Many of these are embedded RTCs inside MCU chips. Initially, we’ll focus on external RTCs. Most of the below, except for RV-3028 require a separate external crystal.
Part #
Manufacturer
Cost (100 pcs)
Interface
Accuracy
Power
Notes
AM1805
Ambiq
$0.65
I2C
2 ppm
22nA
Watchdog timer with hardware reset, 256 bytes ram, auto-calibrates internal osc from xt to achieve lower power
Sometimes it is desirable to run a RTC from a supercap so you don’t have to worry about issues with battery going bad, etc. Below is a comparison of one range of options. The life assumes we will discharge the cap from 3.3V to 1V @ 45nA. The formula is:
That RV-3028 RTC is quite impressive! I hadn’t seen that chip before, thanks!
But there’s quite a few supercaps available at normal retail distributors, like Digi-Key, which come in quite a bit less expensive than your examples if you’re able to accept a bit higher ESR (equiv series resistance), which should be fine for an RTC circuit drawing <1uA. For example, CDA’s CE5R5105HF-ZJ is a 1F 5.5V coin shape supercap for $0.84/ea for qty 100, but the internal resistance is around 30ohms.
Am I misunderstanding something about supercap selection such that this CDA part wouldn’t be reasonable to choose for an RTC circuit?
I assume the 1000 hours means it will hold a charge for that long without self-discharging? Self-discharge is probably the dominant factor with a supercap as the RTC draw is so low.
1000 hours is 41 days.
For the PHVL-3R9V155-R, the lifetime is listed as 0.6 to 2 years:
I assume “endurance” and “lifetime” mean how long the supercap will keep a charge. If these are equivalent, then the more expensive caps do have a significant advantage if you have a scenario where you want to set the time during manufacturing, and have it be valid during installation up do a year later. If you don’t need that, then the cheaper ones seem like a good option.
