• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 48 subscribers
  • Views 587 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.

LMH5401: Input protection for fully-differential amplifier

Nick Alexeev
Intellectual 615 points
Part Number: LMH5401

Folks,

I’m concerned that in my design the inputs of the LMH5401 may exceed the maximum levels. My LMH5401 is powered with ±2.5V . The upstream amplifier stage which feeds the LMH5401 is powered with ±3.3V . I’m concerned that in my design the inputs of the LMH5401 may exceed the maximum levels.

The LMH5401 is in a differential input to differential output topology like fig. 57 in the datasheet. R_gain = 60 Ω . R_feedback = 500 Ω . Output common mode pin is connected to GND. If I apply the rules of fully-differential amplifiers, the input pins should be close to GND always under nominal conditions.

The inputs are 50 MHz sine wave with 800 mV amplitude, nominally.

There may be transient conditions where both inputs can become +3.3V at the same time, or -3.3V at the same time. If I understand it correctly, those would be the worst conditions.

What protection topology could be used?

Any suggestion, insight or reference is really appreciated!

Sincerely,
- Nick

  • 0
    TI__Expert 3320 points

    Hi Nick,

        Is the upstream amplifier in the previous stage rail to rail? Also what's the current capability of that upstream amplifier?
    Also I see one issue with LMH5401 operation, when input is 800mV Amplitude and you are configuring with gain = ~9, output is going to be railed. Have you noticed this? 

    Regards,

    Hruday

  • 0 in reply to Hruday Kumar Mudireddy
    Intellectual 615 points

    Hi Hruday,

    > Is the upstream amplifier in the previous stage rail to rail?

    The upstream amplifier isn’t rail-to-rail. Here’s a more accurate picture of the upstream stage [more accurate than what’s in my original question]. The upstream amplifier is powered with ±5V rails. It outputs can reach from -3.9V to +4.1V .

    > What's the current capability of that upstream amplifier

    The upstream amplifer can output ±64mA nominal, ±105mA short circuit.

    > I see one issue with LMH5401 operation, when input is 800mV Amplitude and you are configuring with gain = ~9, output is going to be railed. Have you noticed this?

    800mV will be the common mode under nominal operation. The differential mode will be on the order of 15mV under nominal operation. Based on the relatively small differential signal (and output common mode set to GND), I’m not expecting the output of the LMH5401 to rail under nominal operation.

    I’m concerned about protection against transient conditions, though.

    Best regards,
    - Nick

  • 0 in reply to Nick Alexeev
    TI__Expert 3320 points

    Hi Nick,
         Thanks for the clarification. I have simulated your circuit for 4V transient. I can see the Voltage on the input nodes of LMH5401(INP and INM) goes to 3.5, which is a problem as this would trigger the internal ESD cells and that may damage. According to datasheet abs max voltages are as follows:


    So in this case max voltage cannot exceed, 2.5V + 0.7V = 3.2V. But again if 4V is transient and for a little time < 1ns or so then the device can survive.
    I would recommend putting an external protection diode at the output of upstream amplifier to +/-2.5V for protection to be on safer side.
    LMH5401 - input.TSC

    Regards,
    Hruday

  • 0 in reply to Hruday Kumar Mudireddy
    Intellectual 615 points

    Hi Hruday,

    > So in this case max voltage cannot exceed, 2.5V + 0.7V = 3.2V. But again if 4V is transient and for a little time < 1ns or so then the device can survive.

    Transients can last for much-longer than one nanosecond in my instrument.  A full seconds is not unlikely.

    > I would recommend putting an external protection diode at the output of upstream amplifier to +/-2.5V for protection to be on safer side.

    I understand.  By any chance, do you have a go-to diode model which you could recommend?
    A clamping diode for a 30MHz signal should have fast recovery, and low capacitance, I imagine.
    Of course, I'll look for a diode myself.  But I wonder what super-pros recommend.

    > LMH5401 - input.TSC

    Is this the simulation file?  What can I open it with?

    Best regards,
    - Nick

  • +1 in reply to Nick Alexeev
    TI__Expert 3320 points

    Hello Nick,

         I do not have suitable diodes to suggest from TI. You are right, a fast recovery diode with a lower cap is needed, you can find them on Digikey/mouser.
    ex: DLLFSD01LPH4
    You can open the simulation file using TINA TI which can be downloaded using the below link
    https://www.ti.com/tool/TINA-TI.

    Regards,
    Hruday 

  • 0 in reply to Hruday Kumar Mudireddy
    Intellectual 615 points

    Hello Hruday,

    > I do not have suitable diodes to suggest from TI. You are right, a fast recovery diode with a lower cap is needed, you can find them on Digikey/mouser.
    ex: DLLFSD01LPH4

    Thank you!  I looked for the suitable diodes.  Noticed that Silicon PN have lower capacitance (down to 0.3pF), while Schottkys have faster recovery (down to sub-nanosecond) and lower turn-on voltage.  Tradeoff.

    Best regards,
    - Nick