TPS22976DPUR
Product Specifications
TPS22976DPUR
Texas Instruments — Power Management & Signal Chain
Product Overview
The TPS22976DPUR 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
TPS22976DPUR dual channel 6 A per channel load switch with 14 mOhm on resistance in the fourteen pin WSON
The TPS22976 puts two independent 6 A load switches in one fourteen pin WSON body measuring 2.00 mm by 3.25 mm, each with its own enable, each with a typical on resistance of 14 mOhm at a 5 V bias. The family also contains the TPS22976A with a different quiescent current and the TPS22976N without the output pull down, so the suffix on the order code decides the behaviour.
Two channels in one body is the reason this part exists. A board that switches a 3.3 V peripheral rail and a 1.8 V memory rail can do it with one component instead of two, and the two channels can be sequenced by driving ON1 and ON2 from separate GPIO lines, or tied together when both rails have to come up at once. The trade is that both channels share a die and a thermal pad, so the dissipation of the pair adds up in one place.
Key parameters taken from the TPS22976DPUR document
| Channels and current | Two independent switches, 6 A continuous per channel |
| Input range | 0.6 V to VBIAS on each input, with VBIAS between 2.5 V and 5.7 V |
| On resistance at 5 V bias | 14 mOhm typical at the 200 mA test current, 18 mOhm maximum at 25 C, 22 mOhm to 85 C and 23 mOhm to 105 C |
| On resistance at 2.5 V bias | 18 mOhm typical, which is the row for a bias rail at 2.5 V or 3.3 V |
| Quiescent current, both channels | 37 microamp typical and 49 microamp maximum at a 5 V bias with both switches on |
| Quiescent current, one channel | 35 microamp typical and 43 microamp maximum, so the second channel costs only a few microamp |
| Quiescent current on the A version | 85 microamp typical and 106 microamp maximum with both channels on, which is the price of its faster response |
| Shutdown current | Sub microamp typical on VBIAS and 1.4 to 3.4 microamp maximum on VIN depending on the rail |
| Thermal shutdown | The die turns both switches off above the threshold and back on after the junction falls about 20 C |
| Output pull down | 230 Ohm typical and 280 Ohm maximum per channel on this version, absent on the N version |
| Rise time | 1770 microsecond typical with 1000 pF on CT at 5 V, 285 microsecond with a 0.6 V input |
| Turn on and delay | 2390 microsecond turn on and 620 microsecond delay at 5 V with the same CT capacitor |
| Package | WSON DPU fourteen pins, 2.00 mm by 3.25 mm, with exposed pad |
| Packing | 3000 pieces on a large reel, NIPDAU finish, moisture level 1 with unlimited floor life, marking 22976 |
| Temperature | Minus 40 to 105 C in the addendum |
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
One pad for two channels changes the thermal arithmetic. Each channel at 6 A and 23 mOhm dissipates 828 mW, so both loaded channels put 1.66 W into a body that is barely larger than the single channel part. The pad has to carry that into copper, and the copper under the middle of the body is shared. A board that loads both channels at the same time needs a plane under the part with vias, and it should measure the case temperature on the first article rather than trusting a calculation. A board that loads one channel heavily and the other lightly is in the same position as two single switches, with the caveat that the thermal shutdown now turns both channels off.
The A version is not a faster version of the same part. Its quiescent current is 85 microamp typical against 37, which is more than double, and the reason is that it keeps more of its internal circuits alive. For a mains powered board that difference is irrelevant. For a battery product that leaves the switches enabled in standby, it is a choice that has to be made with the current budget in front of it, and the two versions are not interchangeable on a bill of materials without that check.
The independent enables are the feature that most boards use, and the sequencing question is where they get it wrong. Each ON pin is a high impedance input with a threshold of about 1.1 V, so a GPIO that goes high impedance during reset can leave a channel in an undefined state until the pin is driven. A pull down on each ON pin is the cheap answer, and it also stops a channel from coming up when the processor is held in reset. The two channels share a bias rail, so the bias has to be valid before either enable is asserted, and the recommended operating conditions print the bias range but not a sequencing rule, which means the rule comes from the application section.
The 14 mOhm typical figure is measured at 200 mA and at 5 V bias, and it is the best number in the document. At the 6 A the part is rated for, the resistance is closer to the 18 to 23 mOhm maximum rows, and at a 2.5 V bias the typical moves to 18 mOhm. A voltage drop budget that uses 14 mOhm on both channels at 6 A with a 3.3 V bias will be about 40 per cent optimistic. The right way to build the budget is to take the maximum row for the bias the board uses and the hot corner, then check the total against the rail tolerance.
- Both channels loaded at once, since the dissipation adds inside one body.
- Which version of the family the order code carries, because the A version draws more than twice the quiescent current.
- The bias rail and the resistance row it implies, since 14 mOhm only applies at a 5 V bias.
- Pull downs on both ON pins, so a floating GPIO cannot enable a rail during reset.
- The CT capacitor per channel, which sets the inrush each output sees.
- The 14 pin land pattern, which is not the same as the eight pin body used by the single channel members.
Which projects this part fits
This part fits boards that switch two rails of the same product together: a 3.3 V and a 1.8 V pair on a processor board, a pair of peripheral ports on an industrial gateway, or two sensor rails on an instrument. It also suits redundant supplies where each channel feeds a separate load and the firmware wants to shed one of them under thermal pressure, because the two enables are independent and the thermal circuit protects both.
Questions that come up before a purchase order is released
Does one channel shutting down affect the other?
The thermal shutdown turns both switches off, because they share a die. An overcurrent on one channel that heats the die enough will take the other channel down with it. A board that cannot tolerate that has to use two single channel parts.
Can the two channels be paralleled for 12 A?
The vendor does not support it. Two switches in parallel do not share current evenly, and the internal thermal circuit sees the sum. A 12 A load needs a single switch rated for it or an external FET arrangement.
What is the difference between the plain, A and N versions?
The plain version has the output pull down and 37 microamp typical quiescent current, the N version has no pull down, and the A version has a higher quiescent current with a different internal behaviour. The pin out is the same, so a substitution has to be checked against the current budget and the power down behaviour.
How are the two rise times set?
There is one CT pin per channel on the fourteen pin body, and each sets its own ramp. A board that needs both rails to come up together uses the same capacitor on both and drives both enables at once, and a board that needs a sequence uses different capacitors and different enable timing.
Is the part available on a small reel?
The large reel row is 3000 pieces and the small reel row is 250, both in the same DPU body. A prototype build can take the 250 piece reel without buying a full production quantity, which is the practical answer for a first article.
How SUPERB handles this line
We supply the TPS22976DPUR on original 3000 piece reels and can split to the 250 piece reel for a prototype, with the marking 22976 recorded on the packing list so the version can be confirmed at goods in. The power management cross reference page lists the dual and single members side by side, BOM consolidation covers the case where several boards in one family use different members of this line, and date code management is how we keep a mixed set of reels traceable through a build.