I want to use a supercapacitor to provide backup power to a DS3234 RTC.
I am stuck as how to charge the capacitor and then not have it discharge through the charge circuit when the power goes off.
The obvious solution would be a diode but even these have reverse leakage and the cap would only hold up for a few days.
Has anyone out there used a supercap with a RTC sucessfully? (I thought that this would be relatively common).
I found a couple of complex circuits using a couple of ICs and several passives but I was hoping to keep it simple and small.
What about the cap's own self-discharge rate, wouldn't that be greater than any diode's leakage?
Rob
Yes, supercaps are usually only good for relatively short outages.
Still, some simple calculations show that a 1F supercap going from 2.5V to 2V with the 25nA leakage current typical of a 1n4148 will last more than 200 days. (that spec is with 20V reverse voltage, and it seems to go down as voltage goes down.)
Or you could use a relay in the charging circuit. Then by placing the NO contacts in series with a diode in the feed to the super capacitor, when power goes off the relay opens the contacts and absolutely none of the microscopic reverse current through the diode feeds back to the charge circuit. Sometimes we are guilty of over-engineering. A diode is all you need !
What is wrong with a battery?
mirith:
What is wrong with a battery?
Gawd. Just be practical and all.
Also, I might be wrong, but won't the frequency change as the voltage level of the super-cap drops? The reason for the battery is to continue supplying power to the 32-ish kHz crystal, so you don't lose track of time. However, over time, the voltage will drop and this would change the crystal's response. It would be fine for short periods of time, but not for long outages.
If I attempted it, I'd use a pnp to charge the cap which powers the rtc... the transistor will also behave like a diode.
Is leakage even an issue? How llong will the rtc be running from the cap? Days?