Main page Knowledge graph
77 GHz radar SoC · baseband circuits

IF chain and ADC circuits

From the mixer output to digital samples: the high-pass filter that equalises range, the PGA and anti-alias filter, ΣΔ and SAR ADC design, decimation, key parameters, failure modes and the PCM/WAT items behind them. Includes an ADC SNR budget calculator.

00

IF chain and ADC circuits

IF · basebandADC

After the mixer, the beat signal lives at IF (roughly 0.1–20 MHz). The IF chain shapes it for the ADC: a high-pass filter removes leakage and very near reflections, a programmable gain amplifier sets the level, an anti-alias filter stops folding, and the ADC with its decimation filter delivers samples to the digital front end. The system view is in the signal chain; the RF side is on the RF front-end page.

Mixer + TIAI and QHPF1st–2nd orderPGA / VGA0–48 dBAAF2nd–4th orderADC driverCM, driveCT-ΣΔ ADC≈ 400–800 MS/sDecimationCIC + half-bandto DFEIF test DACEach Rx channel has an I and a Q copy of this chain; typical IF band 0.1–20 MHz (up to 40 MHz for short-range modes).
One IF chain per I and Q path of every Rx channel. The IF test DAC (pink, test-only) injects tones for BIST path C.

High-pass filter: equalising the range

In FMCW the IF frequency is proportional to range, and echo power falls as 1/R⁴. So close targets and Tx→Rx leakage arrive at low IF with enormous power, and far targets at high IF are weak. A 2nd-order HPF rises 40 dB per decade below its corner, which cancels the 40 dB per decade fall of the echo and flattens the spectrum. The ADC then needs far less dynamic range.

200-20-40-60-80-10010 k100 k1 M10 Mecho power ∝ 1/R⁴HPF gainat the ADCIF frequency (Hz), proportional to rangerelative level (dB)
Illustrative levels for a fixed target size. Below the 300 kHz corner (about 1 m for a 60 MHz/µs slope) the filter cancels the R⁴ law; above it, echoes fall as usual. Corners are programmable per mode.
f_IF = S · 2R / c S = chirp slope (Hz/s) R at the corner: R_c = f_c · c / (2S) e.g. f_c = 300 kHz, S = 60 MHz/µs → R_c ≈ 0.75 m 2nd-order HPF: |H(f)|² = (f/f_c)⁴ / (1 + (f/f_c)⁴)
  • Circuit: an RC high-pass at the TIA output (1st order) plus a second pole in the PGA feedback, or an active-RC Sallen-Key stage.
  • Fast settle: a switch shorts the resistor for the first microsecond of each chirp, so the DC step at ramp start dies quickly.
  • Trim: R or C banks set the corner per mode and absorb process spread (±20 % sheet R).

PGA, anti-alias filter and ADC driver

The PGA (circuit on the RF page) is set by the AGC so the strongest expected echo sits 6–10 dB below ADC full scale. The anti-alias filter must attenuate anything near fs ± BW that would fold into the band. A continuous-time ΣΔ ADC has an inherent anti-alias response, so the AAF can be a simple 2nd-order stage; a SAR needs a sharp 4th–6th order filter and a driver that can charge its sampling capacitor in time.

ADC architecture

ADC typeRate / bandwidthResolutionStrengthsWeaknesses
Continuous-time ΣΔ400–1000 MS/s, 10–40 MHz BW11–13 ENOBInherent anti-aliasing (relaxed AAF), resistive input easy to drive, low powerSensitive to clock jitter and DAC mismatch; loop stability over corners
SAR25–100 MS/s10–12 ENOBVery low power, mostly digital, scales with processNeeds a sharp AAF and a strong driver (switched-cap input); comparator noise
Pipelined / pipelined-SAR100–500 MS/s10–12 ENOBHigh speed at good resolutionHigher power, calibration of inter-stage gain
IF inΣ∫ loop filterorder L = 2–4Quantiser3–5 bit, fsDecimation÷ OSRN-bit words at 2·BWFeedback DACNRZ / RZdigital− (subtract)
A continuous-time ΣΔ loop. The loop filter has high gain in the IF band and the feedback forces the quantiser output to track the input there; quantisation noise is pushed to high frequency where the decimation filter removes it.
OSR = fs / (2·BW) Ideal SQNR (order L, N-bit quantiser): SQNR ≈ 6.02·N + 1.76 + (2L+1)·10·log10(OSR) − 10·log10(π^(2L) / (2L+1)) Jitter limit (sampling, sine at f_in): SNR_j = −20·log10(2π · f_in · σ_j) Thermal (kT/C for switched-cap input): v_n² = kT / C ENOB = (SNDR − 1.76) / 6.02

ADC budget calculator

OSR–
Ideal SQNR–
Jitter-limited SNR at band edge–
Thermal SNR (entered)–
–
Total SNR → ENOB–
Dominant limit–

The jitter formula is the sampling-clock limit for a sine at the band edge; in a CT-ΣΔ the feedback DAC edge jitter behaves similarly for NRZ DACs. Typical radar designs are thermal- or jitter-limited, not quantisation-limited, which is why the ADC clock comes from a clean PLL output.

Saturation detectors

A strong echo very close to the radar, Tx-to-Rx leakage, or another radar can drive a stage of the receive chain into clipping. Clipping spreads energy into harmonics and spurs over all range bins: weak targets are lost and ghost targets appear, and nothing in the FFT result says why. Saturation detectors flag it, so the AGC can lower the gain, the DSP can drop the affected chirp, and the safety monitor knows when the receiver is blind. Each stage clips at a different level, so each stage that can saturate first gets its own detector.

