to bias current will be more on Vin+ pin due higher impedance), and this will drive the op-amp output to 10.8V. You can also try adding a pull-up resistor directly to zener cathode, rather than the op-amp output. Because in your case when you add a pull-up resistor to op-amp output, it has a low impedance path to the op-amp output, and the zener might not be able to ramp up to 6.8V.

    Regards,
    Aditya Gosavi

  • 0 in reply to Aditya Gosavi
    Prodigy 150 points

    Hello Aditya,

    Your answer discussed possible workarounds. (By the way, pull-up to Zener is not a good idea as it will degrade line regulation.)

    Your reply does not answer the main question -- why on earth the op-amp with SUPPOSEDLY 2pA input current develops about 300nA of input current on both inputs under certain conditions.

    Such behaviour definitely is NOT what a user would expect. In my opinion, it can be considered a "bug" in the OPA810 design. 

    At least it should be addressed in the datasheet. For example, it might contain a warning something like "during output reaching negative rail, input current may increase to 300nA" or "input current may increase up to 300nA on power-up" or whatever the best reflects the actual mechanism of this phenomenon. Obviously TI lab and TI engineers are in position to work out exactly under which condition input current may increase and openly write about it.

    Regards, Alex