A real time clock IC is one of the quietest components in any embedded design — a small chip, often no larger than a thumbnail, that keeps counting seconds whether the rest of the system is on, off, or asleep. Most engineers select one early in a project and rarely think about it again. That invisibility is a sign the design is working. But getting to that outcome requires some deliberate choices upfront: how much accuracy does the application actually need, what keeps the clock running through a power failure, and what happens when the chosen part reaches end of life five or ten years down the road?
This article covers all three dimensions — the technical fundamentals of RTC IC accuracy, the design decisions around backup power, and the practical realities of sourcing these parts across a product’s full lifecycle. It is written as a reference for engineers and procurement teams working with embedded systems in industrial, medical, IoT, and commercial applications.
What Is a Real Time Clock IC?
A real time clock IC is a dedicated integrated circuit whose sole purpose is to track the current time and date continuously, including during periods when the main system power is removed. Unlike a general-purpose timer or oscillator — which generates a periodic signal for specific tasks — an RTC maintains a running calendar: seconds, minutes, hours, day of the week, date, month, and year. Most implementations also handle leap year correction automatically.
The distinction between an external dedicated RTC IC and the peripheral RTC block embedded in a microcontroller matters more than it might seem. An MCU’s built-in RTC peripheral depends on whatever 32.768 kHz crystal the designer places on the board, and its accuracy over temperature is generally uncompensated. A dedicated external RTC IC, by contrast, is optimized for low-power autonomous timekeeping and — in higher-end versions — includes on-chip temperature compensation circuitry. The result is a measurable difference in long-term drift.
How RTC ICs Work
Inside every RTC IC there are three core functional blocks. An oscillator, driven by a 32.768 kHz quartz crystal (either external or integrated), generates a stable reference frequency. A counter chain divides that frequency down to a 1 Hz heartbeat, then increments a set of registers representing seconds through years. Those registers are exposed over a serial communication bus — almost universally I²C or SPI — so that the host microcontroller can read the current time or set an alarm with a handful of register writes.
The choice of 32.768 kHz as the standard frequency is deliberate: it is exactly 2¹⁵, which means a simple 15-stage binary divider chain produces a clean 1 Hz output with no remainder. The crystal itself is a tuning-fork type, physically small and designed for very low drive levels — which is what makes it compatible with the microamp-level operating currents that keep backup power consumption negligible.
Understanding RTC Accuracy: ppm, Crystals, and Temperature
Accuracy in an RTC IC is expressed in parts per million (ppm), where 1 ppm of error equals roughly 31.5 seconds of drift per year. A standard uncompensated 32.768 kHz tuning-fork crystal has a parabolic frequency-versus-temperature characteristic, with its peak accuracy near 25°C and degrading progressively as temperature moves in either direction. At room temperature, a typical commercial crystal carries a tolerance of around ±20 ppm — equivalent to gaining or losing approximately 10 minutes per year under stable laboratory conditions. In a real product that sees temperature swings, drift can be considerably worse.
Several design approaches exist to close this accuracy gap. The most common in dedicated RTC ICs is the temperature-compensated crystal oscillator (TCXO), where an on-chip temperature sensor and compensation circuit continuously adjust the oscillator to counteract the crystal’s parabolic response. The DS3231, for example, integrates a TCXO that holds accuracy to ±2 ppm across the 0°C to +40°C commercial range — less than 63 seconds of drift per year. Some ultra-precision modules achieve ±1 ppm or better. A second approach, available on several RTC ICs including parts from Microchip and STMicroelectronics, is a digital calibration register that the firmware can update periodically based on measured drift. This is less automatic than a TCXO but still yields meaningful improvement over a bare crystal for applications that can tolerate occasional software correction.
It is worth being explicit about the underlying error sources. Crystal tolerance at 25°C introduces a fixed offset. The crystal’s temperature coefficient introduces a dynamic, temperature-dependent component. Load capacitance mismatch between the PCB layout and the crystal’s specified load adds another offset — if the circuit presents a different capacitance than the crystal expects, the oscillation frequency shifts accordingly. Finally, aging causes a slow, monotonic drift over years as the crystal’s mechanical properties change. In a well-designed board using a TCXO-equipped RTC, temperature and load capacitance effects are handled in silicon; aging and the residual room-temperature tolerance are the dominant remaining error terms.
Backup Power Design: Coin Cells, Supercapacitors, and Rechargeable Cells
When system power is removed, the RTC IC switches internally to a backup supply rail (typically called VBAT) and continues running at drastically reduced current — commonly in the range of hundreds of nanoamps to a few microamps, depending on the device. Without this backup supply, the registers reset and the clock loses all time context; on restoration of main power, the system sees an invalid timestamp and must be re-synchronized. For many applications, that is unacceptable.