LNAMixer + TIAHPFPGAAAFADCDFEsamples → FFTPeak detRFWindowTIA outWindowPGA outOver-rangeADC codeBurstdigitalSaturation status registersticky flag + clipped-sample count, per Rx, per chirpAGC: lower PGA gainDSP: mark chirp invalidSafety: report if persistentPink: test and monitor hardware. Every stage that can clip gets a detector, because each one clips at a different input level.Flags are cleared at each chirp start; the counter shows how badly a chirp clipped, not only that it did.
Saturation detectors along one Rx channel. The flags are sticky for one chirp and come with a count of clipped samples.
DetectorWhereCircuitCatches
RF peak detectorLNA output (optional) or mixer inputDiode or square-law MOS detector + comparator, like the power detectorVery strong echoes or an interferer that compresses the LNA before anything at IF clips
Window comparatorTIA output (after the mixer)Two fast comparators on the differential output against ±V_th from a small DAC, with a latchTIA saturation from leakage or a target at very short range (below the HPF corner effect)
Window comparatorPGA / AAF outputAs above, threshold tied to the ADC input rangeGain set too high by the AGC; strong echo above ADC full scale
Over-range detectorADCSAR: output at max/min code. ΣΔ: quantiser at its extreme level for several clocks, or integrator-overload comparatorsADC clipping; ΣΔ loop overload (the loop is reset if it goes unstable)
Burst detectorDFE (digital)Magnitude or slope of samples above a threshold, with a sample counterInterference bursts from other radars; marks samples for zeroing (interference)
Window comparator: flag = (v_out > +V_th) OR (v_out < −V_th) V_th = k · V_lin,max , k ≈ 0.8–0.9 ADC over-range: count samples with |code| ≥ code_FS − margin (margin ≈ 1 dB below full scale) AGC action: clipped samples > N_max in a chirp → reduce PGA gain by one step (2–6 dB) for the next chirp

Detector specifications

ParameterTypicalWhy
Threshold80–90 % of the linear swing; ADC at −1 dBFSFlag before the stage compresses (P1dB), not after
Threshold accuracy±5 % after trimComparator offset and DAC error; trimmed with the IF test tone
Response time< 1 sample period (≈ 25 ns at 40 MS/s)Catch single-sample clipping and short interference bursts
Hysteresisa few % of thresholdAvoid chattering on noise near the threshold
Flag typeSticky per chirp + counterThe DSP reads it after the chirp; the counter tells how many samples clipped
Loading on the signal path< 0.1 dB, no visible kickbackComparator input capacitance and kickback must not distort the IF signal

Testing the detectors

  • Threshold accuracy: the IF test DAC (BIST path C) steps a tone up through the threshold; the amplitude where the flag sets is the measured threshold, which is trimmed and checked against limits.
  • RF detector: the RF loopback with a low attenuator code gives a known high level at the LNA (level plan).
  • Latent faults: at key-on, each detector is forced (test-mode override or a large test tone) to prove the flag, the counter and the clear logic work (key-on tests).
  • Production: threshold and response time are tested at wafer sort; a detector that never fires is otherwise invisible in normal tests.

Problems with saturation detection

ProblemEffectFix
Threshold drifts with temperature or offsetFalse flags (AGC lowers gain without need, losing range) or missed clippingTrim at test, use a bandgap-referenced DAC, check over temperature
Detector too slowShort interference bursts not seen; corrupted chirps usedFaster comparators, or rely on the digital burst detector
Comparator kickbackSpurs and distortion on the IF signalPre-amplifier or series resistor in front of the comparator
Flag not clearedEvery later chirp marked invalidClear on chirp start; key-on self-test of the clear logic
Only the ADC is monitoredTIA or PGA clips while the ADC is still in range, so clipping goes unseenDetector at every stage that can saturate first

Key parameters

ParameterTypicalSet by
HPF corner100 kHz – 1 MHz, programmableNear-range suppression vs short-range target loss; R/C codes
HPF settling after chirp start< 1–2 µs (fast-settle switch)Corner frequency; DC offset step at ramp start
PGA gain range / step0–48 dB, 1–2 dB stepsSee the VGA on the RF page
AAF2nd–4th order, f_−3dB ≈ 1.2× IF BWAlias rejection at fs − BW; relaxed for CT-ΣΔ
Input-referred noise (IF chain)a few nV/√HzTIA/PGA resistors and op-amp noise
In-band linearityHD3 < −60…−70 dBcOp-amp loop gain, output swing
Channel match (gain / phase)±0.5 dB / ±3° before calibrationR/C matching; corrected in DFE
ADC SNR / SNDR65–75 dB in the IF bandQuantisation, thermal, jitter
ADC SFDR> 70–80 dBcDAC mismatch, input switch linearity
Overload recoverywithin one chirpSaturation detectors + AGC

Decimation and the hand-over to the DFE

  • CIC filter first (no multipliers, decimates by 4–16), then half-band FIR stages (each ÷2), then a compensation FIR that flattens the passband droop.
  • Output: 16-bit complex samples (I + jQ) at 2× the IF bandwidth, per Rx channel, into the DFE and range FFT.
  • The DFE also removes residual DC, corrects I/Q gain and phase (coefficients from the loopback I/Q test), and flags ADC clipping for the AGC.

