• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 19 subscribers
  • Views 1272 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.

SN74LVC1G34: SN74LVC1G34DBVR low R between 5V & GND pins

Joe Saggio
Prodigy 40 points
Part Number: SN74LVC1G34

Thank you for reading my issue!

I have attached the schematic showing how the SN74LVC1G34DBVR is designed into the product.

I've had (2) of these chips exhibit very low R between 5V and GND pins on failed PCBs. One showed 9.48 Ohms and the other roughly 14 Ohms.

On the PCB, in circuit, right at power up these 2 devices pulled the 5V supply down to "GND" and caused a blow of the 5V regulators on those 2 PCBs. Note: The PCBs with blown regulator and "errant" U4 had been powered up 2 times at the factory - once to load firmware and once to verify response after assembly into enclosure - with no PCB damage DETECTED. The failures occurred at power-up at customer site....since the PCB is in an enclosure when it arrives at the customer, I can't imagine ESD would be an issue at the customer not a big deal, just need to understand how this could happen.

Our PCB is in a conductive enclosure and the PCB GND is attached to the enclosure through PCB mounting holes. PCB is screwed into enclosure. These devices are installed into large water pipes where the PCB is external to the pipe, but electrically attached through the enclosure to the pipe...the pipes may be "grounded". THe PCB is powered from whatever 24V supply the customer has chosen..could be up to 250W at 24V.  

The SN74LVC1G34DBVR chips do not look damaged and they do have a 100 Ohm protection resistor on their output - I have verified those resistors are there and are the proper value. I have also measured and verified the health of TVS1. 

I have several PCBs here where U4 is OK and not pulling 5V supply down...out of 100 PCBs MFG, thus far I have found 2 that are LOW R and 1 that is 1.4M Ohm...what should I measure between pins 5 & 3?

Finally, and Interestingly, we have never seen this anywhere except at 2 customer sites...

  • 0
    TI__Guru* 97046 points

    Hi Joe,

    Resistance isn't really a measurement that makes sense for a supply pin.  Don't get me wrong - it certainly shouldn't be showing 10 ohms, but when you ask "what resistance should it show?" it doesn't really make sense.

    The best method to test the Vcc pin is to follow the conditions in the datasheet Icc spec:

    Apply 5.5V between Vcc and GND, short the input (pin 2) to GND, and measure the current into the Vcc pin. The value should remain below 10uA (typically less than 1uA).

    Obviously the 10 ohm measurement is showing a problem -- if you hooked 5V up to that you'd likely see 500mA and shortly thereafter an exciting smoke show (maybe with some fire, too).

    I have seen similar damgae before only in two cases:

    (1) ESD strikes -- it sounds like you've ruled this out as a likely cause, however it's possible that the device received an ESD hit during manufacturing and was "walking wounded" until it went into normal operation - and failed entirely.

    (2) Vcc/GND were reversed.  This usually is caused by a shared ground connection between a high power inductive load and the logic device.  Sudden changes to the inductor current can cause huge swings in ground voltage and subsequent damage to the device.  There are other ways this can happen, too (for example, switching the Vcc and GND wires on a module), but the inductive load issue is the one I've seen to be the hardest to track down.

    If you want to get TI to really dig into this, you can return the devices through the FA process at your distributor. It will take some time to go through the process, but we will eventually get the parts and test/deprocess them to see what really happened.

    To be clear up front, this will only say what happened to our device and can't explain everything -- often an FA returns the result of "Electrical Overstress" and we're back to trying to figure out what went wrong in the system or manufacturing process to damage the device.

  • 0 in reply to Emrys Maier
    Prodigy 40 points

    Hi Emrys,

    Thank you for the reply - much appreciated. Agreed on current draw through supply pin to gnd pin. 

    My understanding on the 2 or 3 parts that exhibit this is that their subsequent outputs (the TTL Pulse out in sch) are left unused/open at the customer site. Even if the Pulse output was tied to ground, the on-board 100 Ohm would limit current to 50mA at 5V and should not cause the part to fail like this - I'm just trying to rule out other potential issues we did not touch on. 

    To your point regarding an inductive load sharing the 5V rail/GND; the only other devices on the 5V rail are a TI RS485 Transceiver chip and a 3.3V regulator. I suspect we can't classify either of those as a high power Inductive load...but I understand what you are referring to. I have watched the 5V rail power up many times on a scope and not seen anything suspicious there relative to a gnd/5V swap. :-) 

    The 485 chip is there for future use, is not used now but is powered. The designer chose to tie DE RE enable pins together so the entire chip cannot be put into shutdown. These facets of the product design were inherited when I took the project on - there is an error I found where the 485 chip is terminated with 0 ohm resistors - the first run of 100 PCBS were stuffed with 0 for R11, R12, R13. I attached that snippet. U3, at power up, could conceivably draw up to 250mA (if I remember correctly) to GND due to that nasty termination issue - I have since removed R11/R12 on any boards that ship - but some are out in the field with that nominal "stuff". I wonder if that kind of power-on event could cause an issue with U4 since they are the only 2 chips that share the 5V rail other than the 3.3V regulator. 

    Note: I have tested more than several PCBs with the nominal PCB "stuff" - including U3, U4 and errant R11 and R12 and power cycled them repeatedly with a 24V bench supply @3A current limit and have not been able to get this to happen in our test lab or test bench.  

    I do not monitor or control the MFG process but I will suggest they ensure proper ESD control measures mats / wrist straps are in place -  I do see how that could be an issue. 

    Any further thoughts? 

    Thank you

  • 0 in reply to Joe Saggio
    TI__Guru* 97046 points
    No, I think you have it pretty well covered. Let me know if you find anything else out.