• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 484 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.

OPA4991: Slew rate not matching datasheet spec

Abigail Leonardi
Expert 3390 points
Part Number: OPA4991
Other Parts Discussed in Thread: TLV9364, OPA4992

Tool/software:

Hi,

One of my customers recently reached out to me with an issue where the slew rate they are seeing from the device is off by an order of magnitude than what they expected from the datasheet. They were unable to duplicate the curves seen in the data sheet in figure 6-11.

I can ask for their schematic to review and as well as an image of their test setup when they tried to replicate the curves, but is there anything else that would be helpful to check in resolving this? 

Or are there any known dependencies which may influence the slew rate performance of the device? 

Thank you for your help here. 

  • 0
    TI__Guru**** 164706 points

    Hello Abigail,

    This 6-11 chart? Slew rate doesn't apply as this is small signal results, In theory, zero signal.

    Slew rate in data sheet is large signal unity gain. 10V step in this typical chart.

    Slew rate for small signal can be different, see application note for slew rate. [APP] Ramping Up on Slew Rate

    For better small signal slew rate try, TLV9364 

  • 0 in reply to Ron Michallick
    TI__Expert 3390 points

    Hi Ron,

    I relayed your information back. The customer has let me know their devices are not meeting slew rate nor the GBW spec. They were wondering if TI could share how these results were obtained so they may duplicate it?

    Thanks!

  • 0 in reply to Abigail Leonardi
    TI__Guru**** 164706 points

    Abigal,

    GBW and SR are just typical specs with no limits, so there isn't actually a specific data sheet claim at all. 

    For 6-11, I suggest setting up a gain between a hundred and a thousand, input frequency to get to a roll off area. Be sure to use a very small input, use a divider to help. 

    For 6-35,36, setup unity gain a provide a 10V step input.

  • 0 in reply to Ron Michallick
    TI__Expert 3390 points

    They detailed their setup to me as following, does the output seem unexpected given the setup? Also is there a reason for slew not being applicable for small input signals? 

    For 6-11 is there a reason the small input in necessary as well?

    Customer details: 

    In our circuit, we are using a standard non-inverting configuration. The feedback resistor is 19k, and the resistor from - to ground is 1k. A simple gain of 20 V/V. This is loaded with a 10k resistor at the output to ground. Our voltage rails are +/- 12V (24 V span total). We can't even get 20 kHz of bandwidth out of this circuit. In fact the signal rolls off (-3 dB re 1 kHz) at around 10 kHz. This is with 1Vp-p fed into the + input of the op-amp. This circuit should be nowhere near running out of GBW (4.5MHz), or slew (21V/us).

     

  • 0 in reply to Abigail Leonardi
    TI__Guru**** 164706 points

    Abigail,

    For a full gain of 20, the output needs have enough slew rate to keep up with the input signal. Only 1/20th of the output makes it back to IN- pin to match the IN+ pin.

    10kHz at 1Vpp (at input) requires 30mV/us at IN-, so 0.6V/us at output (20X). If IN- is keeping up with IN+ well, then the difference is small at between the inputs. With a small difference, 0.6V/us is all you get. So everything is normal here. 

    If input difference becomes larger then it can slew faster, however this increased input difference voltage comes at a cost of reduced output.  Less than expected forward gain of 20.

    GBW is all about a virtually zero amplitude signal gain. SR is all about a large step input response. This real world application has neither of these. 

    Have they tried TLV9364?

  • 0 in reply to Ron Michallick
    TI__Expert 3390 points

    Thanks for the quick reply Ron, running this by them

  • 0 in reply to Abigail Leonardi
    TI__Expert 3390 points

    Is the TLV9364 a drop in alternative they would be able to use with better slew rate? They already have populated boards and will need to figure out if there is a solution which can be directly swapped out

  • +1 in reply to Abigail Leonardi
    TI__Guru**** 164706 points

    Abigail,

    Yes drop in. Only notable difference is TLV9364 is not rated for input rail to rail. I don't believe rail to rail input is required here. OPA4992 is an rail to rail input upgrade to TLV9364.

  • 0 in reply to Ron Michallick
    TI__Expert 3390 points

    Thank you for all the help Rob, really appreciate it!