TMS320F28377DPTPQ
Product Specifications
TMS320F28377DPTPQ
Texas Instruments — Power Management & Signal Chain
Product Overview
The TMS320F28377DPTPQ 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
TMS320F28377DPTPQ dual core C28x controller at 200 MHz in the automotive 176 pin HLQFP
The TMS320F28377DPTPQ is the automotive grade member of the F2837xD family in the 176 pin HLQFP body. It carries two C28x cores at 200 MHz, each with a single precision floating point unit, a trigonometric math unit and the VCU-II complex maths extension, plus two control law accelerators that also run at 200 MHz and execute independently of the main cores. The memory is one megabyte of flash, which is 512 KB per core, and 204 KB of RAM split into 72 KB of local shared memory, 128 KB of global shared memory and a 4 KB message RAM used for inter processor communication.
The analog side is what separates this generation from the earlier fixed point parts: up to four 16 bit converters at 1.1 MSPS each, which gives 4.4 MSPS of system throughput when all four run, and the same blocks can be switched to 12 bit mode for 3.5 MSPS each and 14 MSPS of system throughput. Each converter has a single sample and hold, differential inputs in 16 bit mode and single ended inputs in 12 bit mode, and the post processing block does offset calibration, setpoint error calculation and trip generation without waking a core. Eight windowed comparators with 12 bit references and three buffered 12 bit outputs complete the analog chain.
Key parameters taken from the TMS320F28377DPTPQ document
| Cores | two TMS320C28x at 200 MHz, FPU, TMU type 0 and VCU-II on both |
| Accelerators | two CLA type 1 at 200 MHz, independent instruction stream |
| Flash | 1MB, 512 KB per core, ECC protected |
| RAM | 204 KB total: 72 KB local shared, 128 KB global shared, 4 KB message RAM |
| Converter | four 16 bit converters at 1.1 MSPS each, 4.4 MSPS system, or 12 bit at 3.5 MSPS each and 14 MSPS system |
| Comparator and DAC | eight windowed comparators with 12 bit DAC references, three buffered 12 bit DAC outputs |
| PWM | 24 channels, 16 high resolution channels, high resolution on both A and B of eight modules, dead band on both standard and high resolution |
| Capture and encoder | six eCAP, three eQEP, eight sigma delta filter channels with two parallel filters each |
| Logic | configurable logic block with four tiles |
| Memory interface | two EMIF units with ASRAM and SDRAM support, USB 2.0 with MAC and PHY, 12 pin uPP |
| Serial | two CAN, three SPI to 50 MHz, four SCI, two I2C, two McBSP |
| GPIO | 97 shared pins on the 176 pin body against 169 on the 337 ball body |
| Supply | 1.2 V core, 3.3 V input output |
| Order code facts | TMS320F28377DPTPQ, HLQFP PTP 176 pins, NIPDAU, Level-3-260C-168 HR, minus 40 to plus 125 C, Active, AEC-Q100 qualified |
The table 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 the first article.
Where the numbers matter in a real build
The two cores are not symmetric in the way a general purpose processor is. CPU1 boots, owns the flash pump and the security configuration, and releases CPU2, so the split has to be settled at boot rather than at run time. The shared RAM is the only practical channel between them and it needs a protocol with an owner bit per buffer, because two cores writing the same 128 KB block without arbitration is the most common way this family produces a fault that cannot be reproduced on a bench. The message RAM exists for short commands and it is small on purpose.
The 16 bit converter mode is differential, and that changes the front end rather than the software. A board that was laid out for single ended signals with a shared ground reference will show offset and noise in 16 bit mode that never appeared in 12 bit mode, because the negative input now has to be driven and the common mode range is bounded. The document prints the differential input range and the common mode limits per channel, and the first article should measure both, plus the sampling switch settling with the actual source impedance of the sensor, not with a signal generator at 50 ohm.
The automotive letter in the code is a process statement, not just a temperature. The Q suffix means AEC-Q100 qualification at minus 40 to plus 125 C, and it comes with the level 3 moisture class at 168 hours of floor life. A 176 pin HLQFP at that body size absorbs moisture during storage, and a part that sits on a shelf outside the bag for a week will need a bake before reflow. The safe production rule is to open the bag at the line, reflow within the printed floor life, and record the open time in the traveler so a second shift cannot restart the clock.
The high resolution PWM channels are the reason this part appears in charger and inverter control. Sixteen of the 24 channels carry the high resolution extension, and the document is specific that high resolution is available on both the A and B outputs of eight modules, which matters when a complementary pair has to stay matched. The dead band unit works on both standard and high resolution outputs, so the dead time does not lose resolution, and the trip zone path can shut the outputs down without core intervention. A first article that only checks the duty at fifty percent will not see the resolution at all.
- Both core rails at the pin under full load, and the sequencing order between 1.2 V and 3.3 V.
- Inter processor message latency and the shared RAM arbitration behaviour under simultaneous access from both cores.
- Differential input offset and noise on each 16 bit channel with the real sensor impedance connected.
- High resolution duty linearity at two percent and at ninety eight percent on a complementary pair.
- Trip zone response time from comparator edge to gate driver input, measured with the core idle.
- Moisture sensitivity handling record: bag open time, bake if any, and reflow profile against the level 3 window.
Which projects this part fits
This family is the control device in traction inverter and motor control work, in medium and short range radar signal chains where the fast converter and the logic block do the timing, in large commercial HVAC drives, in automated sorting equipment, in solar and energy storage converters, and in automotive charger and pump controllers. For SUPERB customers the common thread is a product that needs two control loops at once, one at tens of kilohertz with a floating point transform and one at the switching frequency with tight dead time, on a board that also has to pass an automotive qualification.
Questions that come up before a purchase order is released
Is the PTPQ the automotive qualified row?
Yes. The Q letter is the AEC-Q100 option at minus 40 to plus 125 C, and the addendum prints this string as Active with a NIPDAU finish and level 3 moisture class. The non Q rows of the same body carry the same package and a narrower qualification, so the order string, not the package, decides what arrives.
How much flash is available to each core?
One megabyte in total, presented as 512 KB per core. The array is ECC protected and the security module has two zones, so a product can keep its own image and a third party image on the same die without one being able to read the other. That is the feature to check before letting a partner load code on the same device.
Why does the 176 pin body show fewer IO pins than the 337 ball body?
Because the larger body brings out more of the peripheral set. The document prints 97 shared GPIO on the 176 pin HLQFP against 169 on the 337 ball package. A port from one to the other is not a straight copy: the pin mux table has to be read per signal, and a peripheral that exists on the die may simply have no pin on the smaller body.
What tray quantity should the line expect?
The orderable block prints 90 pieces per JEDEC tray for this string, while the packaging table prints 40. Both numbers appear in the same document, so the carton label is what the incoming count should be checked against, and a line that plans feeders from the first number will be short on a 40 piece tray.
Can the two control law accelerators replace one of the cores?
They can carry a complete control loop, but they have no flash of their own and they are loaded by a core at boot, so they cannot be the only processor on the device. The usual split is the heavy maths on a CLA and the sequencing, communication and safety logic on a core, which leaves the second core free for the outer loop or for a protocol stack.
How SUPERB handles this line
We supply the TMS320F28377DPTPQ on original trays with the automotive letter and the moisture class verified on the label, and we handle it as a level 3 part from receipt to the line, including bake records where the bag has been open beyond the printed window. For a production program the useful conversation is about the annual volume and the grade, not the unit price, because the automotive row is the one that has to be booked ahead. Start from the MCU cross reference page when the question is which member of the family can ship first, and from the incoming inspection plan page when the concern is verifying what arrives.