Issues and failure modes

IssueSymptomCauseFix / catch
HPF not settledGhost targets and raised noise at the start of each chirpCorner too low; DC step at ramp startFast-settle switch, skip first samples, check with IF tone (path C)
ADC overloadClipped spectrum, spurs across all range binsStrong near target or leakage above full scaleSaturation detector → AGC reduces PGA gain
AliasingTargets folded from beyond max rangeAAF too weak at a cornerFilter order, CT-ΣΔ inherent filtering
Jitter-limited SNRNoise floor rises with IF frequencyADC clock jitterClean clock from the PLL domain, not from digital
ΣΔ loop instabilityIdle tones or a noisy spectrum at large inputsExcess loop delay, coefficient drift at cornersLoop design margin, reset on overload
Channel mismatchAngle error, ghost in angleR/C mismatch, clock skew between ADCsLoopback calibration vector in the DFE
Spurs from digitalFixed-frequency tones in all channelsCoupling from clocks and data busesSeparate supplies, frequency planning, shielding

PCM / WAT items that matter

WAT itemAffectsWhy
Poly / diffusion sheet R and matchingHPF and AAF corners, PGA gain stepsRC time constants and resistor ratios
MIM / MOM capacitance and matchingFilter corners, ΣΔ integrator coefficients, DACLoop coefficients set by R·C products
Vth, gm (analog transistors)Op-amp gain-bandwidth and offsetLoop gain → linearity, settling
Flicker noise (1/f corner)Low-IF noiseNear-range noise floor
Ioff / junction leakageSwitched-cap nodes, HPF fast-settle switchLeakage droop and offset

Test: IF response, gain steps and ADC SNR/SFDR are measured on-chip with the IF test DAC and the FFT (BIST path C).

Values are typical for 77 GHz automotive radar SoCs and meant for planning; check your process, product and customer requirements.
01

Glossary of abbreviations

Abbreviations are underlined with dots the first time they appear in each section. Hover to see the meaning; click to jump here. Use Search (Ctrl+K) to find an abbreviation anywhere on the page.

Radar system & ADAS

Abbr.Stands forNote
FCWForward Collision WarningWarns of an imminent frontal collision; stage before AEB.
FCTAFront Cross-Traffic AlertFront corner-radar function at junctions and driveways.
RCTBRear Cross-Traffic BrakingBrakes automatically when RCTA detects cross traffic while reversing.
DOWDoor-Opening WarningWarns before a door is opened into an approaching cyclist or car.
VRUVulnerable Road UserPedestrians, cyclists and motorcyclists.
TTCTime To CollisionRange divided by closing speed; triggers warnings and braking.
LRRLong-Range RadarFront radar, ≈ 200–300 m, narrow field of view.
MRRMid-Range Radar≈ 80–160 m, ±40–60° field of view.
SRRShort-Range RadarCorner radar, ≈ 30–80 m, ±75–90° field of view.
NCAPNew Car Assessment ProgrammeConsumer safety ratings (Euro NCAP, C-NCAP, IIHS …).
GSRGeneral Safety Regulation (EU) 2019/2144Makes AEB and other assistance functions mandatory in the EU.
CPDChild Presence DetectionIn-cabin sensing (often 60 GHz radar) that detects a child left in the car.
ACCAdaptive Cruise ControlKeeps distance to the car ahead; uses long-range front radar.
ADASAdvanced Driver-Assistance SystemsThe vehicle functions the radar serves.
AEBAutonomous Emergency BrakingBrakes automatically when a collision is imminent.
BSDBlind-Spot DetectionCorner-radar function.
LCALane-Change AssistCorner-radar function that warns of fast approaching vehicles.
RCTARear Cross-Traffic AlertRear corner-radar function when reversing.
L2SAE automation Level 2 (L3 = Level 3)Partial (L2) and conditional (L3) driving automation.
FMCWFrequency-Modulated Continuous WaveA linear chirp; range comes from the beat frequency f_b = S·2R/c.
FoVField of ViewAngular coverage in azimuth and elevation.
RCSRadar Cross SectionTarget reflectivity in dBsm (car ≈ 10, pedestrian ≈ −10…0).
dBsmDecibels relative to 1 m²Unit of RCS.
EIRPEffective Isotropic Radiated PowerTx power + antenna gain; limited by regulations.
MIMOMultiple-Input Multiple-OutputN_tx × N_rx channels form a larger virtual array.
TDMTime-Division MultiplexingOne Tx per chirp; v_max falls by a factor of N_tx.
BPMBinary Phase Modulation0/180° codes across chirps separate the Tx signals.
DDMADoppler-Division Multiple AccessEach Tx gets a slow-time phase ramp and separates in Doppler.
DoADirection of ArrivalAngle estimation over the virtual array.
DBFDigital BeamformingFFT or matched steering over array channels.
CFARConstant False Alarm RateAdaptive detection threshold (CA = cell-averaging, OS = ordered-statistic).
FFTFast Fourier TransformRange, Doppler and angle FFTs build the radar cube.
SNRSignal-to-Noise Ratio
MUSICMUltiple SIgnal ClassificationSubspace super-resolution DoA algorithm.
ESPRITEstimation of Signal Parameters via Rotational Invariance TechniquesSuper-resolution DoA algorithm.
IAAIterative Adaptive ApproachSuper-resolution spectral / DoA estimator.
DBSCANDensity-Based Spatial Clustering of Applications with NoiseGroups detections into objects.
EKFExtended Kalman FilterNonlinear tracking filter.
UKFUnscented Kalman FilterNonlinear tracking filter using sigma points.
IMMInteracting Multiple ModelTracker that mixes several motion models.
HWAHardware AcceleratorOn-chip FFT / radar processing engine.
RTSRadar Target SimulatorEcho generator for HIL and EOL testing.
HILHardware-in-the-LoopTesting the real sensor against simulated scenarios.
EOLEnd-of-LineFinal production test of a module or vehicle.

