Conclusion 1
1 arc second is 1/3600° (4.848e-6 rad), and 1/3 arc second is three times tighter; both are the same arc-second intent cluster and belong on this canonical page.
Hybrid Tool + Report
1 Arc Second Calculator and Decision Guide
This single canonical page answers two intents at once: immediate conversion/calculation and deeper precision decision support. Run the tool first, then use the report sections to verify whether your stack can defend the target.
Published: 2026-05-23 | Last reviewed: 2026-06-17
Arc-Second Calculator + Feasibility Check
Tool LayerEnter a target angle (for example 1 arc second or 1/3 arc second), then compare it with your backlash, encoder, and thermal floors. The output gives conversions, interpretive verdict, and next action.
Method boundary: the combined floor is modeled with root-sum-square uncertainty logic, while profile/control factors are screening multipliers. Treat outputs as planning guidance until verified by installed-axis test data. See methodology.
Quick presets
Empty state: run the tool to get unit conversion, uncertainty margin, and a practical next-step recommendation.
Alias target
1 arc second
Degree equivalent
2.778e-4°
Radian equivalent
4.848e-6 rad
Core decision output
Margin ratio + next action
Direct Answer: What Is 1 Arc Second?
Intent MergeThe phrase 1 arc second is handled here as an alias of the canonical “arc second” topic. The related phrase 1/3 arc second (also searched as 1 3 arc second resolution) is the same decision cluster at a tighter target. The numeric conversion is exact, but feasibility is conditional on uncertainty floors.
Conversion ladder
One arc second sits in the micro-radian class. One-third arc second is three times tighter, so installation and thermal terms dominate even faster.
Key numeric anchor
Arc-second
1.000000\"
Degree
2.778e-4°
Radian
4.848e-6 rad
Versus 1 arc-min
60x smaller
Linear shift at 100 mm radius
0.485 µm
Tighter related alias
1/3 arc second = 9.259e-5°
Executive Summary
Report LayerConclusion 2
Resolution and accuracy are not interchangeable: VIM defines accuracy as a concept (not a single numeric quantity), so one headline spec never closes the risk by itself.
Conclusion 3
Conversion is exact, but production confidence is not. Compare target against backlash, encoder floor, thermal drift, and coverage assumptions together.
Conclusion 4
Interface and mounting constraints can materially widen effective floor: the same encoder family can expose different position values per rev by interface and installation conditions.
Conclusion 5
If floor/target margin is above 1.0, treat claims as conditional and require staged validation before RFQ sign-off.
Suitable Profiles
- Teams defining rotary-axis uncertainty budgets before procurement.
- Projects where a 1 arc second or 1/3 arc second claim affects part yield or inspection pass rate.
- Applications with access to thermal and bidirectional repeatability tests.
Unsuitable Profiles
- Rough indexing applications where arc-minute tolerance is already acceptable.
- Teams that cannot measure backlash, thermal drift, or encoder floor in installed conditions.
- Cases expecting control tuning to fix a mechanically loose assembly.
Decision signals
The margin ratio in the tool output is the fastest signal for whether you should continue tuning or escalate architecture.
Research Update: Evidence Added Before Release
Reviewed 2026-06-17The page answers the 1 arc second alias and provides a working calculator. The latest research update improves evidence separation, model limits, radius sensitivity, and supplier proof requirements so the result can support procurement decisions.
| Audited gap | Decision risk | Added evidence | Evidence status |
|---|---|---|---|
| Model multipliers needed clearer evidence status | Users could treat profile/control factors as standard-defined constants rather than screening assumptions. | Added explicit multiplier boundaries in the tool result and a model-assumption table that separates NIST/JCGM/ISO-backed facts from ServoRotary screening heuristics. | Backed by NIST TN 1297 for uncertainty framing; multiplier values are labeled as heuristic and require measured replacement before acceptance. |
| Radius impact was under-covered | A single 100 mm example hides how the same 1 arc second angle becomes a larger linear tolerance problem at larger fixtures. | Added radius sensitivity rows for 50, 100, 250, 500, and 1000 mm plus arc/chord boundary notes. | Backed by NIST SP 330 plane-angle relationship theta = s/r; chord comparison is derived geometry. |
| Supplier evidence requirements were not operational enough | Readers could know that more evidence is needed but still send RFQs without comparable test conditions. | Added an RFQ evidence packet checklist with required fields, failure modes, and minimum acceptable proof. | Backed by VIM definitions, NIST TN 1297 uncertainty reporting, and ISO 230-family test scope. |
| Counterexamples needed stronger decision framing | High encoder bit depth or low backlash catalog language could still be over-read as full installed-axis capability. | Expanded decision notes around HEIDENHAIN interface-dependent position values, Renishaw installed accuracy conditions, and gearbox backlash arc-minute classes. | Backed by public manufacturer pages/PDFs reviewed in this stage; vendor-specific production transferability remains marked as unknown. |
Concept Boundaries You Must Keep Separate
VIM + NISTMost arc-second mistakes begin when terms are mixed. This table maps formal metrology definitions to procurement decisions, so the same phrase does not hide different risk assumptions.