The most common backup solution is a non-rechargeable lithium coin cell, typically a CR2032. Primary lithium cells offer a nominally 3 V output, a capacity around 220 mAh, and a self-discharge rate well under 1% per year at room temperature. At the backup current levels of a moderate-accuracy RTC such as the DS3231 (approximately 3 µA), a CR2032 can theoretically sustain timekeeping for several years. Higher-end, ultra-low-power devices such as the RV-3028 draw only around 45 nA in backup mode, at which point the cell’s self-discharge — not the RTC’s current draw — becomes the binding constraint on effective battery life. One practical limitation of primary lithium coin cells is that they cannot withstand IR reflow soldering temperatures, so they must be installed in a holder or hand-soldered after the main PCB reflow process.
Supercapacitors (electrochemical double-layer capacitors, or EDLCs) offer an alternative that avoids the regulatory and logistics complications associated with lithium cells in some supply chains. They can withstand reflow soldering, recharge rapidly from the main supply, and have no end-of-life disposal constraints. The trade-off is energy density: a supercapacitor holds far less total energy than a coin cell, so its backup hold-up time is measured in days rather than years. This makes supercapacitors well-suited for applications where power interruptions are brief and predictable — industrial equipment with planned maintenance windows, for instance — but unsuitable for products that may sit unpowered in a warehouse for months before deployment.
A third option is a rechargeable coin cell (such as the ML2032 or Panasonic MS series). Some RTC ICs include an integrated trickle-charge circuit that slowly replenishes the rechargeable cell from the main VCC rail whenever power is present. This provides multi-year backup capability without requiring a holder, provided the charging and discharging characteristics are well-matched to the cell chemistry. The trade-off is that rechargeable lithium cells have a lower nominal voltage than primary cells, and the trickle-charge rate must be carefully selected to avoid overcharging.
In summary, the backup power decision depends on three variables: how long the product may be unpowered, whether the PCB process constrains component placement, and whether regulatory requirements around lithium cell transport or disposal are relevant to the supply chain.
Common RTC IC Families and Their Trade-offs
A handful of parts have dominated the RTC IC landscape for years, and they remain the first reference points for most embedded designs.
DS1307 (Analog Devices / Maxim Integrated): The DS1307 is an I²C device that has been in production long enough to appear in countless reference designs and code libraries. It uses an external 32.768 kHz crystal and operates from a 5 V supply, which limits its applicability in 3.3 V systems without additional circuitry. Its accuracy is bounded by the uncompensated crystal — typically ±20 ppm or worse, roughly 10 minutes of drift per year under normal conditions. For applications where timestamp precision is not critical and the host system can periodically re-synchronize the clock, the DS1307 remains a cost-effective choice. For applications requiring tighter accuracy, it is the wrong starting point.
DS3231 (Analog Devices / Maxim Integrated): The DS3231 incorporates a factory-calibrated TCXO and requires no external crystal, which also eliminates the PCB layout sensitivity associated with crystal placement. It communicates over I²C and operates across a wide supply voltage range. Its ±2 ppm accuracy over the commercial temperature range — under 1 minute of drift per year — makes it the practical accuracy benchmark for the class. The integrated temperature sensor can also be read directly over I²C, which is occasionally useful in system designs that need a rough ambient temperature reading without a dedicated sensor.
PCF8563 (NXP Semiconductors): The PCF8563 is a low-power I²C RTC with an operating voltage range of 1.0 V to 5.5 V, making it compatible with essentially any logic-level supply in use today. Its backup current consumption is among the lowest in the standard RTC IC class. It includes programmable alarm and timer functions and a programmable clock output at 32.768 kHz, 1.024 kHz, 32 Hz, or 1 Hz. Accuracy without compensation is similar to other crystal-based RTCs, but its combination of low power and wide voltage range makes it a strong candidate for battery-operated and IoT products.
RV-3028-C7 (Micro Crystal): The RV-3028 represents the current generation of ultra-low-power precision RTCs. At 45 nA backup current and ±1 ppm accuracy at 25°C, it occupies a niche for applications where both precision and battery longevity are critical simultaneously — wearables, environmental sensors, and remote data loggers chief among them.
M41Txx series (STMicroelectronics): ST’s M41T family covers a range of I²C RTCs with varying feature sets, including watchdog outputs and serial EEPROM on some variants. Several members of this family have reached or are approaching end-of-life status, which is a relevant consideration discussed further below.
