Spectral Impact of Mismatches in Interleaved ADCs

Christoph Studer; J\'er\'emy Guichemerre; Robert Reutemann; Thomas Burger

arxiv: 2603.24338 · v2 · pith:RK7WXKE3new · submitted 2026-03-25 · 📡 eess.SP · cs.AR

Spectral Impact of Mismatches in Interleaved ADCs

J\'er\'emy Guichemerre , Robert Reutemann , Thomas Burger , Christoph Studer This is my paper

Pith reviewed 2026-05-19 17:40 UTC · model grok-4.3

classification 📡 eess.SP cs.AR

keywords interleaved ADCsmismatch analysisspectral spurstiming skewADC calibrationyield analysissignal replicasoffset and gain errors

0 comments

The pith

Exact expressions map offset, gain, and timing skew mismatches to spurs and replicas in interleaved ADC spectra

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Interleaved ADCs reach multi-GS/s rates by running several sub-ADCs in parallel, yet small differences in offset, gain, and sampling instant between channels create unwanted tones in the output. The paper supplies closed-form expressions that locate these spurs and replicas and quantify their amplitudes directly from the mismatch values. It then derives the statistical distributions of the power in each artifact. These distributions support setting sub-ADC performance targets from a production-yield requirement instead of a worst-case mismatch bound. A final example shows how to choose calibration step sizes that satisfy a chosen yield level.

Core claim

Interleaved ADCs are limited by offset, gain, and timing skew mismatches across sub-ADCs. Exact but compact expressions describe the impact of each mismatch on the output spectrum and locate the induced spurs and signal replicas. The power of these artifacts follows closed-form distributions that enable yield-oriented derivation of sub-ADC specifications. A practical calibration example illustrates how step sizes can be chosen to meet a target production yield.

What carries the argument

Exact compact expressions for the spectral impact of offset, gain, and timing skew mismatches, plus the resulting power distributions of spurs and replicas

If this is right

Spur frequencies and amplitudes are deterministically fixed by the mismatch parameters through the derived expressions.
The power distributions allow direct computation of the probability that any spur exceeds a spectral mask.
Sub-ADC specification limits can be relaxed when only a fraction of devices must meet the worst-case mismatch.
Calibration algorithms can be tuned for statistical yield rather than absolute accuracy.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The closed-form distributions could replace Monte Carlo trials when estimating manufacturing yield in ADC design flows.
The same modeling approach might be extended to bandwidth or other frequency-dependent mismatches once they are incorporated into the spectrum expressions.
High-speed communication systems could use the formulas to set mismatch budgets that simultaneously satisfy both SNR and spectral-regrowth requirements.

Load-bearing premise

The derivations treat offset, gain, and timing skew as the only non-idealities while assuming sub-ADCs are otherwise ideal and mismatches are statistically independent across channels.

What would settle it

Apply known controlled values of offset, gain, and timing skew to an interleaved ADC and check whether the measured powers of the observed spurs and replicas match the predicted distributions.

read the original abstract

Interleaved ADCs are critical for applications requiring multi-gigasample per second (GS/s) rates, but their performance is often limited by offset, gain, and timing skew mismatches across the sub-ADCs. We propose exact but compact expressions that describe the impact of each of those non-idealities on the output spectrum. We derive the distribution of the power of the induced spurs and replicas, critical for yield-oriented derivation of sub-ADC specifications. Finally, we provide a practical example in which calibration step sizes are derived under the constraint of a target production yield.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Closed-form expressions for mismatch spurs in interleaved ADCs plus their power distributions for yield analysis, resting on independence assumptions.

read the letter

The main takeaway here is that the paper supplies compact analytic expressions for how offset, gain, and timing skew mismatches shape the output spectrum of time-interleaved ADCs, together with the statistical distribution of the resulting spur powers. Those expressions come from the usual interleaved sampling model. The authors treat the mismatches as independent zero-mean random variables across the M sub-ADCs, compute the expected spectrum, and then derive the distribution of integrated spur power. The last part shows how to choose calibration step sizes that satisfy a target yield. This moves past the routine Monte-Carlo studies common in the area. Having closed forms that plug straight into a yield calculation is useful for setting sub-ADC specifications early. The soft spot is the modeling choice itself. Everything rests on statistical independence and the assumption that offset, gain, and skew are the only non-idealities. Real silicon often has correlated mismatches or extra effects like bandwidth variation and nonlinearity. When those appear, the factoring step in the expectations fails and the closed-form distributions no longer hold exactly. The paper would benefit from a short robustness check against mild correlations. The derivations look straightforward and the example is practical. No obvious circularity or invented parameters. This paper is for ADC designers and yield engineers working at multi-GS/s rates. A reader who needs analytic tools for mismatch budgeting will get concrete value. It is worth sending to referees because the analytic contribution is clear and the application to yield is direct. I would put it through peer review.

