CD74HC4053 cmos switch: off-state resistance is very low

[exe]

Hi!

I can't make CD74HC4053 working. In off state it shows 2-3uA leakage. What am I doing wrong?  . I'm attaching the schematic.

If I measure resistance in OFF state, my DMMs go nuts: https://youtu.be/_osrkhBMWwA . I measured leakage, it shows 2-3uA. I tried to heat the IC with hotair, no change in leakage. In on-state, switches show 35 Ohms. All switches behave similarly. There is leakage between switches, which I find weird.

What I tried, none of it worked:
1. I changed the IC from TI (in DIP package) to and smd version from nexperia
2. Changed breadboard. Breadboard without IC shows no measureable leakage.
3. Added decoupling caps on supply rails
4. Checked voltage levels with DMM
5. Changed power supply to an old linear one (presumably low noise)
6. Replaced wires
7. Checked voltage rails
8. Replaced DMM leads
9. Cleaned flux as much as I could and dried the breakout board.
10. Tripple-checked everything, including IC orientation, can't find any issues.

Help

[exe]

Photos of actual setup. First picture shows TI's ic on the breadboard. The second picture show another setup. I removed the IC to check if there is any leakage from breadboard, or breakoutboard. There is not leakage.

May be it's ESD damage, soldering another IC... This time I'll hotair gun so I don't touch it with my soldering iron.

[exe]

Changing IC didn't change anything. So, I think I'm doing something fundamentally wrong. Can anyone spot a problem on pictures?

How I do measurements: I tie pin E to Vcc (this should put switches in high-Z), then measure resistance between pins 4 and 5.

Power condition: Vcc is set to +5V, Vee is -5V, GND is GND.

[Kleinstein]

With 2 supplies one may cause more trouble. The HC4053 also work with just a positive supply. So try VEE = GND for a 1st test.

Some bench meters may exceed 5 V in the ohms mode - a handheld meter usually uses lesss than 2 V and should not cause trouble.

When wired as shown pins 4 and 5 should be connected. Pins 3 and 5 would be off.

[Jay_Diddy_B]

Hi,

I think the issue here is that the measurement source, the DMM, is floating with respect to circuit GND. So the common voltage is undefined.

Try measuring with pin 4 connected to GND (pin .

If this works, you explore the effect of common mode voltage by connecting a potentiometer from +5 and -5V connect the wiper to pin 4. You can see how the leakage current changes with common mode voltage.


Regards,
Jay_Diddy_B

[exe]

Hi,

I think the issue here is that the measurement source, the DMM, is floating with respect to circuit GND. So the common voltage is undefined.

Try measuring with pin 4 connected to GND (pin .

Oh, man, thanks a ton! Yes, it did help. Also, measurements are making much more sense now. Like resistance of 35 Ohm that I measured before is suspiciosly low. Or, afaik, results were changing depending in which polarity I was connecting the DMM, which made me puzzled.

Tomorrow I need to absorb the information you gave me. Probably I'll have more questions . Thanks again!

[Alti]

I think the issue here is that the measurement source, the DMM, is floating with respect to circuit GND. So the common voltage is undefined.

+1

This is a CMOS chip which means that there are two moses passing current. The impedance is not only a function of supply voltage but also a common voltage. Connecting autorange multimeter that expects a constant resistance to something that has resistance varying with common voltage gives a disco effect. Switch off autorange.