I'm using the SN74AHC1G32DBVRE4 in an application where the inputs to the gate may be energized, but the VCC input to the chip may be disconnected (i.e. floating, open-circuit). What, if anything, can I say about the state of the output in this condition? Is it high-impedance? If at least one of the inputs is at a voltage > V_ih does the output rise to the same voltage? Does it let out the smoke?
\$\begingroup\$
\$\endgroup\$
4
-
3\$\begingroup\$ You may well power the gate via its inputs. \$\endgroup\$– Andy akaCommented Mar 28 at 20:02
-
1\$\begingroup\$ What @Andyaka has mentioned is a real and often surprising troubleshooting realization for some people. For certain logic types/internal topologies, it is possible to power the entire logic chip though a 'high' on an input. This usually happens when internal protection structures on the inputs of these chips shunt the high-level input to the internal power supply rails, powering the rest of the chip. \$\endgroup\$– Chris KnudsenCommented Mar 28 at 20:08
-
1\$\begingroup\$ @ChrisKnudsen Not with this chip though. It's especially meant to handle that. \$\endgroup\$– JustmeCommented Mar 28 at 20:11
-
1\$\begingroup\$ vicatcu - Hi, I'm fairly sure that the p/n you mentioned is specific to TI and not any second-source. Therefore to help others read its datasheet, I linked the one from TI. However if I'm wrong and you're using a device from another manufacturer, please update the datasheet link. TY \$\endgroup\$– SamGibson ♦Commented Mar 28 at 20:22
Add a comment
|
1 Answer
\$\begingroup\$
\$\endgroup\$
6
The chip accepts inputs up to 5.5 even without supply.
It is specifically allowed to use it for e.g. level conversion of 5V logic to lower logic voltage levels.
The output will be high impedance, but it can't exceed VCC so basically you can't e.g. tie the output with a pull-up to another supply.
No smoke will come out. Chip won't be powered from input.
-
\$\begingroup\$ Justme - Hi, Are you sure that the 74AHC1G32 is specifically allowed for operation with the supply disconnected? I thought the 74LVC family is for that situation (and have used there). For example, TI specify behaviour of the 74LVC1G32 (datasheet here) with Vcc = 0 and they refer multiple times to "Ioff" (no supply power) in that datasheet, but not in the 74AHC1G32 datasheet. Can you see why I think the 74LVC family is the one with defined & specified behaviour for a "no supply power" situation, rather than the 74AHC? TY \$\endgroup\$– SamGibson ♦Commented Mar 28 at 20:48
-
\$\begingroup\$ @SamGibson I am quite positive. E.g. max ratings for input voltage is not determined by VCC and input clamp current is determined only for Vin < 0V. Input voltage range up to 5.5V with no mention of supply limitations. Leakage current given with Vin 5.5V and 0V for 0V to 5.5 supply range. "Input signals are high impedance when Vcc = 0V". "Allows down voltage translation". "Accepts input voltage to 5.5V". My only concern is, "inputs can expect voltages to 5.5V at any valid Vcc". So the valid range is a bit vague, is it down to 0V or just to 2V. \$\endgroup\$– JustmeCommented Mar 28 at 21:02
-
1\$\begingroup\$ Justme - Thanks for the reply. That last part in your comment is one of my concerns too, as well as the lack of mentioning Ioff for the 74AHC at all. Whereas it's mentioned several times as a feature for the 74LVC family. If behaviour without power was fully specified for the 74AHC family too, why wouldn't TI mention it? See section 8.1 in the 74LVC1G32 datasheet: "The Ioff circuitry disables the outputs, preventing damaging current backflow through the device when it is powered down.". Not mentioned for the 74AHC1G32. See what I mean? \$\endgroup\$– SamGibson ♦Commented Mar 28 at 21:11
-
1\$\begingroup\$ @SamGibson Ah, yes the Ioff feature. The difference is, AHC1G32 does not tolerate output being pulled higher than VCC even when unpowered VCC=0V. LVC1G32 disconnects output when unpowered, so it allows output of unpowered chip pulled high to 5.5V when VCC=0V. \$\endgroup\$– JustmeCommented Mar 28 at 21:28
-
\$\begingroup\$ Justme - Interesting, thanks. That may be an important difference in some designs. \$\endgroup\$– SamGibson ♦Commented Mar 28 at 22:20