• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 267 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.

OPA392: Input voltage exceed both power supply rails

Heiko Bruggner
Prodigy 10 points
Part Number: OPA392
Other Parts Discussed in Thread: LM6132, OPA928

Hi,

we are designing a new analog frontend and plan to use the OPA392 because of its very low input bias.

Unfortunately, the absolute max supply voltage for the OPA392 is +/-3V. In rare cases, our input signals can reach up to +/-7V.

Now we are curious if it's possible to limit the current to below 10mA using a series resistor and the OPA392 will still be fine?

Other devices like the LM6132 have similar characteristics regarding their Inputs pins, but the datasheet clearly states, that it's possible:

"The LM6132/34 can be driven by voltages that exceed both power supply rails, thus eliminating concerns over exceeding the common-mode voltage range"

We assume that the OPA392 is equipped with internal clamping diodes, so if the input voltage exceeds the supply voltage but limited to 10mA using a series resistor, there should be no damage?

Thanks,

Heiko

  • +1
    TI__Mastermind 22347 points

    Heiko,

    1. You are correct that limiting the input current to less than 10mA normally will protect the device from damage.  However, when the overstress signal is applied to the input the overstress is routed to the power supply via ESD diodes.  Normally, we recommend adding a TVS diode to the supply as an additional protection.  This is done because many LDO supply regulators cannot sink or respond quickly enough to the overstress signal routed from the input to the supply.  This may seem counter intuitive as the overstress signal is on the input, but I have seen many real world cases where this is required.  It also has the added benefit of protecting against supply transients.
      1. Choose the TVS standoff voltage to be equal to the nominal supply voltage.
      2. Ideally the breakdown will be less than the absolute maximum supply voltage (i.e. less than 6V).
    2. Yes.  The OPA392 has standard ESD structures (see figure 7-1).
    3. Generally, I recommend limiting the current to less than 10mA (10mA is the limit).  Ideally 1mA.
    4. This video covers this topic:  https://www.youtube.com/watch?v=Yz6vq-vBNnI 
    5. You may want to consider OPA928 with a wider supply.  This is a new device with extremely good IB specifications.  This device has a wider supply range so you could design the circuit so that it did not experience overstress.  If that is possible, I prefer that solution.

    I hope this helps!  Best regards, Art

  • 0 in reply to Art Kay
    Prodigy 10 points

    Hi Art,

    thanks for your great answer. This helps a lot. Also, the hint with the TVS diodes and your video is of great value.

    Regarding the OPA928 - this is a very interesting device. A view points make it difficult for us to come to a clear decision.

    - The Input offset voltage drift is only specified between -40°C and +125°C. For the OPA392 there is a spec between 0°C and +80°C. For sure that spec is much better because of the smaller temperature range, but we are only working in this small area and it would be very interesting how the OPA928 behaves in this area, to compare it against the OPA392.

    - The differential input impedance seems pretty low with 100MOhm compared to 10TOhm of the OPA392. Is this correct, as the common mode Impedance is in the same range and the differential input impedance is that much lower? We plan to use this device as a buffer into a low impedance load, so we would see the differential resistance at our signal input.

    - Is there a rough timeline available when the device will be in production? Would be great to see if it fits our schedule.

    Thanks a lot & Best Regards,

    Heiko

  • 0 in reply to Heiko Bruggner
    TI__Mastermind 22347 points

    Heiko,

    1. Regarding the offset drift:  I will mention your comment to the development team (i.e. that it would be good to specify drift over 0C to 85C).  Over the full range the two products are somewhat comparable(0.6uV/C vs 0.8uV/C).
    2. I will loop the product definition expert for this new device in to this E2E for feedback on some of your questions.  Regarding the differential input impedance this is largely due to AOL.  The common mode impedance is related to a change in bias current vs common mode input.  See https://e2e.ti.com/support/audio-group/audio/f/audio-forum/1075480/ina134-input-differential-impedance .  For parameters such as common mode and differential impedance the pre-release data sheet may be an estimate based on early design simulations, so I will defer to the product definer in case there is different information on this.
    3. I believe the release of this device is expected in the next few months.  The product definer can provide a better estimate.

    Best regards, Art

  • 0 in reply to Heiko Bruggner
    TI__Prodigy 720 points

    Hi Heiko,

    Thank you for your inquiry and feedback. Here are my comments:

    1) We made an initial decision to keep all dc precision specifications (with the exception of IB) to a wide temperature range -40C to 125C. With that said, I do think it will be worth showing a more limited temperature range. We will consider adding it to the final version of the document. I expect the drift to be ±0.1, typ and ±0.8, max for 0C to 85C.

    2) As Art mentioned, the preliminary data sheet contains only preliminary numbers. We will update with the final value in the production version of the data sheet. I expect the existing number to increase, however, the differential input impedance should not present an appreciable error source when compared to the common-mode resistance that is more directly related to IB. 

    3) The OPA928 should be released in the next couple of months. The current plan is March 2024.

    Please let me know if you have any further questions.

    Best,

    Daniel

  • 0 in reply to Daniel Terrazas
    Prodigy 10 points

    Hi Daniel,

    thanks for your valuable feedback. Please find my comments in blue.

    1) We made an initial decision to keep all dc precision specifications (with the exception of IB) to a wide temperature range -40C to 125C. With that said, I do think it will be worth showing a more limited temperature range. We will consider adding it to the final version of the document. I expect the drift to be ±0.1, typ and ±0.8, max for 0C to 85C.

    Perfect, I had the similar assumption, thanks for feedback.

    2) As Art mentioned, the preliminary data sheet contains only preliminary numbers. We will update with the final value in the production version of the data sheet. I expect the existing number to increase, however, the differential input impedance should not present an appreciable error source when compared to the common-mode resistance that is more directly related to IB. 

    The common-mode resistance is pretty important, as for my understanding it's in principle the resistance to the ground path (as shown in the very rough draft below). From our understanding, Zid is the resistance in between the positive and the negative terminal. We plan to use it as a buffer for a "low" impedance ADC, as shown in the graph below. Our understanding is, that the ( Zid + Zesr + Zadc ) is parallel to Zic , so that the current through ( Zid + Zesr + Zadc ) massively increases the current drawn from our input signal. In that case it would outnumber IB completely. Or do we have a misunderstanding of Zid?

    AT the moment, this is the only point which speaks against the OPA928. From all other points, it's much better than the OPA392.

    3) The OPA928 should be released in the next couple of months. The current plan is March 2024.

    That should definitely work with our schedule. We plan to manufacture first rough prototypes in Q2

    Thanks,

    Heiko

  • 0 in reply to Heiko Bruggner
    TI__Prodigy 720 points

    Hi Heiko,

    I hope you are well.

    1) Thank you, no problem.

    2) Your understanding is correct, however, the differential input impedance only sees the offset of the input transistors (see below), or about 25uV, whereas the common-mode input impedance sees the full signal range. This is largely why the input common-mode impedance usually yields the more significant error source. Assuming, of course, that you are operating within the linear region of the amplifier and have sufficient AOL (to Arts point). 

    With that said, I do agree that the differential input impedance on the data sheet is quite low for this part. This number will certainly be revised and much larger than the current spec of 100MΩ.

    3) Good to hear! We will keep you updated.

    Best,

    Daniel