Texas Instruments Data Converters: ADC & DAC ICs

A Texas Instruments ADC turns a real-world analog signal — voltage, current, temperature, RF — into the digital code a processor can read, and a TI DAC does the reverse. The portfolio runs from a simple 16-bit I²C ADC like the ADS1115 to a 6.4 GSPS RF-sampling converter, so the real task is matching architecture, resolution and sample rate to the signal in front of you, then protecting that performance through the layout.

This guide separates precision from high-speed converters, lists real TI parts with their headline numbers, and explains why the converter is rarely the part that limits your accuracy.

• SAR ADCs give low latency and per-sample conversion; delta-sigma ADCs trade latency for very high resolution.
• The ADS1115 is the go-to small 16-bit, 860 SPS, I²C ADC for sensor and instrumentation work.
• High-speed/RF-sampling parts such as the ADC12DJ3200 reach 6.4 GSPS and digitise RF directly.
• More bits do not help if the reference, clock jitter or driver amplifier is noisy. The signal chain sets the real accuracy.

What TI Data Converters Do: ADC and DAC Basics

Two numbers frame every converter: resolution (how many bits, and therefore how finely it splits the input range) and sample rate (how many conversions per second). A third, often decisive, number is effective number of bits (ENOB), which is the resolution you actually get once noise is included.

TI builds three broad ADC architectures. Successive-approximation (SAR) converters give a clean, low-latency sample per conversion and suit multiplexed, per-channel sensing. Delta-sigma converters oversample to reach 24-bit and higher resolution, ideal for slow, precise measurements like temperature, pressure and energy metering. High-speed and RF-sampling converters trade resolution for raw speed to capture wideband and radio signals directly. DACs mirror these as precision and high-speed families.

Precision vs. High-Speed TI ADCs: How to Choose

Start from the signal bandwidth and the accuracy you need, then pick the architecture. The table shows representative TI ADCs across the range.

TI ADC	Architecture	Key specs	Typical use
ADS1115	Delta-sigma, 16-bit	Up to 860 SPS, I²C, PGA ±256 mV to ±6.144 V	Compact sensor and instrumentation front ends
ADS127L11	Delta-sigma, 24-bit	Up to 400 kSPS wideband (1067 kSPS low-latency), 18.6 mW	Wideband precision measurement
ADS131M08	Delta-sigma, 24-bit	8-channel simultaneous sampling, 32 kSPS	Energy metering and power monitoring
ADC12DJ3200	RF-sampling, 12-bit	Up to 6.4 GSPS, JESD204 interface	Radar, software-defined radio, wideband test

The honest trade-off here is latency versus resolution. A SAR converter answers each sample immediately, which a control loop needs. A delta-sigma converter buys 24-bit resolution by oversampling and filtering, which adds group delay — fine for a temperature reading, a problem inside a fast feedback loop. Choose the architecture for the role, not the bit count.

TI DACs: Precision and High-Speed

On the output side, precision DACs such as the DAC80508 (a 16-bit, 8-channel device) set stable reference and control voltages, while high-speed DACs like the DAC38RF family synthesise wideband and RF waveforms over a JESD204 interface. The selection logic mirrors the ADC side: settling time and glitch energy matter for fast outputs, while DC accuracy and drift matter for slow reference outputs.

Driving and Powering a TI ADC: The Signal Chain

Here is the point most converter selections miss: the ADC is usually not what limits your accuracy. Three other things do — the driver amplifier, the voltage reference and the clock.

• The driver amplifier must settle to the ADC’s resolution within the acquisition window. A slow or noisy op amp caps your real sample rate and ENOB long before the converter does, which is why Texas Instruments amplifiers are part of every high-performance converter design.
• The reference sets the absolute scale. Reference noise and drift translate directly into measurement error, so a 24-bit ADC on a noisy reference may deliver only 18 effective bits.
• Clock jitter destroys SNR in high-speed converters. At high input frequencies, a few picoseconds of jitter sets a hard ceiling on signal-to-noise ratio no matter how good the ADC is.

Powering the converter cleanly closes the loop: a low-noise LDO from the Texas Instruments power management ICs line, often fed by a buck converter upstream, gives the quiet rail the ADC needs. Converters sit between Texas Instruments sensors on the input and TI processors on the output, where a JESD204 host ingests the high-speed data stream.

An instrumentation client once shipped a 24-bit design that delivered only about 18 effective bits. The ADC was blameless. Fixing the reference bypassing, adding a proper low-noise LDO and choosing an ADC driver that settled in time recovered roughly four bits. The converter datasheet promised 24; the system delivered what the rest of the chain allowed.

Data Converter Layout and Assembly Pitfalls

Layout protects the bits you paid for. The recurring mistakes:

1. Splitting the ground plane to “isolate” analog and digital. This usually breaks the return path and makes noise worse. A single, well-managed ground plane with thoughtful component placement almost always wins.
2. Skimping on reference and supply decoupling. Place bypass capacitors tight to the reference and supply pins; this is where lost ENOB hides.
3. Ignoring controlled impedance on high-speed links. JESD204 and LVDS lanes need controlled impedance — typically 100 Ω differential — routed per the IPC-2221 spacing and geometry guidance, or the eye closes.
4. Treating a high-pin-count converter BGA as a routine part. Verify its joints with X-ray and track MSL handling, the same as any fine-pitch device.

Common TI ADC and DAC Mistakes

1. Choosing resolution by the datasheet number and ignoring achievable ENOB.
2. Using a delta-sigma ADC inside a fast control loop that cannot tolerate its latency.
3. Driving a fast ADC with an amplifier that cannot settle in the acquisition window.
4. Powering a precision converter from a noisy switching rail with no LDO post-regulation.
5. Splitting ground planes and breaking the return path.

Frequently Asked Questions About TI Data Converters

What is the difference between a SAR and a delta-sigma ADC?

A SAR ADC converts one clean sample at a time with low latency, suiting multiplexed and control applications. A delta-sigma ADC oversamples and filters to reach very high resolution (24-bit and beyond) at the cost of latency, suiting slow, precise measurements.

Which TI ADC is best for a sensor project?

For most compact sensor and instrumentation work, the ADS1115 — 16-bit, up to 860 SPS, I²C, with a built-in PGA — is a popular default. For higher-resolution or faster precision measurement, a delta-sigma part like the ADS127L11 fits.

Why am I not getting the full resolution of my ADC?

Because the rest of the signal chain limits it. Reference noise, clock jitter and a driver amplifier that does not settle in time all reduce the effective number of bits. A 24-bit converter on a poor reference can deliver far fewer usable bits.

What is an RF-sampling ADC?

An RF-sampling ADC runs fast enough to digitise radio-frequency signals directly, replacing the traditional mixing stage. TI’s ADC12DJ3200 samples up to 6.4 GSPS and is used in radar, software-defined radio and wideband test.

Should I split the ground plane around a data converter?

Usually no. Splitting the plane tends to break the signal return path and increase noise. A single, continuous ground plane with careful placement of the analog and digital sections is generally the better approach.

Choose Your TI Converter, Then Protect the Bits

Pick the architecture from your bandwidth and accuracy needs, then design the driver, reference, clock and clean rail around it — that is where the real performance lives. Send your Gerber and BOM for a DFM review so impedance, decoupling and X-ray inspection are handled before the first build, and return to the Texas Instruments component hub to complete the signal chain. For impedance and spacing standards, see IPC.