CD14538B: Date code marking of CD14538BE

I just verified that every single board that I have that is unreliable has the same makings on it, about 200 or so boards. Has there been any customer reported issues with this lot reported to TI in the past?

Hi Kyle,

We haven't seen any returns or reports that I am aware of. Would you care to share a schematic or more details as to what the issue is?

I believe that the "AKG3" portion of the package markings is the lot code. If all problematic devices share that code then they are likely from the same lot. Who did you purchase this lot from previously?

Best,

Malcolm

Above is the schematic and scope waveforms of the pins detailed in the postit below the scope. The thing that is a tad frustrating about this is that the part triggers about 20-50% of the time, with identical specs to the onsemi/fairchild part. There is not an issue of reliability with the parts mentioned below, only the TI parts. 

Below is scope waveforms of the onsemi/fairchild part, which is identical to the scope waveforms of the toshiba part.

We purchased these parts through mouser in March of 2024. The only other supplier we have used in the past five years is digikey.

I remember seeing this issue in another thread. Would you be able to link that thread here so I can recall what's already been said? Unless I'm mistaken.

Best,

Malcolm

CD14538B: Should pins 1 and 15 be grounded or left ungrounded? - Logic forum - Logic - TI E2E support forums

Above is the linked thread. I posted this thread last week looking for support and was told by another user that it was most likely a damaged chip. I do not see how ~200 chips can be damaged unless it was a bad lot or a difference between the input structures of the TI chip compared to Onsemi and Toshiba. The only difference between the datasheets of these chips are delay times and the Toshiba chip has a lower operating temperature. We adhere to strict ESD policies here.

Best,

Kyle

Hi Kyle,

I'll run it by some coworkers. Could you confirm for us if they are all from the same lot or not?

Best,

Malcolm

Malcolm,

They all have the same package marking below.

Package Markings:

1AAKG3MG4

CD14538BE

Best,

Kyle

Hi Kyle,

Could you please see here: https://www.ti.com/support-quality/additional-information/customer-returns.html

I think in this case the proper thing to do moving forward is initiate a failure analysis.

Best,

Malcolm

Hi Malcom,

Before doing so, I was wondering if there is a minimum current into the chip needed to trigger the chip? I did not see a spec for this on the datasheet.

Best Regards,

Kyle Hansen

CMOS devices have high-impedance inputs; the specified input current is less than 1 µA.

Clemens,
The reason for my question about input current was that I have been experimenting with different input resistances. Some of the resistances send the mono into an abnormal state, which will trigger occasionally but not on every rising edge (What I am seeing in my application). Then, there are other resistances that will not trigger the mono at all. Lastly, there are resistances that the mono triggers every single time. I currently see that resistances of about 0-5 ohms and 25- 50 ohms will trigger the mono unreliably; resistances of 5-25 and 10meg and above will not trigger the mono at all; otherwise, it triggers as expected. I cannot explain this behavior and was curious if it was a current issue of some sort.

Best Regards,

Kyle

Hi Kyle,

Yes since these are CMOS inputs I would expect leakage currents of max 1-5 uA.

Best,

Malcolm

Update:

I have received a new lot from Digikey. I am getting a consistent brake trigger within my circuit with the newly purchased TI parts. Although this may be the case, I am experiencing some very strange behavior still; the part triggers consistently but will sometimes trigger a pulse that is much shorter than the RC circuit specifies. See below. Any further help would be greatly appreciated.

Perfect Trigger:

(TR+ = blue, magenta = Q*, and Yellow = Q of the d-flip-flop)

Short Pulse Trigger:

(TR+ = blue, magenta = Q*, and Yellow = Q of the d-flip-flop)

Short Pulse Trigger zoom:

Blue = Q*, magenta = TR+, and Yellow = Q of the d-flip-flop)

Please be aware that I have the scope configure differently for the "zoom" scope plot.

Best Regards,

Kyle Hansen

Hi Kyle,

1. Could you confirm the top marking for these devices, to see if it is indeed a different lot 
2. Could you provide the RC value on the RxCx pin (size of the resistor and the capacitor). Also, could you probe the pin during this instance so we can see what the RC discharge looks like
3. What peak voltage level does the input trigger signal reach? ESD protection is limited on these devices and I could see the input potentially causing latchup if it goes above 12V. 

Probably even if this is a different lot initiating a return here is also going to be the best thing to do just to see if there's some problem with these devices. Unfortunately I get the feeling that the process has changed and this old CD4000 device just doesn't work as expected anymore.

Best,

Malcolm

Hi Malcom,

1. The markings are:

43A160MG4

CD14538BE

2. The values of R and C on the pin are 1Megaohm and .1microfarad, respectively. I will not be back into the office until Thursday; therefore, I will not be able to provide scope waveforms until then. 

3. The input voltage peak is about 12V. The CD14538 is powered by the same regulator that the TR input signal is coming from. A switch is used to switch the 12V regulator signal on and off at the TR input. The power to the CD14538 is not affected by the switching on and off at the TR input.

Best Regards,

Kyle Hansen

Hi Kyle,

The only thing I would consider is measuring the voltage/leakage at the RC pin to see if it's consistently charging/discharging to the full 12 volts. But otherwise I am out of ideas and it sounds like these devices simply aren't working properly. It would be great if you could get a failure analysis done with us and we can determine if there are any physical issues with these lots.

Best,

Malcolm

Malcolm,

Below is a scope shot of TR in yellow, RC in pink, and Q* in blue. It does appear that there is substantial overshoot in the TR+ input signal that I did not notice before. The offset of the probe is a little off, so I expect this overshoot to be ~15V. Also, there seems to be no discharge of the RC network. I also measured the Vdd pin with a peak voltage of ~13V.

Best,

Kyle

Malcolm,

Thank you for your help and helping me address the issue I was experiencing.

Best Regards,

Kyle Hansen