• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 18 subscribers
  • Views 875 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.

SN74HC244: Output drive current appears divergent from mfg code 53CXGFK to samples of 6CCTY1K and 76AR5JK - did a process change occur after 2015?

Mark Pallant
Prodigy 20 points
Part Number: SN74HC244

We have a systemic fault in our mature product function that appears to be the result of unequal current supply drive between VCC and GND.  The TI SN74HC244DW IC is used to interface an a keyboard matrix - columns signaled and rows read. The issue is where the pulses are combined through multiple key presses (simultaneously) to obtain 2nd and 3rd functions - the logic high on the input isn't high enough to switch the 244 to a logic high output whereas on the older devices, this logic level was always over 2.3V.  The input level is a sum of the combined outputs of the same IC (column drive signals).  The fault popped up mid production run and was identified by having a mix of two date codes - the good ones all being older than 2016 ('bad' device marking 6CCTY1K). Last known good ID is 53CXGFK. 

Initial thought was a bad batch - changed out 16 units to MFG code 76AR5JK, but same result.  Is there any known change between 2015 and 2016 with either wafer fabrication or package assembly?  XRAY didn't reveal anything obvious, but there's plenty of characteristics that aren't visible (like bond wire material).

If there was a change, we'll probably have to have another look at the application with our customer, but again, this is a mature product that we've been manufacturing for over a decade. 

Any insights would be appreciated.

Thanks!

Mark Pallant

  • 0
    TI__Mastermind 44570 points
    Hi Mark,

    I can definitely look into this. First, could I get more info on the application of this device ( a schematic would be helpful as well).
  • 0 in reply to Dylan Hubbard
    Prodigy 20 points

    Hi Dylan,

    I've attached the relevant portion of the schematic (cut'n paste). There's IT and possibly ITAR considerations that prevent my providing the larger picture.  The left side nets interface with the microprocessor and the right side with the keyboard matrix (3x5).  The loss of function is where COL1 and COL3 are summed on ROW2 and ROW3.  The resultant pulse sums are too near switching levels of the inputs to provide the expected output logic level at each 'bit' timing.  A good 244's summed pulses at the input pins have a much better resolution (amplitude change) for the combined output pulses and aren't in the unstable / don't care levels of the inputs. The data is essentially lost in this condition making the specific key function fail.   System logic levels are 5V and 0V. 

    Regards,

    Mark Pallant

  • 0 in reply to Mark Pallant
    Prodigy 20 points

    Mark Pallant said:

    Hi Dylan,

    I've attached the relevant portion of the schematic (cut'n paste). There's IT and possibly ITAR considerations that prevent my providing the larger picture.  The left side nets interface with the microprocessor and the right side with the keyboard matrix (3x5).  The loss of function is where COL1 and COL3 are summed on ROW2 and ROW3.  The resultant pulse sums are too near switching levels of the inputs to provide the expected output logic level at each 'bit' timing.  A good 244's summed pulses at the input pins have a much better resolution (amplitude change) for the combined output pulses and aren't in the unstable / don't care levels of the inputs. The data is essentially lost in this condition making the specific key function fail.   System logic levels are 5V and 0V. 

    Regards,

    Mark Pallant

  • 0 in reply to Mark Pallant
    Prodigy 20 points
    Ok, I'm having trouble actually sticking the file in.. I'll try again in a bit
  • 0 in reply to Mark Pallant
    TI__Mastermind 44570 points
    Thanks for this. Let me get this information to the right people and see if I can track down any changes in the fabrication or assembly process.
  • 0 in reply to Dylan Hubbard
    TI__Mastermind 44570 points
    Hi Mark,

    After doing some digging there were no changes to the device that would change any electrical specs. I didn't see anything wrong with your schematic, however the device is not operating within datasheet spec at 2.3 V input. The user takes a risk when operating below the VIH spec.
  • 0 in reply to Dylan Hubbard
    Prodigy 20 points

    Hi Dylan,

    Thank you for digging into that!  I was grasping at any possible event correlation and was excited to see the white paper's revision date of mid 2016 which was after our last 'good' lot (by that I mean they worked a certain way that pleased the customer) and before the subsequent purchased lots that weren't.

    I've had things like this happen before where a specification not listed happens to cause the design to function a certain way and becomes dependent upon the unpromised attribute, and then when it doesn't, everyone gets whiny.  Last time it was a die size change - changed the input capacitance which affected the entire loopback response - simply a bad design that had to be revised.  In this case, we merely convinced the customer to drop a certain key-combination function (not relevant to the end user's operation).

    So, problem solved.  Thank you!

    Best Regards,

    Mark Pallant