AB1805
Epson alternative: RX8901CE
The AB18X5 family is the low-current benchmark in this field, and we are not going to pretend otherwise. What the headline figure does not say is that the crystal is not included and that the supply stops at 3.6 V. The comparison covers the SPI relative AB1815 as well.
Features of the AB18X5: 256 byte battery-backed RAM, countdown timer, watchdog with hardware reset, supercapacitor trickle charger, an integrated power switch of about 1 Ω and a programmable analogue voltage comparator.
AB1805 versus RX8901CE
| AB1805 / AB18X5 — Abracon | RX8901CE — Epson | |
|---|---|---|
| Manufacturer | Abracon | Epson (Seiko Epson) |
| Interface | AB1805 I²C (400 kHz) · AB1815 SPI (2 MHz) | I²C (400 kHz) |
| Crystal integrated | no — external 32.768 kHz crystal required | yes — DTCXO, factory calibrated |
| Accuracy | XT mode ±2 ppm digitally calibrated · RC mode ±16 ppm · autocalibration 3 – 35 ppm | ±3.0 ppm across −40 to +85 °C (DTCXO) |
| Timekeeping current | 55 nA at 3 V (XT) · 14 nA (RC) · 22 nA (autocalibration) | 240 nA typ. · 1.5 µA max. |
| Supply | 1.5 – 3.6 V (VCC and VBAT) | 1.6 – 5.5 V · clock down to 1.1 V |
| Max. temperature | −40 … +85 °C | −40 … +105 °C |
| Package | QFN-16, 3 × 3 mm · 9 mm² footprint | CE 3.2 × 2.5 × 1.0 mm · 8.0 mm² footprint |
| Extras | 256 byte RAM, watchdog, power switch, trickle charger, comparator | 32 timestamps (FIFO or 3 channels), 256 byte SRAM, self-monitoring |
What speaks for and against the change
We sell Epson — and we still say where the competitor is better. You would notice it at the first datasheet comparison anyway.
Honestly
- 55 nA in crystal mode and 14 nA in RC mode are well below anything Epson offers. On a coin-cell design that is a real argument and we will say so
- An integrated power switch of about 1 Ω, a watchdog with hardware reset and a programmable voltage comparator go beyond a plain RTC
- Both bus variants exist in the same family — I²C as AB1805, SPI as AB1815
The difference
- The crystal is not included. Add a 32.768 kHz crystal, its load capacitance matching, two capacitors and the layout risk of the oscillator loop — and the 55 nA only hold with a crystal that behaves
- Supply: the AB1805 stops at 3.6 V. It is not a 5 V part. The RX8901CE runs to 5.5 V
- Accuracy: ±2 ppm is a calibration range at one temperature, not compensation across the range. Autocalibration lands between 3 and 35 ppm. Epson's ±3.0 ppm holds across −40 to +85 °C
- Temperature: +105 °C against +85 °C
- 256 byte SRAM and 32 timestamps are on the Epson side too — that line is a draw, not a loss
Frequently asked questions
What is the Epson alternative to the AB1805?
The RX8901CE is the closest match, with the RX8130CE where a reset output and charge control matter more than accuracy. Neither is a drop-in: package, pinout and register map differ.
Does the AB1805 really run on 55 nA?
In crystal mode at 3 V, yes — and 14 nA in RC mode. Those are the lowest figures in this comparison. What the number does not include is the external crystal the part needs, along with its load capacitance matching and the layout risk in the oscillator loop.
Is the AB1805 5 V tolerant?
No. The datasheet specifies 1.5 to 3.6 V for both VCC and VBAT. In a 5 V design it needs a level shifter, or a different part. The RX8901CE runs from 1.6 to 5.5 V.
Who supplies an alternative to the AB1805 in Europe?
Timing Devices Europe in Ladenburg near Heidelberg, Germany stocks the complete Epson RTC programme in its own warehouse — physically on site, not at the manufacturer. Samples, comparison and series demand in one go, answered the same working day.
Samples and a comparison for the AB1805.
Tell us which version you use today, together with quantity and date. We will put together the Epson equivalent with its datasheet and the migration effort — from our own warehouse in Ladenburg near Heidelberg, Germany.