SN74LVC126A: Internal diode

Hi Danilo,

Danilo A. said:

In the absolute maximum rating, a footnote (2) specifies that the voltage on Vo can exceeded this spec if input current ratings are observed. Can you provide more details on the voltage range if the current is lower or equal to the rated value?

We recommend not exceeding the maximum limits of the device's process node, which would be 7V total across the device (described as -0.5V to +6.5V in the absolute maximum ratings table). Voltages beyond this could cause immediate damage to the device, even if current is limited. For example, applying 1000V could instantly break down the gate oxide of the device, causing permanent damage to the device, even if the current applied is quite small.

Realistically though, it would be very difficult to apply a value larger than Vcc + 0.7V to the output as the clamp diode at the output would turn on, providing approximately 0.65V drop from the supply to the output, ie clamping the output to VCC + 0.7V or less.

Danilo A. said:

According to an answer in the TI Tech Forum here: https://e2e.ti.com/support/logic-group/logic/f/logic-forum/983169/sn74lvc126a-output-spec?tisearch=e2e-sitesearch&keymatch=SN74LVC126A.  How accurate is this answer? Can we exceed Vcc + 0.5 or not?

The diode in this case is not added to the device, but is a parasitic diode in the positive output driver (p-channel MOSFET). In either case, there is a diode there, and you can exceed VCC so long as the current through the diode is limited.

Clemens does make a mistake here - forgivably so in my opinion as he likely did not check the datasheet first and answered from memory. He indicates that this device has the Ioff feature which will prevent current from flowing into the output when VCC = 0V. This is true for most LVC family devices, however this device does not include this feature.

Yes, you can exceed Vcc + 0.5V, so long as the current is limited through the output diode.

Danilo A. said:

In that Forum answer, is D3 cathode terminated to ground or some other circuitry?

The D3 diode in that answer is connected from the output (anode) to the supply (cathode). The push-pull (or 3-state) CMOS output structure will look something like this in all devices:

The pFET (top) like all other MOSFETs has a parasitic diode, not shown in the above schematic:

This is the diode we are discussing. There is also a diode on the nFET, connecting from GND to the output, and the device includes an additional clamp diode in parallel with the nFET diode at the output for ESD protection.

Danilo A. said:

We see in testing that if Vo is in between 2-4V from an external source, there is feedback voltage driven through the part which keeps Vcc around 150mV when the part is removed from power (we want VCC at 0). What is the internal circuitry of this part? We want to know why a voltage on Vout is causing feedback voltage through the part. I have attached the circuit for more details.

I expect that the details above should suffice to explain this -- the voltage at Vcc will be coming from the current supplied by the diode and will depend on the system design. It's likely that the VCC connection of the logic gate connects to other parts of the circuit which are dissipating the extra current to keep the supply at only 150mV -- otherwise it would be more like 1.3V or 3.3V when applying 2V or 4V, respectively.

Danilo A. said:

Lastly, what type of behavior can we expect from this device when the device is OFF and Vout is non-zero?

The output will act as though there is a diode connected from the output to VCC, because there is one.

You can replace the SN74LVC126A with the SN74LV126A, which does have the I_off feature. (In the linked answer, I did not make clear that LVC and LV are different families.)