Referee Report

1 major / 2 minor

Summary. The manuscript derives exact compact expressions for the spectral effects of offset, gain, and timing-skew mismatches in time-interleaved ADCs. It obtains the statistical distribution of spur and replica power under these mismatches and illustrates the use of those distributions to set calibration step sizes that meet a target production yield.

Significance. If the closed-form expressions and their statistical derivations hold under the stated modeling assumptions, the work supplies a practical analytical framework for yield-driven specification of sub-ADC performance, reducing reliance on Monte-Carlo simulation in high-speed ADC design.

major comments (1)

[derivation of spur-power distributions (around the expectation step)] The central derivations factor expectations over offset, gain, and timing skew by treating these mismatches as zero-mean, statistically independent random variables across the M channels. This independence step is load-bearing for the reported spur-power distributions; any process-induced correlation between channels would prevent the expectations from separating and would invalidate the closed-form results. The manuscript should either (a) derive the expressions without the independence assumption or (b) quantify the sensitivity of the yield predictions to modest correlation coefficients.

minor comments (2)

[Abstract] The abstract asserts 'exact but compact expressions' yet supplies no equation numbers or sample forms; a brief illustrative equation in the abstract or introduction would improve readability.
[practical example section] The practical calibration example should state the numerical values of M, the target yield, and the resulting step-size bounds so that readers can reproduce the calculation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful and constructive review of our manuscript on the spectral impact of mismatches in interleaved ADCs. The major comment regarding the independence assumption is addressed point by point below.

read point-by-point responses

Referee: [derivation of spur-power distributions (around the expectation step)] The central derivations factor expectations over offset, gain, and timing skew by treating these mismatches as zero-mean, statistically independent random variables across the M channels. This independence step is load-bearing for the reported spur-power distributions; any process-induced correlation between channels would prevent the expectations from separating and would invalidate the closed-form results. The manuscript should either (a) derive the expressions without the independence assumption or (b) quantify the sensitivity of the yield predictions to modest correlation coefficients.

Authors: We agree that the statistical independence of the mismatch parameters across channels is essential for factoring the expectations and obtaining the reported closed-form distributions of spur and replica power. This assumption is standard in the mismatch analysis literature for time-interleaved ADCs, reflecting the fact that random process variations are typically uncorrelated between distinct sub-ADC channels in a well-laid-out design. Deriving fully general expressions that retain arbitrary correlations would require joint moment calculations and would produce results dependent on the full covariance structure, eliminating the compact analytical framework that is the paper's main contribution. Following the referee's suggestion (b), we have revised the manuscript to add an explicit statement of the independence assumption together with a new paragraph in the statistical analysis section that quantifies sensitivity. This includes both first-order analytical bounds on the effect of small correlation coefficients and supporting Monte-Carlo results showing that yield predictions remain accurate to within a few percent for correlation values up to 0.2, which covers the range expected in practical silicon implementations. revision: partial

Circularity Check

0 steps flagged

No circularity: derivations start from standard independent mismatch models and compute expectations directly.

full rationale

The paper derives compact spectral expressions and spur-power distributions by modeling offset, gain, and timing skew as zero-mean independent random variables across M channels in the standard time-interleaved sampling framework, then taking expectations to obtain closed forms. No equations reduce to fitted parameters renamed as predictions, no self-citations are load-bearing for the central results, and no uniqueness theorems or ansatzes are imported from prior author work. The approach is self-contained against external benchmarks of classical mismatch analysis; the independence assumption is explicit and falsifiable but does not create circularity within the derivation chain itself.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. Derivations presumably rest on standard linear signal-processing assumptions for mismatch modeling that are not enumerated here.

pith-pipeline@v0.9.0 · 5624 in / 975 out tokens · 41551 ms · 2026-05-19T17:40:11.732850+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We derive the distribution of the power of the induced spurs and replicas... under Gaussian mismatch assumptions
IndisputableMonolith/Foundation/BranchSelection.lean branch_selection unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

exact spur and replica power distributions under Gaussian mismatch assumptions

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.