RF & analog front end

Abbr.Stands forNote
HPBWHalf-Power BeamwidthAngle between the −3 dB points of an antenna or array beam.
PDNPower Delivery NetworkRegulators, planes, decoupling and package paths that supply the chip.
PIPower IntegrityKeeping supply voltage within limits under load steps and noise.
EMIElectromagnetic InterferenceUnwanted emission or pick-up of electromagnetic energy.
BCIBulk Current InjectionImmunity test (ISO 11452-4) that injects RF current into a cable harness.
TIMThermal Interface MaterialGap pad or paste between board or package and the heat sink.
RFRadio FrequencyHere the 76–81 GHz signal.
LOLocal OscillatorThe chirp signal that drives the Tx and the Rx mixers.
IFIntermediate FrequencyThe beat / baseband signal after the mixer, DC–20 MHz.
PAPower AmplifierTx output stage, 12–14 dBm.
LNALow-Noise AmplifierFirst Rx stage; sets most of the NF.
NFNoise FigureNoise the receiver adds, in dB.
VGAVariable-Gain AmplifierProgrammable-gain stage (IF), or signed VGAs in the phase rotator.
TIATransimpedance AmplifierCurrent-to-voltage stage after a passive mixer.
HPFHigh-Pass FilterIn IF: removes leakage and DC, and equalises R⁻⁴.
LPFLow-Pass Filter
AAFAnti-Aliasing FilterLow-pass filter in front of the ADC.
PPFPolyphase FilterRC-CR network that generates I/Q.
I/QIn-phase / QuadratureTwo copies of a signal 90° apart.
IRRImage Rejection RatioHow well an I/Q path suppresses the mirror frequency.
SSBSingle SidebandA frequency shift in one direction only, e.g. the loopback offset.
BPSKBinary Phase-Shift Keying0/180° switching; produces ±f_m sidebands.
VMVector ModulatorI/Q-weighted phase rotator.
RTPSReflection-Type Phase ShifterHybrid coupler terminated by tunable loads.
P1dB1 dB compression pointInput (or output) level where gain drops by 1 dB.
IIP3Input third-order intercept pointLinearity figure; cascades with the gain before each stage.
PAEPower-Added Efficiency(P_out − P_in) / P_DC of a PA.
PsatSaturated output powerMaximum output power of a PA.
DSBDouble SidebandNoise figure counting both sidebands (typical for zero-IF mixers).
PGAProgrammable-Gain AmplifierIF VGA with digitally set gain steps.
AGCAutomatic Gain ControlLoop that sets the VGA gain from the signal level.
HD22nd-order Harmonic Distortion (HD3 = 3rd)Harmonics show up as ghost targets at 2×/3× range.
AM/PMAmplitude-to-phase conversionPhase change caused by amplitude or supply changes.
RLReturn LossMatch quality, in dB.
CplCoupler (directional coupler)Taps a small, known fraction (−15…−20 dB) of a signal for detection or loopback.
PDPower DetectorDiode or square-law RF-to-DC converter.
APCAutomatic Power ControlClosed loop from the power detector to the VGA/PA gain code.
DPDDigital PredistortionWideband PA linearisation used in comms; radar uses static gain/phase LUTs instead.
VSWRVoltage Standing-Wave RatioMismatch at the antenna/load; changes the delivered power.
CWContinuous WaveAn unmodulated single-frequency tone.
ENRExcess Noise RatioCalibrated output level of a noise source.
SiGeSilicon-Germanium (BiCMOS)Bipolar-plus-CMOS process used for earlier radar chips.
CMOSComplementary Metal-Oxide-SemiconductorRFCMOS = CMOS process with RF / mmWave capability.
RFCMOSRF-capable CMOSToday's single-chip radar process (e.g. 45/40/28/22 nm).
NMOSN-channel MOS transistor (PMOS = P-channel)
PTATProportional To Absolute TemperatureVoltage or current used by temperature sensors.
BGBandgap referenceTemperature-stable voltage reference (BG1/BG2 = two independent ones).
PVTProcess, Voltage, TemperatureVariation corners a design must cover.

LO, clock & synthesis

Abbr.Stands forNote
PLLPhase-Locked LoopLocks the VCO to the crystal and generates the ramp.
ADPLLAll-Digital PLLPLL with a TDC and a digital loop filter.
VCOVoltage-Controlled Oscillator≈ 19–20.25 GHz, multiplied ×4 to 77 GHz.
PFDPhase-Frequency DetectorCompares the divided VCO with the reference.
CPCharge Pump; also Chip ProbeIn a PLL: converts PFD pulses into loop-filter current. In test flows: wafer-level probe test.
TDCTime-to-Digital ConverterDigital phase detector of an ADPLL.
XTALCrystal (reference oscillator)40/50 MHz reference for PLL, ADC and digital.
DDSDirect Digital SynthesizerDigitally generated tone (IF test tone).
KvcoVCO gainTuning sensitivity in MHz/V; varies per band.
AFCAutomatic Frequency Calibration (band calibration)Picks the VCO capacitor-bank band before the PLL closes.
MMDMulti-Modulus DividerProgrammable divider in the PLL feedback, dithered by the ΣΔ modulator.
CMLCurrent-Mode LogicFast differential logic used for the first divider (prescaler) stages at 20 GHz.

