• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 3 answers
  • Subscribers 49 subscribers
  • Views 362 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.

TLV7042: TLV7042 Field Failures: INA- to VEE Resistance Change

Brad Carson
Prodigy 40 points
Part Number: TLV7042
Other Parts Discussed in Thread: ESD441

I have a window comparator design using TLV7042DDFR experiencing significant field failures where the resistance between INA- and VEE changes significantly.  

On TLV7042 parts removed from returned PCBAs, I measure ~1.5k Ohms between INA- and VEE (both measurement polarities).

On unused, good TLV7042 parts, I measure ~360k Ohms from VEE to INA- and "OL" from INA- to VEE.

Does this change from at least 360k Ohms (good part) down to ~1.5k Ohms (damaged part) indicate evidence of any particular component stress condition (e.g. over-voltage condition, over-current condition, over-temperature condition)?

My implementation is essentially the same as documented in Figure 8-3. TLV704x-Based Window Comparator from the datasheet.

  • 0
    Prodigy 40 points

    Also of note: INB+ and INA- are connected together (per Figure 8-3 / window comparator) but I only measure a resistance change for INA- to VEE.  INA- is first to fail instead of INB+ if that provides any additional clue.

  • 0 in reply to Brad Carson
    TI__Genius 11701 points

    Hi Brad,

    Sorry to hear that you're running into issues. The inputs have internal protection diodes to VEE. Are the inputs going negative by any chance? Negative inputs can turn on these diodes and they can start to conduct and lead to damage. Due to the weekend, I will review further on Monday. Thanks for your patience. 

  • 0
    TI__Mastermind 24560 points

    Brad

    Thanks again for your post and sorry to hear of the problem. As Chi has noted, it does sound like some sort of electrical over stress may have occurred. Please provide details of what is connected to the inputs (range of expected voltage). Likewise, it possible please provide a schematic. We can tske our support offline if proprietary. Finally, do you have someone from TI that services your company. If so. you can get them involved as well. 
    Chuck

  • 0 in reply to Chuck Sins
    Prodigy 40 points


    TI_TLV7042_e2e.pdf

    Thank you for promptly looking into this.  Here is a schematic snapshot.  The input to INA-/INB+ is the middle node of a voltage divider between VCC (VCHGR nominal 4.34V) and a battery pack 10k Ohm B-3435k thermistor to GND.  The inputs to the TLV7042 shouldn't be able to swing beyond the VCC/VEE range.

  • 0 in reply to Brad Carson
    TI__Genius 11701 points

    Hi Brad,

    From a schematic standpoint, I do not see any glaring issues. As you said, the inputs shouldn't be beyond the VCC/VEE range. My only concern is that it looks like the signal coming from off the board into the battery connector (and therefore BAT_NTC) and it's possible the signal could be causing some sort of transient voltage spikes that may be damaging the lower ESD diodes on the inputs. Usually, we recommend adding an input resistance (so resistor to pin 2 and 5) to limit the current in case this ever occurs. 

    At this point, because of I have more follow up questions, I think it is best to take this offline by email/call so we can better assist you with your issue. I will be reaching out to you shortly offline. Please accept my friendship request. 

  • 0 in reply to Chi Nguyen1
    Prodigy 40 points

    I can look at adding input resistance but that would mean a board spin.  With consideration of the specific diode properties internal to the TLV7042, can you propose any TI component that could be placed in location C11 (from BAT_NTC to GND) to suppress transient voltage spikes quicker / at lower levels than the diodes internal to the TLV7042 and protect the internal diodes?  

  • 0 in reply to Brad Carson
    Guru 287370 points

    You can put an external diode in parallel with the internal one. This works best if there is a resistor between them so that most of the current flows through the external one. Without a resistor, you have to use a diode with a lower forward voltage, i.e., a Schottky diode (with higher leakage current).

    If you also need to protect against positive voltages, a clamping diode to VCC would be best, but you have no footprint for that. A TVS/ESD diode will work if the clamping voltage is low enough (the TLV704x's absolute maximum rating is 7 V). There are no diodes designed for 4.3 V; you have to use a 5.5 V device with a low clamping voltage like the ESD441. (But any such device is a plain silicon diode in the other direction, so it would not protect against negative spikes as well as a Schottky diode.)

  • 0 in reply to Clemens Ladisch
    TI__Mastermind 24560 points

    Excellent suggestions as usual Clemens.  Thanks for your support.

    Brad

    Thanks for posts as well and your support of TI.

    As Chi mentioned, there is nothing obvious from the schematic that we can highlight.  However, when we receive customer returns and we see signals coming from off a board or through a connector, there is always the concern that some voltage outside of the expected range could get coupled in and thus the suggestions that you are seeing from Chi and Clemens.  These are best practice suggestions since we really can't tell without seeing waveforms if there is concern for electrical overstress.  I believe Chi has reached out offline to offer further support if needed, so I will be closing this post with the thought that we will continue offline if needed.

    Chuck