TPS22993RLWR
Product Specifications
TPS22993RLWR
Texas Instruments — Power Management & Signal Chain
Product Overview
The TPS22993RLWR 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
TPS22993RLWR quad channel 1.2 A load switch with GPIO and I2C control in the twenty pin WQFN
The TPS22993 is four load switches in one body, each rated 1.2 A, with inputs from 1.0 V to 3.6 V and a bias rail that runs from 4.5 V to 17.2 V so it can sit on a 2S, 3S or 4S battery stack. Control is either four GPIO pins or an I2C bus, and the device is characterised over minus 40 to 85 C.
It is the part a board uses when four small rails have to be switched and the processor has run out of pins, or when the firmware wants to read back what each switch is doing. The I2C interface carries the enable bits and the status, the four ON pins still work as direct enables, and the ADD pins set the bus address so several of these can sit on one bus. The switches are plain low resistance FETs with no current limit, which is the point: they are meant to switch a defined load, not to protect it.
Key parameters taken from the TPS22993RLWR document
| Channels and current | Four independent switches, 1.2 A continuous per channel |
| Input range | 1.0 V to 3.6 V on each VIN, which covers a single cell rail and a 3.3 V rail |
| Bias range | 4.5 V to 17.2 V on VBIAS, which is what makes it usable on a 2S to 4S stack |
| Logic supply | VDD from 1.62 V to 3.6 V, the rail that sets the I2C and GPIO levels |
| On resistance | 15 mOhm typical and 20 mOhm maximum at 25 C, 22 mOhm maximum over the full temperature range |
| Quiescent current, four channels | 16.7 microamp typical and 30 microamp maximum at a 7.2 V bias |
| Quiescent current, one channel | 7.7 microamp typical and 15 microamp maximum, so the idle channels cost little |
| Shutdown current | Below 6 microamp with all four channels off, as printed in the feature table |
| Control | Four GPIO enable inputs plus an I2C interface for the same enables and for status |
| Addressing | ADD1 to ADD3 pins set the bus address, so several devices can share one bus |
| Interface speed | The I2C timing table covers the standard and fast modes and the document prints the bus timing separately from the switch timing |
| Package | WQFN-HR RLW twenty pins |
| Packing | 3000 pieces on a large reel, moisture level 1, marking 22993P |
| Temperature | Minus 40 to 85 C ambient as printed in the recommended operating conditions |
Every row above is read from the vendor document, not from a distributor summary. Where a window is printed, the window is what a production test should use, and the typical value is only a starting point for a first article.
Where the numbers matter in a real build
The bias rail is the unusual part of this device and it is where most of the questions come from. The switches themselves work from 1.0 V to 3.6 V, but the gate drive comes from a separate bias pin that tolerates up to 17.2 V. On a battery product the bias comes from the stack, and the switches then pass a low rail that may come from a buck converter running off the same stack. That arrangement is fine as long as the bias is valid before an enable is asserted and as long as the bias never falls below the input plus the gate headroom, which on a discharging battery is the case that a protection circuit has to catch.
Two control paths for the same four switches sounds redundant until a firmware problem is considered. The GPIO pins are hardware enables that work with no bus traffic, so a processor that hangs cannot leave a rail stuck on if the pins are driven from a supervisor. The I2C interface is what the firmware uses to read status and to change enables in normal operation. A board that wires only the bus and leaves the GPIO pins floating has a rail whose state depends on the bus, and that is the arrangement that fails when the bus is held low by a stuck device.
The quiescent current is quoted per channel state, which is more useful than a single number. With all four channels on the bias current is 16.7 microamp typical at 7.2 V, and with one channel on it is 7.7 microamp, so an idle channel adds about 3 microamp. The number that matters for a battery product is the shutdown current, printed below 6 microamp with everything off, and the VDD rail current, which is separate and small. None of these are large, but they add up across a board with several of these devices, and the maximum column is what belongs in the budget.
There is no current limit per channel. Each switch is a FET with a controlled gate, and a short on one output is limited by the source and by the die temperature. On a 1.2 A channel with a 22 mOhm maximum resistance the dissipation at full load is only 32 mW, so a short that pulls an amp through the switch is the case that heats the die. The document gives the thermal information for the twenty pin body and the application section explains the derating, and a product that has to survive a permanent short on a peripheral connector needs an external element rather than relying on these switches.
- Bias rail behaviour during a battery discharge, because the switches need headroom above the input.
- The state of the four GPIO enables during reset, since a floating pin leaves a rail undefined.
- The I2C address straps, because two devices on one bus with the same address is a production failure that only shows at test.
- Per channel loading, since one channel at 1.2 A and the others idle is a different thermal case from four channels at 400 mA.
- Shutdown current of the whole device, which is the figure the standby budget needs.
- The 20 pin land pattern and the exposed pad, which is the only thermal path for the quad die.
Which projects this part fits
This part fits battery powered instruments that switch four sensor or peripheral rails from a stack, notebook and tablet boards that switch a set of low voltage rails under firmware control, and any product where the processor pin count is the constraint. It also fits telecom and industrial boards where a supervisor has to be able to force rails off independently of the firmware, because the GPIO enables work without the bus.
Questions that come up before a purchase order is released
Can the switches be used above 3.6 V?
No. The input range stops at 3.6 V and the absolute maximum on the input pins is 4 V. A 5 V rail needs a different member of the family or an external divider arrangement, and the bias pin cannot be used to lift that limit.
What happens if the bias rail drops?
The gate drive falls with it and the on resistance rises. Below the minimum bias the switches are no longer characterised, and the recommended operating conditions set 4.5 V as the floor. A battery product should treat a bias below that as a protection event.
Do the GPIO pins override the I2C bits?
Both act on the same enable, and the document describes how the two are combined. The practical rule is to pick one path per rail in the firmware and to leave the other in a known state, otherwise a bus write and a pin change can fight each other.
Is there a current limit?
No. The four channels are plain switches. Overcurrent is limited by the source and by the thermal behaviour of the die, and a product that needs a defined limit has to add one.
How many devices fit on one bus?
The ADD1 to ADD3 pins give eight addresses, so up to eight of these can share a bus without a multiplexer. The address pins have to be strapped, not left floating, because a floating strap reads as an undefined level.
How SUPERB handles this line
We supply the TPS22993RLWR on original 3000 piece reels with the marking 22993P recorded at goods in, and we can quote the lower voltage members of the quad line when a board does not need a 17 V bias. The power management cross reference page lists the quad switches by channel count, current and bias range, kitting service is how a mixed set of these devices is delivered against one build, and incoming inspection plan covers the address strap and marking checks we run before a reel goes to the line.