Selecting the Right RTC IC for Your Application
RTC IC selection is straightforward once the application requirements are clearly defined. The following parameters deserve explicit answers before comparing datasheets:
- Required accuracy: Is ±10 minutes per year acceptable, or is ±1 minute per year necessary? The former is achievable with a standard crystal-based RTC; the latter requires a TCXO or digital calibration.
- Operating temperature range: Commercial range (0°C to +70°C) or industrial (-40°C to +85°C)? TCXO-based parts typically specify accuracy across the industrial range; uncompensated crystal-based parts degrade significantly at temperature extremes.
- Supply voltage: 5 V legacy systems versus 3.3 V or lower modern MCU supply rails determine which parts are directly compatible without level-shifting.
- Communication interface: I²C is by far the most common for RTC ICs and consumes only two signal lines. SPI variants exist and are occasionally preferred when the I²C bus is already congested or when the MCU lacks a dedicated I²C peripheral.
- Backup power strategy: Coin cell, supercapacitor, or rechargeable cell — and whether the RTC IC itself includes trickle-charge circuitry.
- Integrated features: Alarms, countdown timers, programmable interrupt outputs, square-wave outputs, and embedded EEPROM or RAM are available on various devices. Include only features the application will actually use, since unused circuitry adds cost without benefit.
- Component lifecycle status: Check the manufacturer’s product lifecycle status at design time, not at production time. Parts that are currently in growth or maturity carry lower supply risk over a multi-year program than parts already in decline.
Where RTC ICs Are Used
The range of products that depend on an RTC IC is broader than the embedded hobbyist framing often suggests. In industrial automation, RTCs provide the time-stamping infrastructure for process event logs, shift scheduling, and maintenance interval tracking. In medical devices — patient monitors, infusion pumps, diagnostic recorders — accurate timestamping is often a regulatory requirement, not just a design convenience, because treatment records must be traceable to verified clock times. Smart meters in energy infrastructure use RTCs to enable time-of-use tariff calculations, which require accurate timekeeping across multi-year deployment lifetimes without network synchronization. IoT sensor nodes use RTCs to timestamp locally logged data before periodic upload, particularly in deployments where network connectivity is intermittent. Even automotive applications increasingly rely on RTCs for ADAS event logs and infotainment system clocks that must survive ignition-off states.
In many of these contexts — industrial, medical, automotive, metering — the product’s expected service life is measured in years or decades, not months. That time horizon is directly relevant to component sourcing strategy, because an RTC IC selected in year one of a design program may be approaching end-of-life by year seven or eight of production.
Obsolescence and Lifecycle Sourcing for RTC ICs
RTC ICs are an interesting category from a lifecycle perspective. Because they are simple, well-understood devices, they tend to stay in production for a long time — but when a manufacturer does discontinue a line, the options narrow quickly and the impacts can be disproportionate to the part’s cost. A low-cost RTC IC embedded in a medical monitor or industrial controller can trigger a board redesign that costs far more than the component itself.
Several well-known RTC parts have already gone obsolete or reached “not recommended for new designs” (NRND) status. Parts from STMicroelectronics’ M41T series, older Dallas Semiconductor parts that predate Maxim’s acquisition, and some Epson/Seiko Instruments RTC modules fall into this category — confirm the published status for the specific part number before relying on it, since lifecycle states change. Engineers maintaining legacy products or managing long-running production programs regularly encounter the scenario where the original part is no longer available through any in-production channel.
The response options when an RTC IC goes obsolete are well-established, though none is without cost. A last-time buy (LTB) at the time of the end-of-life notice can extend runway for programs with known production volumes, but it ties up capital, consumes warehouse space, and carries its own risks if demand projections are wrong. A form-fit-function equivalent replacement — a pin-compatible or software-compatible substitute from a different manufacturer — is the cleanest solution when one exists, but RTC register maps and communication timing are not standardized across vendors, so driver changes may be required even for nominally equivalent parts, and final approval of any substitution belongs to the customer’s engineering team. In some cases a board redesign to a modern, actively supported part is the most cost-effective long-term answer, particularly for high-volume programs with many remaining production years.
Proactive lifecycle monitoring — tracking manufacturer product change notices (PCNs) and end-of-life notifications for every critical component on a BOM — is the recommended practice, but it is genuinely difficult to execute systematically across a large component portfolio. The window between an EOL notification and a manufacturer’s final shipment date has narrowed over the years, compressing the time available to evaluate alternatives or secure bridge stock. Engineering and procurement teams that treat lifecycle risk as an ongoing background activity, rather than a crisis response, consistently fare better than those who address it only after production stops.
