Low leakage 100nF caps?

[Chris Wilson]

A circuit has the following warning about the need to use low leakage caps. How do I check the ones I have ARE low leakage please?

"TO LIMIT INFLUENCE OF 74HC86 INPUT
MAX LEAK CURRENT OF 0.2?A @ 6V @ 55°C
(EQUIVALENT TO 30 Megohms)
ALL TIMING RV'S: 2.2 Megohms MAXI
USE LOW LEAKAGE CAPACITORS !!"

Thanks!

[mikeselectricstuff]

You'll struggle to find 100nF that isn't low leakage. I don't know if they still make 100nf electrolytics, but pretty much any other type (ceramic, polycarb, polyprop etc.) is fine.

[Chris Wilson]

Thanks Mike, I used new branded ceramics from Farnell, so should be OK. I appreciate the advice (and your videos...) Cheers.

[SeanB]

If you want really low leakage use caps rated at over 250V, 400V or 630v film units are guaranteed to have a very low leakage at rated voltage, and at CMOS levels it is pretty close to being equal to the PCB surface leakage. 16V ceramic units can have a quite high leakage, and with the lovely voltage dependence of capacitance timing will be pretty poor with supply voltage variations. If you are going to use ceramic use 100V units, though they will be a lot bigger than 16V units

[Harvs]

If you're using brand named caps, then this should be fairly easy to calculate.

They typically spec the insulation resistance in Mohm - uF's, so you can just divide the insulation resistance by the capacitor value in uF (so 0.1 for 100nF.)

For example this X5R dielectric:
http://www.ece.ucdavis.edu/vcl/asap/asap_v1/docs/X5R_D.pdf

I get 5Gohm.  So at 5V you should get about 1.2nA leakage.

Also note, depending on the dielectric you get, this is highly temperature dependent.