LTC2642IMS-12#TRPBF
Product Specifications
LTC2642IMS-12#TRPBF
Analog Devices — Analog & Mixed-Signal ICs
Product Overview
The LTC2642IMS-12#TRPBF from Analog Devices is a high-performance analog / mixed-signal IC providing precision signal acquisition, conversion, and conditioning. Analog Devices (ADI) is a world leader in high-performance analog, mixed-signal, and DSP ICs for instrumentation, healthcare, automotive, and communications.
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-resolution data conversion with excellent SNR and THD
- Wide dynamic range and high linearity across operating conditions
- Integrated voltage reference with low temperature drift
- Flexible digital interface — SPI, I2C, or parallel options
- Industrial/automotive temperature range with guaranteed specs
- Low power consumption with power-down modes
Key Specifications
| Resolution | bits |
| Channels | — |
| Sample Rate | kSPS |
| Interface | — |
| Supply Voltage | V |
| Operating Temp | °C |
| Package | — |
| INL | LSB |
| SNR | dB |
| THD | dB |
Applications
LTC2642IMS-12#TRPBF: A 12-Bit DAC With No Output Buffer, and Why That Is the Point
The LTC2642IMS-12#TRPBF is the 12-bit member of the LTC2641 and LTC2642 families of unbuffered voltage-output converters. It runs from a single 2.7 V to 5.5 V supply, guarantees monotonic operation over temperature, draws 120 uA, settles to 16 bits in 1 us, holds glitch impulse to 0.5 nV seconds, accepts a reference from 2 V to the supply rail, and puts a 0.5 nV second class of disturbance on the output rather than the larger glitch of a buffered stage. The serial interface is three wire and runs to 50 MHz. The LTC2642 adds matched scaling resistors so an external precision amplifier can produce a plus and minus full scale around a midpoint, and it powers up at midscale, while the LTC2641 powers up at zero. Both come in 10-lead MSOP and 3 mm by 3 mm DFN packages.
The unbuffered output is the reason this family exists. There is no internal amplifier to add offset, drift and settling of its own, so the converter's own linearity is what the board sees, and the 120 uA supply current does not grow with a buffer's quiescent draw.
A 12-bit code on 4.096 V is 1 mV, and the load is part of the error
Twelve bits over 4.096 V gives exactly 1 mV per code. On a 2.5 V reference it is 610 uV. The output is a switch network with no amplifier after it, so any load current flows through the internal switch resistance and appears directly as a gain error. A 60 kohm load at 4.096 V draws 68 uA, which is more than half the converter's own supply current, and the data sheet's guaranteed 60 kohm drive figure is the boundary of that behaviour rather than a comfortable operating point. A 10 kohm load draws 410 uA and turns the switch resistance into a visible fraction of a code.
That is why the part is usually followed by an amplifier, and then the amplifier's offset becomes the error budget. At 1 mV per code, an amplifier with 1.25 mV of offset costs more than one code before the converter contributes anything, so the follower in this position has to be a chopper-stabilised type with microvolt offset. The saving of the unbuffered output is only real when the stage after it is chosen with the same care.
A midscale power-up is 2.048 V on a 4.096 V span
This order code powers up at midscale, so the output sits at half of the reference until the first valid write. In a gain trim or a bias circuit that is not a safe state; it is half the range driving whatever follows on every power cycle, including the cycles that happen while the processor is still booting. The CLR pin can be pulled low to force midscale, which is the same value, so the only way to reach zero at power-up in this family is the LTC2641 rather than the LTC2642. A buyer who orders by family name and expects a zero start gets the opposite behaviour.
The matched resistors in the LTC2642 are what make a bipolar span possible without a second converter. With an external amplifier and the internal resistors, the output becomes plus and minus the reference around a midpoint, and the resistor matching, not the amplifier, sets the symmetry. The settling time of 1 us to 16 bits supports an update rate near 1 MHz, and the 50 MHz interface is far faster than that, so the converter and not the bus is the limit.
| Resolution | 12 bit in this order code, 14 and 16 bit in the same family |
| Code size | 1 mV at 4.096 V, 610 uV at 2.5 V, 76.3 uV at 5 V for 16 bit parts |
| Output | Unbuffered, guaranteed to drive 60 kohm, output current adds gain error |
| Settling | 1 us to 16 bits, so about 1 MHz of useful update rate |
| Glitch | 0.5 nV seconds, no output amplifier to add its own step |
| Supply | 2.7 V to 5.5 V single, 120 uA, 0.6 mW at 5 V, 1.05 Ah over a year |
| Reference | 2 V to the supply rail, output swings 0 V to the reference |
| Bipolar option | Matched scaling resistors in the LTC2642, external amplifier required |
| Power-up state | Midscale in this part, zero in the LTC2641 |
| Interface and package | Three wire to 50 MHz, 10-lead MSOP and 3 mm by 3 mm DFN |
Two questions buyers ask
Is this a drop-in for the LTC2641?
No. The power-up state differs, midscale against zero, and the pinout differs because the LTC2642 brings the scaling resistors out to a tenth pin. A board wired for the LTC2641 has no place to land those resistors and a board wired for the LTC2642 expects them. The converter core is the same, the pin function map is not.
Can it drive an ADC reference pin directly?
No. A reference input draws a charge spike at every conversion and is not a 60 kohm load, so the unbuffered output develops a disturbance at the conversion rate. The part that belongs in that position is a reference with an output amplifier rated for the transient, and the converter here belongs on a setpoint that an amplifier follows.
Ordering and supply
The IMS-12 suffix is a 12-bit part in the MSOP body on the industrial temperature grade, and TRPBF is tape and reel, lead free. The same body at 16 bits is a different order code with a code size four times finer, so a line item that names only the package and the family can deliver a converter with the wrong weight per bit. The power-up state and the presence of the scaling resistors are the two fields worth writing on the purchase order next to the part number.
We source this family for production and for repair lots. Start at the analog mixed signal ICs family page, compare a candidate against the ADC and DAC cross reference, and where a converter has to go into a board that is already built, the checks in how to verify a drop in replacement put the power-up state and the pin map ahead of the socket fit.
Request Quote — In Stock, 100% New & Original, Full Warranty