# Why is my XNOR output always low?

I recently bought some Texas Instruments TTL chips. So far they've been OK, except the XNOR (SN74HC266N) chips (I've tried 3, including one that's never been in my "real" circuit) always output low no matter what the state of the inputs are.

They weren't working right in the real circuit, so I pulled them out to test them in isolation. Maybe I'm not testing them right, but here's how I did it:

----
Input1 -    - Vcc
Input2 -    -
MM Probe (+) -    -
-    -
-    -
-    -
Ground -    -
----

Where Input1 and Input2 where tied either to ground or Vcc of my supply (5.25v as measured by the meter). I expect the "Y1" pin above to measure near Vcc with respect to ground if Input1 and Input2 are either both high or both low, but the meter measures ground no matter what.

Am I testing this thing wrong, or do I have a bad run of chips? Am really new at digital electronics, so I'm fairly certain I've done something wrong.

"Open Drain" output devices operate like a switch. They don't source current of their own. They are usually used to either drive higher current devices, or to connect to a bus along with lots of other devices.

To use as a standard output device you have to supply the output with a voltage. This is in the form of a pull-up resistor (a resistor connected between the output and the supply voltage) so that when the output is "off" the resistor provides power to the output, and when the output is "on" the output pin is connected to ground. Note that the output actually operates inverted so that a logic 1 is an open switch.

• Thanks for the great explanation of both the "open drain" concept and the pull-up. It's been a long time since I took DC/AC electronics so I forgot what a pull-up resistor was. I had no idea what open drain meant but saw it in the datasheet and wondered if it was significant. Dec 29 '11 at 3:18

The SN74HC266N is CMOS, not TTL, and has open-drain outputs. Try putting a pull-up resistor (1k, say) on the output. You should then get the expected outputs, according to the truth table in the data sheet.

• I really wish I could accept two answers. The resistor value was a help, too! Thanks! I'd upvote you but I don't have enough rep yet. Dec 29 '11 at 3:16
• @Lytithwyn - If you're driving logic inputs (microcontroller, HCMOS) from the output a 10k$\Omega$ is even better, as it keeps power consumption down, like HCMOS logic does. Dec 29 '11 at 9:23
• @stevenh I'm actually driving a SN74HC08 quad AND chip. Which do you recommend for that? Dec 29 '11 at 13:39
• 10k should be fine. Dec 29 '11 at 14:56
• I just wanted to note here that I had to stick with the 1k. I tried a 10K but that outweighed the resistance of my load enough to cause the output voltage of the XNOR to be low. Mar 4 '12 at 20:59