Data conversion & digital

Abbr.Stands forNote
SNMStatic Noise MarginNoise an SRAM cell tolerates before flipping; read SNM is the critical one.
SEUSingle-Event UpsetBit flip caused by a particle strike (soft error).
SERSoft-Error RateRate of radiation-induced bit flips, in FIT per Mbit.
MBUMulti-Bit UpsetOne particle flips several neighbouring cells.
RDFRandom Dopant FluctuationRandom number of dopant atoms in a small channel; main cause of Vth mismatch.
LERLine-Edge RoughnessRough gate edges that add Vth variation in small devices.
NPSFNeighbourhood Pattern-Sensitive FaultMemory cell fails only for certain data in its neighbours.
SEC-DEDSingle-Error Correction, Double-Error DetectionECC code used on memory words (e.g. 64 + 8 bits).
ROMRead-Only MemoryMask-programmed memory for boot code and tables.
DRFData-Retention FaultMemory cell loses its value after a pause.
WLWord LineRow select line of a memory array.
BLBit LineColumn data line of a memory array (BL and its complement BLB).
OSROversampling Ratiofs / (2·BW) of a ΣΔ ADC; each doubling adds (2L+1)·3 dB of SQNR.
NTFNoise Transfer FunctionHow a ΣΔ loop shapes quantisation noise out of the signal band.
SNDRSignal-to-Noise-and-Distortion RatioENOB = (SNDR − 1.76) / 6.02.
GBWGain-Bandwidth productOp-amp figure of merit; sets filter and PGA accuracy at IF.
CMRRCommon-Mode Rejection RatioHow well a differential circuit rejects signals common to both inputs.
CSI-2MIPI Camera Serial Interface 2High-speed serial link used for raw radar data.
TSNTime-Sensitive NetworkingDeterministic automotive Ethernet (IEEE 802.1).
ADCAnalog-to-Digital ConverterMain Rx ADCs (ΣΔ) and the aux SAR ADC.
DACDigital-to-Analog ConverterRotator I/Q weights, bias, trim, test-tone generation.
SARSuccessive-Approximation Register ADCSlow, precise ADC used for monitoring.
CTContinuous-Time (CT-ΣΔ)Sigma-delta ADC with a continuous-time loop filter.
ENOBEffective Number of Bits
SFDRSpurious-Free Dynamic RangeTone relative to the largest spur.
THDTotal Harmonic Distortion
DRDynamic Range
INLIntegral NonlinearityDeviation from the ideal transfer line.
DNLDifferential NonlinearityDeviation of each step from 1 LSB.
LSBLeast Significant Bit (MSB = most significant)
DFEDigital Front EndDecimation, DC and I/Q correction, interference handling.
CICCascaded Integrator-CombMultiplier-free decimation filter.
FIRFinite Impulse Response filter
DSPDigital Signal Processor
MCUMicrocontroller Unit
CPUCentral Processing UnitE.g. the RF control CPU that runs chirps and monitors.
SoCSystem-on-Chip
ECUElectronic Control UnitVehicle computer that receives radar data.
SRAMStatic RAM (RAM = random-access memory)
ECCError-Correcting CodeMemory protection for functional safety.
DMADirect Memory Access
LUTLook-Up TableE.g. phase-code → (I, Q) DAC codes, Tx power vs. temperature.
FWFirmware (SW = software)
OTPOne-Time Programmable memoryHolds trims and IDs, like eFuse.
SPISerial Peripheral InterfaceConfiguration / test bus from host or ATE.
CANController Area Network (CAN-FD = Flexible Data-rate)Vehicle bus for object lists.
ETHEthernet (100/1000BASE-T1)Automotive single-pair Ethernet.
CSICamera Serial Interface (MIPI CSI-2)Streams raw radar data to a central SoC.
LVDSLow-Voltage Differential SignalingRaw-data debug / streaming interface.
GPIOGeneral-Purpose Input/Output
IOInput/Output

Power

Abbr.Stands forNote
LDOLow-Dropout regulatorLinear regulator; one per noise-sensitive domain.
PMICPower-Management ICExternal bucks, supervisors and watchdog.
PSRRPower-Supply Rejection RatioHow much supply ripple an LDO (or circuit) blocks.
UVUnder-Voltage (OV = Over-Voltage)Supply monitors, a safety mechanism.
OVOver-VoltageRail above its window; OV monitor protects gate oxides.
PORPower-On ResetHolds the chip in reset until all rails are valid.
BORBrown-Out ResetResets when a supply sags during operation.
ESREquivalent Series ResistanceOf the LDO output capacitor; its zero affects loop stability.
OCPOver-Current ProtectionCurrent limit on a regulator output.
IqQuiescent currentCurrent a regulator draws for itself with no load.
DLDODigital LDOSwitched pass-device array regulated by a comparator loop.
GNDGround

