• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 540 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:

LMH6518: What's the cause of the offset drift in LMH6518

LMH6518: The potential root causes of the output offset drift issue in LG mode

Yoshikazu Kawasaki
Intellectual 861 points
Part Number: LMH6518

Hello,

My customer seems to have exactly the same issue as this.  They got complaint from their customer(end customer) that LMH6518 had a large offset voltage.  Then, they checked how it works.  If LMH6518 worked for 15 hours continuously, the output offset drifted a lot(~30mV).  Is it expected?  Would you please tell me the potential root causes of the issue?

https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/594736/lmh6518-what-s-the-cause-of-the-offset-drift-in-lmh6518/2296746?tisearch=e2e-sitesearch&keymatch=LMH6518#2296746

Best Regards,

Yoshikazu Kawasaki

  • 0
    TI__Mastermind 22864 points

    Hello Yoshikazu-san,

    When there is a large offset voltage drift over time, as in your case, the cause could be due to device heating.  What load is the LMH6518 driving in the circuit?  A light enough load could cause a device to heat up due to high output current; this would affect the offset voltage due to drifting over temperature.

    The maximum output offset is +/- 40mV, with a typical output offset voltage of +/- 15mV.  With a datasheet spec for output offset voltage drift of -7uV/°C in low gain (LG) mode, I am curious about potential device heating.

    I look forward to continuing our conversation.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 861 points

    Hello Alec-san,

    Thank you very much for your quick reply.

    I also think the root cause would be device heating, but it's too large comparing with the results of ~30mV drift since as you mentioned it is -7uV/℃ in LG mode.  The load shouldn't that big since it just connects to the ADC.

    The attached, you'll find how it was checked by my customer.  Would you please give me your comments if it's caused by the device heating?  If so, would you please tell me why it drifted ~30mV which isn't expected?

    LMH6518 output offset.docx

    Best Regards,

    Yoshikazu Kawasaki

  • 0 in reply to Yoshikazu Kawasaki
    TI__Mastermind 22864 points

    Hello Yoshikazu-san,

    Thank you for this information; I will investigate and use the document from your customer.  After a quick look at the ADC driven by the LMH6518, I recognize the light load is the necessary 100 Ohms differential to match the internal 100 Ohm termination.  I am going to check other parts of the design and setup, as well as review the LMH6518 device with my team.  I will update again with my comments once I explore the cause of this behavior.

    I may ask additional questions.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    TI__Mastermind 22864 points

    Hello Yoshikazu-san,

    I have conferred with my team regarding this concern.  Thank you for bringing this behavior to our attention.  After reviewing your circuit and device behavior we noticed a clarification was necessary with the LMH6518 device datasheet.  The absolute maximum input voltage of +/- 1V is specified for dynamic (AC) signals.  With your use case of a static (DC) signal, the actual absolute maximum input voltage differs from the dynamic case.

    When static signals are used at the input of the LMH6518, the absolute maximum input voltage should be kept below 400mV.  This absolute maximum input voltage range is now +/- 400mV for static (DC) signals.  Operating this device with a DC input signal while following the absolute maximum for AC inputs is the contributing factor to your large output offset voltage results.

    We are already working to add this clarification on input voltage limits to the datasheet.  While we are updating the device document, please utilize this response as guidance for your design.

    Best,

    Alec

  • 0 in reply to Alec Saebeler
    Intellectual 861 points

    Hello Alec-san,

    Thank you very much for your reply.

    I understand the root cause would be the wide input voltage as DC, but my customer has already released their products to the market with the specification of +/-1V(max) as input voltage as specified in the datasheet, so it has a big impact.  They're thinking to use LMH6518 on their another product, but it's difficult because it can't handle only by the diode clamp and the attenuation ratio will be big which will cause the noise performance worse.  They strongly asked to fix this issue by LMH6518 itself, not just by updating the datasheet.  What would you propose them?

    1. LMH6518 test screening to meet +/-1V.

    2. Proposing additional circuit to meet +/-1V.

    3. Proposing another P2P amplifier (hopefully drop in replacement is possible).

    4. Others

    Best Regards,

    Yoshikazu Kawasaki

  • 0 in reply to Yoshikazu Kawasaki
    Intellectual 861 points

    Hello Alec-san,

    I got a message from my customer.  I'd like to talk about this issue off line by connecting you directly, so I asked friendship request.  Please accept it.

    Best Regards,

    Yoshikazu Kawasaki

  • 0 in reply to Yoshikazu Kawasaki
    TI__Mastermind 22864 points

    Yoshikazu-san,

    I have sent you a message to allow us to connect offline.

    FORUM POST UPDATE:  A datasheet clarification for the LMH6518 is being completed; for steady-state input signals the absolute maximum differential input voltage is +/- 400mV.  Transient input signals retain the +/- 1V absolute maximum differential input voltage range.  Any ability of the device to operate within specification while receiving a steady-state input signal exceeding +/-400mV cannot be warranted or guaranteed.  The failure to observe these clarified limits may result in output voltage offset behavior in excess of LMH6518 specifications.

    Best,

    Alec