ADS5545
Product Specifications
ADS5545
Texas Instruments — Power Management & Signal Chain
Product Overview
The ADS5545 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
ADS5545: 14 bit 170 MSPS converter in a 48 pin QFN with internal reference
ADS5545 is a 14 bit converter sampling at 170 MSPS in a 7 mm body, with 1.1 W total power and a stated 11.4 effective bits at a 70 MHz intermediate frequency in a QFN body of 48 pins. The ADS5545 samples one channel to 14 bit at up to 170 MSPS in a 48 pin QFN body of 7 mm by 7 mm, at 1.1 W total power. The document gives 74 dBFS of SNR and 85 dBc of SFDR at a 70 MHz intermediate frequency, with a minimum of 11.4 effective bits at the same point, and offers DDR LVDS or parallel CMOS outputs, programmable gain up to 6 dB and reduced power modes at lower sample rates. The internal reference needs no external decoupling capacitor.
Order code and package
| Order code | ADS5545 |
| Channels | one |
| Resolution | 14 bit, no missing codes |
| Maximum sample rate | 170 MSPS |
| Total power | 1.1 W |
| Output interface | DDR LVDS or parallel CMOS |
| Package | 48 pin QFN, 7 mm by 7 mm |
| Reference | internal, no external decoupling required |
| Gain | programmable up to 6 dB |
| Clock amplitude | down to 400 mV peak to peak |
| Power modes | reduced power at lower sample rates |
| Document | SLWS180C, November 2005, revised May 2007 |
Electrical parameters that decide the application
| SNR | 74 dBFS at a 70 MHz intermediate frequency |
| SFDR | 85 dBc at a 70 MHz intermediate frequency at 0 dB gain |
| ENOB | 11.4 bits minimum at a 70 MHz intermediate frequency |
| Power per sample | 6.5 nJ at 170 MSPS |
| Data rate at 14 bit | 2.38 Gbps at 170 MSPS |
| Sample and hold | internal |
| Output clock | programmable position, useful for capture timing |
| Reference decoupling | not required for the internal reference |
| Temperature | industrial range as stated by the vendor |
| Input drive | differential, internal sample and hold |
| Reduced power | lower rates draw less, per the mode table |
| Package sharing | 48 pin 7 mm body |
Points to check before this part goes on a BOM
Eleven point four effective bits minimum at a 70 MHz intermediate frequency is the honest number, and it is 2.6 bits below the 14 bit resolution. The gap is noise and distortion in the converter and its clock path, so a system that needs 13 effective bits cannot be built on this part no matter how the board is laid out.
The internal reference needs no external decoupling capacitor, which is unusual and useful: it removes a component and a noise injection path from a dense board. The trade is that the reference cannot be trimmed externally, so absolute gain accuracy depends on the internal reference tolerance and on calibration.
1.1 W at 170 MSPS is 6.5 nJ per sample, which is a third of a 16 bit converter at a similar rate. For an eight channel receive board that difference is the difference between forced air and convection, so the resolution requirement often decides the enclosure rather than the other way round.
Programmable output clock position is a capture timing feature rather than a datasheet line. Being able to move the output clock edge relative to the data removes the need for a delay line in the capture path, and on a board where the FPGA timing is tight it can be the item that makes the timing close.
Reduced power modes at lower sample rates mean the part can be run at 125 MSPS for a narrower band and still be the same order code. The mode is a register setting, so a single build can serve two bandwidth variants, which is useful when a product family spans two markets.
The 2.38 Gbps output at 170 MSPS is the capture constraint. A parallel CMOS output at that rate is 14 lines plus clock at 170 MHz, which is beyond most general purpose capture pins, so the LVDS path is the realistic choice and the FPGA needs a source synchronous LVDS receiver with the right delay.
For life cycle planning this is an older document, revised in 2007, and the part sits in a rate class that newer families have moved past. That makes it a candidate for a last time buy or a cross reference exercise, and the replacement has to be compared on effective bits at the intermediate frequency rather than on resolution alone.
The document dates from 2005 and was revised in 2007, so the timing table and the register map are the authority rather than any later application note. A capture interface written against a newer part in the same class usually needs the output clock position and the framing re-checked before it works with this converter.
Where this order code fits
This order code fits cellular infrastructure receive paths for 802.16d and 802.16e systems, software defined radio, power amplifier linearisation, high definition video digitisation and medical imaging receive channels. The 170 MSPS rate with 14 bit resolution is a common middle ground where a 125 MSPS part is too slow for the band and a 250 MSPS part costs too much power. Boards comparing it with a neighbouring part start from our ADC DAC Cross Reference page, and boards replacing a part that has gone end of life start from the Obsolete RF Part Replacement page.
Questions buyers ask
Does the internal reference need an external capacitor? No, the document states no external reference decoupling is required. That removes a component and a noise path from the board.
How many effective bits does it really deliver? 11.4 bits minimum at a 70 MHz intermediate frequency, against a 14 bit resolution. The gap is noise and distortion.
Can it run slower to save power? Yes, there are reduced power modes at lower sample rates, and the rate is set by the clock and the register, not by the order code.
SUPERB quotes this order code with packing and date codes, and the How to Verify a Drop in Replacement page covers how to qualify a replacement for a converter whose document dates from 2007.