Test & production

Abbr.Stands forNote
EVBEvaluation BoardBoard that carries the chip on the bench with supplies, clock, interfaces and RF transitions.
SMUSource-Measure UnitPrecision instrument that forces voltage or current and measures the other.
OTAOver The AirMeasurement through antennas in a chamber instead of a cable or waveguide.
GSGGround-Signal-Ground probeRF wafer probe with a signal tip between two ground tips.
LRRMLine-Reflect-Reflect-Match calibrationOn-wafer VNA calibration method for mmWave probes.
SCPIStandard Commands for Programmable InstrumentsText command set used to automate instruments.
UPHUnits Per HourTest throughput of a tester and handler.
eFuseElectrically programmable fuseSilicided poly link blown by current; stores trim, ID and repair bits.
MTPMultiple-Time Programmable memorySmall NVM that can be rewritten a few times.
NVMNon-Volatile MemoryKeeps data without power: eFuse, OTP, MTP, flash.
VPPProgramming supply voltageSupply used only to program fuses or OTP.
CRCCyclic Redundancy CheckChecksum that detects corrupted data, e.g. trim words at boot.
RMAReturn Material AuthorizationA customer return, analysed through the 8D process.
8DEight DisciplinesStructured problem solving: team, describe, contain, root cause, correct, verify, prevent, close.
MSAMeasurement System AnalysisProves a measurement is precise and accurate enough (Gage R&R, bias, stability).
GRRGage Repeatability and Reproducibility%GRR < 10 % of tolerance is good.
BISTBuilt-In Self-TestOn-chip stimulus + measurement; result is pass/fail or a value checked against a limit.
BISMBuilt-In Self-MeasurementSame on-chip hardware, but the result is a calibrated value (dB, °, MHz) used for trim, calibration and trend.
DFTDesign for Test
ATEAutomatic Test EquipmentProduction tester.
ATPGAutomatic Test Pattern GenerationScan patterns for digital logic.
IDDQQuiescent supply-current testDetects leakage defects.
PMUParametric Measurement UnitATE DC source/measure channel.
CresContact resistanceProbe or socket contact resistance; measured with the Kelvin or two-current method.
FIMVForce Current, Measure VoltagePMU mode used for open/short (diode) tests.
FVMIForce Voltage, Measure CurrentPMU mode used for leakage and power-short tests.
IILInput leakage current, input Low (IIH = input High)Leakage limit per IO pin, typically |I| < 1 µA.
IIHInput leakage current, input High
ESDElectrostatic DischargePad protection diodes / clamps; the diode method uses them for continuity.
VDDPositive supply rail (VSS = ground / negative rail)
VSSGround / negative supply rail
DCRDC ResistanceE.g. of an RF shunt inductor, 0.5–3 Ω.
Hi-ZHigh impedance (Z)Output driver switched off; the pin floats.
OEOutput EnableControl that switches an output between driving and Hi-Z.
IOZOff-state output leakage (IOZH at VDDIO, IOZL at 0 V)Measured on a pin in Hi-Z; typically |I| < 1–10 µA.
VOHOutput High voltage threshold (VOL = Output Low)Tester comparator levels that decide 1 / Z / 0.
VOLOutput Low voltage threshold
HIGHZIEEE 1149.1 instruction that puts all outputs in Hi-Z
EXTESTIEEE 1149.1 external test instructionBoundary cells drive and sample the pins; used for interconnect and walking-Z tests.
RPCTReduced Pin-Count TestingTesting through JTAG / a few pins, so more sites fit on one tester.
BICSBuilt-In Current SensorOn-chip current monitor for IDDQ per power domain.
SPCStatistical Process Control
VNAVector Network AnalyzerLab instrument for S-parameters.
AWGArbitrary Waveform Generator
DMMDigital Multimeter
GPIBGeneral-Purpose Interface BusInstrument control bus (like LAN).

Safety, standards & regulation

Abbr.Stands forNote
PMHFProbabilistic Metric for random Hardware FailuresISO 26262 residual failure rate; < 100 FIT for ASIL B/C, < 10 FIT for D.
SEooCSafety Element out of ContextA part developed to ISO 26262 against assumed system requirements.
FMEAFailure Mode and Effects AnalysisQualitative analysis of failure modes; FMEDA adds rates and diagnostic coverage.
FITFailures In TimeFailures per 10⁹ device-hours.
MTTFMean Time To Failure
REDRadio Equipment Directive (EU 2014/53/EU)EU conformity framework for radio transmitters, including radar.
ASILAutomotive Safety Integrity LevelISO 26262 risk class A–D; radar typically ASIL-B.
ISOISO 26262 / ISO/SAE 21434 / ISO 11452Functional safety / cybersecurity / EMC immunity.
FTTIFault-Tolerant Time IntervalTime from fault to hazard; monitors must react within it.
FMEDAFailure Modes, Effects and Diagnostic Analysis
SPFMSingle-Point Fault MetricISO 26262 coverage metric.
LFMLatent Fault MetricISO 26262 coverage metric.
AECAutomotive Electronics Council (AEC-Q100 / Q200)Qualification of ICs / passives.
ETSIEuropean Telecommunications Standards InstituteEN 301 091 / EN 303 396 radar standards.
FCCFederal Communications CommissionUS rules, Part 95M for 76–81 GHz.
EMCElectromagnetic Compatibility
CISPRComité International Spécial des Perturbations RadioélectriquesCISPR 25: vehicle-component emissions.
OOBOut-of-Band (emission)