Supply-chain support is most useful when it spans the whole of a production program rather than only its difficult end. A distributor carrying verified inventory across active, EOL and obsolete parts can supply a current-production RTC for a new build, hold bridge stock while a longer-term fix is qualified, and keep lifecycle status visible on the bill of materials in between — which is what turns an end-of-life notice into a planned transition rather than a line stop.
WIN SOURCE is a global independent distributor of electronic components, in business since 1999, supporting component supply, shortage sourcing, lifecycle management and alternative solutions across manufacturers including Texas Instruments, STMicroelectronics, NXP, Microchip and Analog Devices. Its WinLink BOM tool flags at-risk parts across a full bill of materials, and it holds AS9120 and ISO 9001:2015 certifications and is a member of ERAI.
A useful starting point when assessing sourcing risk for an RTC IC already on your BOM is to check current stock and lifecycle status across your distributor network. The real time clock category covers both current-production and hard-to-find parts.
Final Thoughts
Real time clock ICs are uncomplicated components by semiconductor standards, but the decisions around them — how much accuracy the application actually requires, what keeps the clock alive through a power interruption, and what the sourcing plan looks like across a multi-year production program — deserve more thought than they typically receive at design time. Choosing a TCXO-equipped part like the DS3231 when a simple crystal-based device would suffice adds cost unnecessarily; choosing an uncompensated part for an industrial data logger that needs reliable timestamps at -30°C creates a reliability problem that surfaces years into the product’s life. The same logic applies to backup power and to lifecycle planning: the trade-offs are well-defined, the information needed to make the right call is available in datasheets and manufacturer lifecycle data, and the cost of getting it wrong usually exceeds the cost of getting it right by a wide margin.
For engineers and procurement teams working with RTC ICs in long-lifecycle applications, the technical selection decision and the supply chain continuity decision deserve equal attention. A part that performs well in the lab but becomes unobtainable during production is not a solved problem.
Frequently Asked Questions
How accurate is a real time clock IC?
Accuracy is quoted in parts per million, and 1 ppm works out to roughly 31.5 seconds of drift per year. A standard uncompensated 32.768 kHz crystal is typically around ±20 ppm at 25°C, or about ten minutes a year, and it gets worse as temperature moves away from room temperature. A TCXO-equipped part such as the DS3231 is specified at ±2 ppm over 0°C to +40°C, which is under 63 seconds a year.
Why do RTC ICs use a 32.768 kHz crystal?
Because 32,768 is exactly 215, so a plain 15-stage binary divider chain turns the oscillator frequency into a clean 1 Hz tick with no remainder and no correction logic. The tuning-fork crystals cut for that frequency are also physically small and run at very low drive levels, which is what keeps backup current in the microamp range.
How long will a CR2032 keep an RTC running?
It depends almost entirely on the part’s backup current. At the few-microamp backup draw of a mid-range RTC, a CR2032’s roughly 220 mAh will support timekeeping for years. On an ultra-low-power device drawing tens of nanoamps, the cell’s own self-discharge becomes the limiting factor rather than the RTC. Note that primary lithium coin cells cannot go through IR reflow, so they need a holder or hand-soldering.
Should I use a supercapacitor or a coin cell for RTC backup?
Use a coin cell when the product may sit unpowered for weeks or months — a unit in a warehouse before deployment, for instance. Use a supercapacitor when outages are short and predictable and you want a reflow-safe part with no lithium transport or disposal constraints; its hold-up is measured in days, not years. A rechargeable cell with an on-chip trickle charger sits between the two.
What are the options when an RTC IC goes end-of-life?
A last-time buy sized against your remaining production, a form-fit-function replacement from another vendor, or a redesign onto an actively supported part. The middle option is rarely as clean as it sounds, because RTC register maps and bus timing are not standardized across vendors, so expect driver work even for a pin-compatible part — and final approval of the substitution belongs to your engineering team.
Part discontinued, obsolete or hard to find?
WIN SOURCE is a global independent distributor of electronic components, supporting component supply, shortage sourcing, lifecycle management and alternative solutions. Visual inspection, non-destructive testing and electrical testing are applied according to material risk, customer requirements and the applicable quality process, and orders are covered by the published return policy and a 3-year warranty on eligible components.
Stock and pricing are updated daily — search a part number or contact us for current availability, lead time and a quote.
© 2026 Win Source Electronics. All rights reserved. This content is protected by copyright and may not be reproduced, distributed, transmitted, cached, or otherwise used without the prior written permission of Win Source Electronics.
