Contact Us

TPS22994RUKR

Brand: Texas Instruments
Texas Instruments TPS22994RUKR is a power management component designed for AI Servers & Low-Altitude Economy (Drone and UAV) and related applications. In stock at Superb Automation — request a quote for competitive pricing and fast delivery.
quote now

Product Specifications

TPS22994RUKR — Power Management & Signal Chain | Superb Automation

TPS22994RUKR

Texas Instruments — Power Management & Signal Chain

Product Overview

The TPS22994RUKR 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 VoltageV
Output VoltageV
Output CurrentA
Switching FrequencykHz
Efficiency%
Quiescent CurrentµA
Topology—
Package—
Operating Temp°C
Protection—

Applications

Industrial Automation Automotive Electronics Telecom Infrastructure Server & Data Center Power Medical Devices Consumer Electronics

Request Quote — In Stock, 100% New & Original, Full Warranty

TPS22994RUKR quad channel 1 A load switch with GPIO and I2C control in the twenty pin WQFN

The TPS22994 is the single cell member of the quad load switch line: four switches rated 1 A each, inputs from 1.0 V up to either the bias minus one volt or 3.6 V, and a bias rail from 2.7 V to 17.2 V that lets the same part sit on a one cell or a four cell stack. It is controlled by four GPIO pins or by an I2C interface, and it is characterised over minus 40 to 85 C.

The difference to its sibling is the bias range and the resistance. The bias starts at 2.7 V, so the part works from a single lithium cell or a 3.3 V rail without a boost, and the on resistance rises to about 41 mOhm typical, which is the price of the lower gate drive. For the 1 A loads this part is meant for, that is 41 millivolt of drop, which is acceptable on a peripheral rail and not acceptable on a core rail.

Key parameters taken from the TPS22994RUKR document

Channels and currentFour independent switches, 1 A continuous per channel
Input range1.0 V to 3.6 V, and never above the bias minus 1 V when the bias is below 4.6 V
Bias range2.7 V to 17.2 V on VBIAS, which covers a single cell up to a four cell stack
Logic supplyVDD from 1.62 V to 3.6 V for the I2C and GPIO levels
On resistance at 7.2 V bias40.5 mOhm typical and 50 mOhm maximum at 25 C, 58.5 mOhm maximum over temperature
On resistance at 5.2 V bias41.5 mOhm typical at a 1.8 V input, rising to 65 mOhm maximum over temperature
On resistance at the low bias endThe 3.3 V input row at a 5.2 V bias prints 60.4 mOhm typical and 71 mOhm maximum, which is the worst case in the table
Quiescent current, four channels21.1 microamp typical and 33.6 microamp maximum at a 7.2 V bias
Quiescent current, one channel11.3 microamp typical and 22.5 microamp maximum, so each idle channel costs about 3 microamp
Logic rail current1.2 microamp typical and 1.9 microamp maximum on VDD with no bus traffic
Shutdown currentBelow 7 microamp with all four channels off, as printed in the feature table
ControlFour GPIO enables plus I2C, with ADD1 to ADD3 setting the address
PackageWQFN RUK twenty pins
Packing3000 pieces on a large reel, NIPDAU finish, moisture level 1 with unlimited floor life, marking 22994

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 input ceiling is relative to the bias, and that is the rule that catches boards out. The recommended operating conditions print two rows: with a bias below 4.6 V the input may not exceed the bias minus 1 V, and with a bias at or above 4.6 V the input may go to 3.6 V. A board that runs the bias at 3.3 V and switches a 3.3 V rail is therefore outside the recommended conditions, because the input would sit above the bias minus one volt. The usual fix is to take the bias from a higher rail on the same board, or to choose the sibling part whose bias floor is 4.5 V and feed it from the stack.

The resistance numbers are printed against the bias voltage as well as the input, and the spread is wide. At a 7.2 V bias the typical is 40.5 mOhm across the whole input range, but at a 5.2 V bias with a 3.3 V input the typical rises to 60.4 mOhm and the maximum over temperature reaches 71 mOhm. On a 1 A load that is 71 mV, which is a fifth of a 3.3 V rail's tolerance. A voltage drop budget built on the 41 mOhm headline will be wrong by 70 per cent in that corner, and the corner is the common one on a 3.3 V board.

Two control paths give the firmware an escape route. The four ON pins are hardware enables that work with no bus traffic, and the I2C interface carries the same enables plus status. A supervisor that pulls the ON pins low during a reset can force all four rails off even if the bus is stuck, which is the behaviour a product needs when the processor hangs with a rail enabled. The address straps have to be driven, not left floating, because a floating ADD pin reads as an undefined level and two devices can then answer the same address on a shared bus.

The quiescent current is quoted for four channels on and for one channel on, and the difference is the cost of an idle channel. At a 7.2 V bias the four channel figure is 21.1 microamp typical and the single channel figure is 11.3 microamp, so each extra enabled channel adds about 3 microamp. That is small, but the maximum column is 33.6 microamp against 21.1, and a battery product should budget the maximum. The VDD rail is separate and costs about 1.2 microamp with no traffic, which is what a standby budget should use.

  • The bias voltage and the input voltage together, because the input ceiling depends on the bias.
  • The resistance row for the bias the board actually uses, since 41 mOhm only applies at a 7.2 V bias.
  • Address strap values, because two devices on one bus with the same address is a test failure that only appears at the end of the line.
  • The state of the ON pins during reset, since a floating pin leaves the rail undefined.
  • Load per channel, because one channel at 1 A and three idle is a different thermal case from four at 250 mA.
  • Shutdown current of the whole device, which is the figure a standby budget needs.

Which projects this part fits

This part fits single cell battery products that switch several small rails, wearable and handheld instruments, and any 3.3 V board that needs four switches under firmware control without a higher bias rail. It also fits boards that already use the 20 pin land pattern of the sibling part and need a lower bias floor, since the pin out is the same and only the silicon inside changes.

Questions that come up before a purchase order is released

Why is the input limited to the bias minus one volt?

Because the gate drive comes from the bias and the switch needs headroom to stay in its low resistance region. Below a 4.6 V bias the input must stay a volt below it. The document prints this as two separate rows in the recommended operating conditions, and it is the first thing to check on a 3.3 V board.

Can this part replace the higher bias sibling?

Only after the bias and resistance checks. The pin out is the same, the current rating is lower at 1 A against 1.2 A, and the resistance is higher. A substitution in either direction is a controlled change on the bill of materials, not a drop in.

Is there a current limit per channel?

No. The switches are plain FETs. A fault current is limited by the source and by the die temperature, and a product that needs a defined limit has to add one externally.

How many devices fit on one bus?

Eight, using the three address pins. The straps must be tied to a defined level, and the same eight addresses are available on the sibling part, so mixing the two on one bus needs a strap plan.

What is the advantage over four discrete switches?

Board area, one bias decoupling network, one bus address and one set of enables. The cost is that a thermal event on one channel affects the shared die, and that a single part failure takes all four rails down.

How SUPERB handles this line

We supply the TPS22994RUKR on original 3000 piece reels with the marking 22994 checked at goods in, and we can quote the higher current sibling when a board has a stack rail available for the bias. The power management cross reference page lists the quad switches by bias range and current, one stop BOM sourcing is how a customer sends one list and gets the switches, the passives and the connectors together, and incoming inspection plan covers the marking and address strap checks we run before a reel is released.