Digital DFT, stress & reliability

Abbr.Stands forNote
UHSTUnbiased Highly Accelerated Stress TestHAST without bias, 130 °C / 85 % RH; alternative to autoclave.
PTCPower Temperature CyclingTemperature cycling with the device powered on and off.
EDREndurance and Data RetentionQualification of non-volatile memory (AEC-Q100-005).
DPADestructive Physical AnalysisOpening and sectioning parts to inspect their construction and damage.
SEMScanning Electron MicroscopeHigh-magnification imaging of cross-sections and surfaces.
EDXEnergy-Dispersive X-ray spectroscopyElement analysis in the SEM: corrosion, contamination, IMC phases.
FIBFocused Ion BeamPrecise local cutting and imaging; also used for circuit edits.
LTPDLot Tolerance Percent DefectiveDefect rate a sampling plan rejects with given confidence; 0 / 231 ≈ 1 % at 90 %.
PCPreconditioningBake, moisture soak and reflow before environmental stress (JESD22-A113).
BLRBoard-Level ReliabilityReliability of the package soldered on a PCB (temperature cycling, drop, vibration).
SACTin-Silver-Copper solderLead-free solder alloy (e.g. SAC305) used for balls.
PPAPProduction Part Approval ProcessAutomotive approval package, includes the qualification report.
DRCDesign Rule CheckVerifies layout against foundry manufacturing rules.
LVSLayout Versus SchematicVerifies that the layout matches the schematic netlist.
PEXParasitic ExtractionExtracts resistance, capacitance (and inductance) of the layout for post-layout simulation.
AMSAnalog / Mixed-Signal simulationSimulation that combines transistor-level analog with digital models.
ECOEngineering Change OrderLate design fix; a metal-only ECO changes only upper masks.
SCRSilicon-Controlled RectifierThyristor; used as a compact ESD clamp, and the parasitic structure behind latch-up.
GGNMOSGrounded-Gate NMOSESD clamp that works by parasitic-bipolar snapback.
SEScan EnableSelects shift (SE = 1) or functional capture (SE = 0) in scan flops.
SAFStuck-At FaultNode permanently 0 or 1.
TDFTransition Delay FaultNode slow to rise or fall; needs at-speed test.
STAStatic Timing AnalysisFinds the longest paths for timing-aware ATPG.
OCVOn-Chip VariationTiming derates in STA that account for local (mismatch) variation across the die.
OCCOn-Chip Clock ControllerGates exact launch/capture pulses from the PLL for at-speed test.
LOCLaunch-On-CaptureAt-speed scheme; also called broadside.
LOSLaunch-On-ShiftAt-speed scheme; SE must switch at speed.
EDTEmbedded Deterministic TestScan compression (decompressor + compactor).
LFSRLinear-Feedback Shift RegisterPseudo-random sequence generator.
PRPGPseudo-Random Pattern GeneratorLFSR that feeds the scan chains in LBIST.
MISRMultiple-Input Signature RegisterCompresses responses into one signature.
STUMPSSelf-Test Using MISR and Parallel Shift-register sequence generatorStandard LBIST architecture.
LBISTLogic BISTOn-chip random-pattern scan test; used at key-on.
MBISTMemory BISTOn-chip March-algorithm test of SRAMs.
BIRABuilt-In Redundancy AnalysisComputes which spare rows/columns repair a memory.
BISRBuilt-In Self-RepairLoads the repair from eFuse at boot.
IRIR dropSupply voltage drop (current × resistance), high during scan shift.
JTAGJoint Test Action Group (IEEE 1149.1)Boundary scan and test access port.
TAPTest Access PortJTAG pins TCK, TMS, TDI, TDO, TRST.
BSDLBoundary Scan Description LanguageDescribes a chip's JTAG boundary-scan cells.
IJTAGInternal JTAG (IEEE 1687)Network for accessing on-chip instruments.
MUXMultiplexerSelects one of N inputs; in test, routes sub-block nodes to shared instruments.
ATBAnalog Test BusChip-wide wire pair (force + sense) that test muxes connect sub-block nodes to.
TGTransmission GateParallel NMOS + PMOS switch; passes the full voltage range.
SPDTSingle-Pole Double-Throw switchOne input to one of two outputs (SP4T = one of four).
BBMBreak-Before-MakeSwitching order that opens the old path before closing the new one.
ICLInstrument Connectivity Language (PDL = Procedural Description Language)IJTAG description files.
ECIDElectronic Chip IDUnique per-die ID for traceability.
HVSTHigh-Voltage Stress TestShort over-voltage on the core rail to activate latent oxide defects.
VLVVery-Low-Voltage testTest below nominal V to expose resistive defects.
VminMinimum operating voltageThe pass/fail edge in a shmoo; an outlier Vmin flags a defect.
PATPart Average Testing (AEC-Q001)Outlier limits at robust μ ± 6σ.
DPATDynamic Part Average TestingPAT limits recalculated per lot or wafer.
GDBNGood Die in Bad NeighbourhoodWafer-map screen that rejects dies in defect clusters.
NNRNearest-Neighbour ResidualCompares a die's parameter with its wafer neighbours.
DPPMDefective Parts Per MillionShipped-quality metric; automotive target ≪ 1.
FTFinal TestPackage-level production test.
SLTSystem-Level TestFunctional test by booting the application.
PCMProcess Control MonitorTest structures in the scribe lines, measured by the fab (see WAT).
WATWafer Acceptance TestFab's electrical test of PCM structures before shipping wafers.
OSATOutsourced Semiconductor Assembly and TestCompany that packages and tests the dies.
QAQuality AssuranceSample-based outgoing inspection and audits.
CpProcess capability (Cp)(USL − LSL) / 6σ: spec width vs process spread.
CpkProcess capability indexmin(USL − μ, μ − LSL) / 3σ; automotive often ≥ 1.33, critical ≥ 1.67.
USLUpper Spec Limit (LSL = Lower Spec Limit)Design limits; outside them the part is out of spec.
LSLLower Spec Limit
UCLUpper Control Limit (LCL = Lower Control Limit)Process limits from history, usually mean ± 3σ.
LCLLower Control Limit
OCAPOut-of-Control Action PlanDocumented steps when an SPC rule fires.
MRBMaterial Review BoardDecides what happens to held material: release, rework, scrap or deviation.
MIMMetal-Insulator-Metal capacitor (MOM = Metal-Oxide-Metal)Precision on-chip capacitors.
MOMMetal-Oxide-Metal (finger) capacitor
BSIMBerkeley Short-channel IGFET ModelIndustry-standard compact transistor model family used in design kits.
VtThreshold voltageGate voltage where the channel turns on; Vt_lin at low Vds, Vt_sat at VDD.
IdsatSaturation drain currentId at Vgs = Vds = VDD, per µm of width; transistor strength.
IoffOff-state leakage currentId at Vgs = 0, Vds = VDD.
DIBLDrain-Induced Barrier LoweringVt drop per volt of Vds in short-channel devices.
SSSubthreshold swing (also SS = slow-slow process corner)mV of Vgs per decade of Id below Vt; ≈ 60 mV/dec ideal at room temperature.
TTTypical-typical process cornerFF = fast-fast; FS / SF = skewed NMOS / PMOS corners.
FFFast-fast process corner
gmTransconductance∂Id/∂Vgs; the transistor's gain.
gdsOutput conductance∂Id/∂Vds; limits voltage gain (intrinsic gain = gm/gds).
CoxGate-oxide capacitance per areaεox / Tox.
ToxGate-oxide thicknessMeasured with C-V structures in WAT.
fTTransit (unity current-gain) frequency≈ gm / 2π(Cgs + Cgd).
fmaxMaximum oscillation (unity power-gain) frequencyNeeds ≈ 3–4× the operating frequency for useful gain.
VbdGate-oxide breakdown voltageVoltage where oxide current jumps past compliance in a V-ramp test.
BVBreakdown VoltageBVdss = drain-source (gate off), BVj = junction.
BVdssDrain-source breakdown voltageTransistor off-state breakdown; limits PA drain swing.
BVjJunction breakdown voltageReverse breakdown of N+/P-well or P+/N-well diodes.
TLPTransmission-Line PulseESD device characterisation with short, high-current pulses.
QBDCharge to breakdownTotal charge an oxide passes before breaking down (reliability).
IMDInter-Metal DielectricInsulator between metal lines/layers; tested with comb structures.
CMPChemical-Mechanical PlanarizationPolishing step that flattens each layer; source of centre, edge and ring signatures.
OPCOptical Proximity CorrectionMask shape corrections; weak spots (hotspots) can fail per reticle shot.
EBREdge Bead RemovalRemoves resist build-up at the wafer edge.
D0Defect densityRandom defects per unit area; baseline yield Y ≈ e^(−D0·A).
HTOLHigh-Temperature Operating Life≈ 1000 h biased at max temperature (AEC-Q100).
ELFREarly Life Failure RateShort high-temperature operation of a large sample.
TCTemperature CyclingPackage and interconnect fatigue test.
THBTemperature-Humidity-Bias85 °C / 85% RH with bias.
HASTHighly Accelerated Stress Test130 °C / 85% RH with bias.
RHRelative Humidity
HTSLHigh-Temperature Storage LifeUnbiased bake.
HBMHuman-Body Model (ESD)
CDMCharged-Device Model (ESD)
LULatch-UpJESD78 test.
EMElectromigrationMetal wear-out under current (wafer-level reliability).
TDDBTime-Dependent Dielectric BreakdownGate-oxide wear-out.
HCIHot-Carrier InjectionTransistor degradation mechanism.
NBTINegative-Bias Temperature InstabilityPMOS threshold drift.

