I'm using a water detection circuit similar to the one found here

I put it all together and it worked pretty well for a hobby project.

The circuit hinges around an AC signal that's generated from a (CD4093B) Here's a link for it at digikey.

However I was wondering if I should be concerned about ESD on the input of the "sensor"

Perhaps I should consider changing the CD4093B over to one with ESD protection such as 74LCX00YMTR again here's a link at digikey.

I prototyped it with a breadboard hence why the first part was a Through Hole but am planning on laying out a board so I could switch over to a Surface Mount part.

I was curious if changing this part is necessary or do the two capacitors that block out the DC from the sensor protect it form ESD as well?

Edit - Added additional thoughts.

I've decided to double footprint my prototype board with both the CD4093 (and include the ESP protection diodes both responses suggested) and also SN74LV132ADR (IC QUAD NAND GATE 14-SOIC)from Digikey.

That way I can compare the two and see if there are any other changes I should address.

I read though the HCMOS Design Considerations provide by dfowler7437 which lead me to the TI Logic Selection Guide After skimming over the first few diagrams I settled on the LVC family. I ordered some SN74LV132ADR and hope with the following changes I should be set.

I'll include a cheap 100mA regulator for the power input of the SN74LV132ADR. Now, since it's a different part from the reference design (and a different voltage 3.3V instead of 12V) will I need to change the R1 and C1 to a different value to maintain the same frequency? (I'm assuming the frequency is tuned to get the sensor to work around the capacitive value of water)

Also, Since I'm changing the NAND gate to a different voltage I believe I need to change from a P Channel to an N channel (and then invert the output from N2 into say N3 or N4) in order to get the N Channel to work with a different Gate voltage. Does this make sense?

  • \$\begingroup\$ added more info for a slight design change to 3.3V \$\endgroup\$
    – user16105
    Commented Jan 15, 2014 at 18:37

2 Answers 2


The CD4093 does have ESD protection. There are diodes from the inputs to power and ground and a small series resistance. The CD4093 also has characteristics that are important for your application but lacking in the 74LCX00: it will operate at 12V and it has Schmitt-trigger inputs.

If you are worried about ESD you could add your own diodes from the sensor terminals to ground and to the 12V power supply. The diodes should be oriented so that if the voltage on one of the sensor terminals goes above the power supply voltage then one diode will conduct (this diode has its anode at the sensor and its cathode connected to the power supply), and another diode will conduct if a sensor terminal goes below ground (this diode has is anode connected to ground and its cathode connected to the sensor terminal).

  • \$\begingroup\$ Thanks, I added the ESD diodes to the design and also am thinking about using a SN74LV132ADR at 3.3V and changing the output so it'll still work with the 12V relay. \$\endgroup\$
    – user16105
    Commented Jan 15, 2014 at 18:40

Add a diode from N2 inputs to the 12V rail to catch positive voltage spikes. D1 already provides negative spike protection. The extra diode will clamp spikes at about 12V providing some protection. You could use a Schottky diode for D1 and the protection diode.

The CD series CMOS parts dont have much ESD protection, the later HC parts do. The HC parts include clamp diodes which can handle about 20mA and are designed to handle the spike which a human might deliver.

You can read more about HCMOS ESD protection in the TI paper linked here: HCMOS Design Considerations

Basically the inputs are diode clamped to 0.5V of the power rails. A typical diode will clamp at around 0.7, but you can use Schottky, which clamp at about 0.5 if you want that extra protection.

Read more about CD40 series parts and ESD sensitivity at this link: Wipeda

  • \$\begingroup\$ I overlooked the point Hass's makes about the input being Schmitt-Trigger, you need that feature so the HC part is not really an option. The Shmitt-Trigger feature is required in your circuit to make an oscillator out of N1. \$\endgroup\$ Commented Jan 8, 2014 at 0:20
  • \$\begingroup\$ Hass and I are saying the same thing, I think.. Add the extra diode... \$\endgroup\$ Commented Jan 8, 2014 at 0:29
  • \$\begingroup\$ Thanks for the link to the HCMOS Design Considerations. I read though that and it gave some good insight. \$\endgroup\$
    – user16105
    Commented Jan 15, 2014 at 15:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.