BQ32002DR
Product Specifications
BQ32002DR
Texas Instruments — Power Management & Signal Chain
Product Overview
The BQ32002DR from Texas Instruments is a power management / signal chain IC delivering high efficiency, precision regulation, and robust protection. Texas Instruments (TI) is a global leader in analog ICs, power management solutions, and embedded processors.
Superb Automation partners with leading brands including Qorvo, TI, ADI, XILINX, NVIDIA, Intel, ST, MACOM, Renesas, and Coilcraft to provide complete BOM and supply-chain solutions.
Key Features
- High-efficiency power conversion with low quiescent current
- Wide input voltage range with precision output regulation
- Integrated over-current, over-temperature, and short-circuit protection
- Small footprint packaging for space-constrained designs
- AEC-Q100 qualified options available for automotive applications
- Flexible sequencing and soft-start control
Key Specifications
| Input Voltage | V |
| Output Voltage | V |
| Output Current | A |
| Switching Frequency | kHz |
| Efficiency | % |
| Quiescent Current | µA |
| Topology | — |
| Package | — |
| Operating Temp | °C |
| Protection | — |
Applications
Request Quote — In Stock, 100% New & Original, Full Warranty
BQ32002DR: a real time clock that runs from a capacitor
The BQ32002DR is a serial real time clock in an 8 pin SOIC body. It counts seconds, minutes, hours, day, date, month and year, applies leap year correction by itself, and reports all of it to the host over an I2C bus running at up to 400 kHz. The register map and the pin functions follow the common industry part, so a board that already carries an older clock can take this one without a change to the host code.
The feature that decides most projects is the backup path. The part switches to a backup supply on its own when the main rail falls, and that backup can be a small capacitor or a non rechargeable battery. A capacitor is the cheaper route and it removes the coin cell from the service schedule, which is why metering and building control boards tend to use it. The backup pin works from 0.9 V up to the main rail, so a supercapacitor charged to 3.3 V keeps the oscillator alive for days rather than hours.
Inside the part there is a 32.768 kHz oscillator, a set of time registers that are all updated in the same instant once a second, an oscillator fail flag, and a calibration register. The single instant update matters on a bus that is read by an interrupt driven host, because a host that reads seconds at 59 and minutes at 00 in two separate transactions will otherwise report a time that never existed.
Key data for the BQ32002DR
| Function | Serial real time clock with leap year correction |
| Interface | I2C, serial clock up to 400 kHz |
| Crystal | 32.768 kHz tuning fork crystal on the oscillator pins |
| Calibration | Programmable from minus 63 ppm to plus 126 ppm |
| Pre-calibration accuracy | Plus or minus 35 ppm typical at 25 C with the reference crystal |
| Time registers | Seconds, minutes, hours, day, date, month and year, all latched together once a second |
| Backup supply | Automatic switchover, backup pin operates from 0.9 V to the main rail |
| Supply current | 65 microamps typical, 200 microamps maximum |
| Supervision | Oscillator fail flag and a stop bit the host can use to halt the clock |
| Package | SOIC, 8 pins, 4.90 mm by 3.91 mm |
| Temperature | Minus 40 to 85 C ambient |
| Ordering | BQ32002DR, 2500 pieces on a large tape and reel, nickel palladium gold finish, moisture level 2 with a one year floor life |
What the calibration range is worth in seconds per day
One part per million is 0.0864 seconds per day. A tuning fork crystal that is 20 ppm fast will therefore gain about 1.7 seconds a day, or roughly ten minutes over a year, and that is before any temperature drift is counted. The calibration register in this part covers minus 63 to plus 126 ppm, which is wide enough to pull a typical crystal back to within a few ppm at room temperature.
Where the range runs out is temperature. A tuning fork crystal turns over near 25 C and its frequency falls as the board moves away from that point in either direction, so a clock trimmed on the bench at 25 C will run slow in a cabinet at 60 C. Nothing in this part compensates for that, because it has no temperature sensor and no second order correction. A product that has to hold time across a wide ambient range either accepts the drift, resynchronises from a network source, or uses a temperature compensated clock instead.
The published plus or minus 35 ppm figure is a pre-calibration number measured with a specific crystal on a reference board. It is a statement about the silicon and that crystal pair, not a guarantee for the crystal that ends up on your board. The crystal load capacitance and the two load capacitors decide most of what is left, and they are the first thing to check when a build comes back running fast.
The oscillator fail flag, and what it cannot tell you
The fail flag is set when the oscillator stops, which is what happens on a first build with no backup fitted or after a backup capacitor has drained during a long storage period. The host reads the flag at start up and knows the time registers cannot be trusted. The flag is latched, so the host has to clear it after it has reloaded the time, otherwise the same warning comes back on every boot.
What the flag cannot do is tell you how long the clock was stopped, and it cannot rebuild the time. That has to come from somewhere else, which in practice means a network time source, a GPS pulse, or a service visit on a product with no connectivity. On a data logger this is the argument for fitting the backup capacitor in the first place, because a clock that never stops never needs a resynchronisation step in the product code.
The stop bit is a separate control. It halts the oscillator under host control, which is useful when a board is being stored or shipped and the clock is not needed. A part left with the stop bit set draws less current but also stops counting, so the same care is needed on the way out of storage as on the way in.
Where it fits in a build
This is a clock for a board that already has a host processor, a 3.3 V rail and a reason to know the date: a meter, a controller, a data logger, an access panel, a piece of test equipment. It is not a fit for a product that needs sub second timestamps across a network, where a disciplined clock is the right answer, and it is not a fit for a product that runs from a rail below 2 V.
The part is also a steady repeat item on metering and building control bills of material, and those programmes usually run for a decade. For a clock that is quoted late in such a programme, the component lead time solutions page sets out the stock, date code and packaging checks we run before a date is promised.
The sourcing side of this part is covered on the date code management. For a clock that may sit in stores before it is fitted, that page covers how reel age is tracked and what we will not ship into a new build.
Questions we get on this code
Can the backup be a battery instead of a capacitor? Yes. The backup pin accepts either, and a non rechargeable cell gives a much longer hold time than a capacitor of sensible size. The trade is that the cell has to be replaced on a service interval, and a product that cannot be opened in the field is better served by the capacitor.
Does the crystal have to be a 32.768 kHz tuning fork type? Yes, that is what the oscillator is built for. A crystal with a different load capacitance will start, but the frequency will sit off the nominal point and the calibration range will be spent correcting an error that should not be there.
How accurate is the clock straight out of the box? Close enough to log a timestamp, not close enough to bill on. The pre-calibration figure is a typical number with a specified crystal, and a production line that wants a tighter guarantee has to trim each board against a reference.
Is the part a drop in replacement for older clocks in the same family? The pin functions follow the common part and the register set is compatible, so in most cases yes. The one item to confirm is the backup pin behaviour on the board being replaced, because some older parts expect a different charge path for the backup capacitor.