PCB, package & materials

Abbr.Stands forNote
TCBThermo-Compression BondingFlip-chip attach with heat and force per die; fine pitch, low warpage.
MUFMolded UnderfillMold compound that also fills the gap under a flip-chip die.
CSAMC-mode Scanning Acoustic MicroscopyUltrasound imaging of voids and delamination inside packages.
AOIAutomated Optical InspectionCamera inspection of RDL, bumps and balls in the assembly line.
KGDKnown-Good DieDie that passed wafer sort and is released for assembly.
WLPWafer-Level PackagePackage built on the whole wafer (fan-in) or a reconstituted wafer (fan-out).
PCBPrinted Circuit Board
RDLRedistribution LayerThin-film Cu wiring that routes die pads to balls (fan-out) or bumps.
UBMUnder-Bump MetallisationMetal stack between die/RDL pad and the solder.
FOWLPFan-Out Wafer-Level PackageDie embedded in mold with RDL fanning out to balls; eWLB is one type.
CSPChip-Scale PackagePackage only slightly larger than the die (FC-CSP = flip-chip CSP).
AiPAntenna-in-PackageAntennas built into package layers.
IMCIntermetallic CompoundLayer formed between solder and UBM/pad; grows with time and temperature.
CTECoefficient of Thermal ExpansionMismatch between materials drives solder and bump fatigue.
MSLMoisture Sensitivity LevelJ-STD-020 rating: how long a package may sit out of dry pack before reflow.
ELKExtreme Low-k dielectricFragile back-end dielectric in advanced nodes; cracks under bump stress.
NSMDNon-Solder-Mask-Defined padPCB pad style that improves solder-joint fatigue life.
DkDielectric constantSets line impedance and antenna tuning; shifts with moisture.
DfDissipation factorDielectric loss; key for RF loss at 77 GHz.
C-SAMC-mode Scanning Acoustic MicroscopyNon-destructive imaging of delamination and voids.
PBOPolybenzoxazolePolymer dielectric used in RDL (like polyimide).
EOSElectrical OverstressDamage from overvoltage or overcurrent beyond ESD events.
BGABall Grid ArrayFC-BGA = flip-chip BGA.
eWLBembedded Wafer-Level Ball-grid-arrayFan-out package with low RF loss.
AoPAntenna-on-Package
LoPLaunch-on-PackageRF launched from package into waveguide / antenna.
GCPWGrounded Coplanar WaveguidePCB transmission line at the ball launch.
SIWSubstrate-Integrated WaveguideWaveguide formed with via fences in the PCB.
FR4Flame-Retardant epoxy-glass laminateStandard low-cost PCB material.
ENEPIGElectroless Nickel, Electroless Palladium, Immersion GoldPCB surface finish.
ImAgImmersion SilverPCB surface finish.
HASLHot-Air Solder LevelingSurface finish; too rough/uneven for 77 GHz.
PBTPolybutylene terephthalate (PC = polycarbonate)Radome plastics.
HFSSHigh-Frequency Structure Simulator (Ansys)3D EM simulator (CST = Dassault's equivalent).