| Concept | Formal boundary | Decision implication | Source | Date |
|---|---|---|---|---|
| Resolution (VIM 4.14) | Smallest change in a quantity being measured that causes a perceptible change in the indication. | Fine counts-per-rev helps command granularity, but noise/friction/deadband can still hide real motion changes. | JCGM 200:2012 (VIM), definition 4.14 | Published 2012; reviewed 2026-05-23 |
| Measurement accuracy (VIM 2.13) | Closeness of agreement between a measured quantity value and a true quantity value; accuracy is not a quantity and is not given a numerical value. | A single number (such as ±10") is only one uncertainty term, not complete system accuracy evidence. | JCGM 200:2012 (VIM), definition 2.13 | Published 2012; reviewed 2026-05-23 |
| Precision and repeatability (VIM 2.15, 2.21) | Precision is agreement among indications under specified conditions; repeatability means those conditions include same procedure, operators, system, and short interval. | Supplier claims must include explicit test conditions, not only best-case snapshots. | JCGM 200:2012 (VIM), definitions 2.15 and 2.21 | Published 2012; reviewed 2026-05-23 |
| Expanded uncertainty coverage (NIST TN 1297 §6) | Expanded uncertainty U = k × uc; k = 2 gives about 95% confidence, k = 3 gives about 99% for a normal distribution. | If suppliers use different k values, their “accuracy” numbers are not directly comparable. | NIST TN 1297 section 6.2 | Updated by NIST 2021-06-02; reviewed 2026-05-23 |
Evidence Table and Boundary Notes
Updated 2026-06-17This section adds source-linked numeric references and explicit limits so “arc second”, “1 arc second”, and “1/3 arc second” claims stay decision-grade, not slogan-grade.
Alias numeric anchor
1.000000\"
Canonical answer for the 1 arc second query cluster.
Degree equivalent
2.778e-4°
Shows why arc-second targets are easy to over-claim.
1 arc-min comparison
60x
1 arc-min is 60x larger than 1 arc second.
| Topic | Evidence | Boundary | Source | Date |
|---|---|---|---|---|
| Exact unit conversion baseline | 1 arc second = 1/3600 degree = π/648000 rad; NIST SI guidance treats degree/minute/second of plane angle through radian conversion factors. | Exact conversion does not prove machine capability at that scale. | NIST SP 330 and SP 811 angle-unit guidance | Reviewed 2026-06-17 |
| Alias query numeric answer: 1 arc second | 1 arc second equals 1/3600 degree, π/648000 rad, and about 4.848136811e-6 rad. | This is the canonical numeric anchor. It does not prove an axis can hold a 1 arc second tolerance under load. | Derived directly from conversion formulas in this page | Model reviewed 2026-06-17 |
| Alias query numeric answer: 1/3 arc second | 1/3 arc second equals 0.333333", 9.259e-5°, and 1.616e-6 rad. | This is the target definition only. Capability still depends on uncertainty floors. | Derived directly from conversion formulas in this page | Model reviewed 2026-06-17 |
| Resolution vs accuracy boundary (formal metrology) | VIM defines resolution as smallest perceptible indication change, and defines accuracy as a concept (not a numeric quantity). | A single published ± value cannot substitute for full uncertainty decomposition. | JCGM 200:2012 (VIM definitions 4.14 and 2.13) | Published 2012; reviewed 2026-05-23 |
| Uncertainty combination basis used in tool logic | NIST TN 1297 defines combined standard uncertainty as the positive square root of the variance sum, and expanded uncertainty as U = k × uc (k often around 2 for interval reporting). | Coverage factor and distribution assumptions must be explicit. This page uses a screening model; its profile/control multipliers are not certified acceptance factors. | NIST Technical Note 1297 §5 + Appendix D | NIST page reviewed 2026-06-17 |
| Interface-dependent quantization in one encoder family | ECN 2000 documentation lists 33,554,432 values/rev on EnDat and 8,388,608 values/rev on FANUC serial interfaces. | Counts/rev claims are interface-bound; use the installed control chain value, not catalog maximum only. | HEIDENHAIN ECN 2000 product information PDF | Document date 2019-06; reviewed 2026-05-23 |
| Backlash context from precision gearboxes | Harmonic Drive HPF series commonly publishes backlash classes below 3 arc-min for selected configurations. | 3 arc-min = 180 arc-sec, which is far above 1/3 arc second class targets. | Harmonic Drive HPF product data | Accessed 2026-05-23 |
| Planetary gearbox class spread | Neugart PSN pages show multiple backlash classes by frame/stage, often in arc-minute ranges (<1 to <4 arc-min class examples). | Catalog class alone is not assembled-axis uncertainty. | Neugart PSN documentation | Accessed 2026-05-23 |
| Encoder class reference (mid-tier precision) | HEIDENHAIN ECN 2000 materials publish ±10 arc-sec system accuracy and 25-bit position values per revolution for selected variants. | Sensor class still requires installed alignment and thermal verification. | HEIDENHAIN ECN 2000 product page | Accessed 2026-05-23 |
| Encoder class reference (baseline integration class) | HEIDENHAIN ECN 100 pages list ±20 arc-sec positioning accuracy and 33,554,432 position values per revolution. | Even this class is 60x of a 1/3 arc second target. | HEIDENHAIN ECN 100 product page | Accessed 2026-05-23 |
| High-accuracy installed system reference | Renishaw VIONiC + REXM references list installed accuracy of ±1 arc-sec for ≥100 mm diameter rings, ±1.5 arc-sec at 75 mm, and ±2 arc-sec at ≤57 mm. | “Up to” values require strict installation and thermal controls; not universal. | Renishaw VIONiC / REXM materials | Reviewed 2026-06-17 |
| Validation framework for machine tools | ISO 230-2, ISO 230-7, and ISO 230-3 provide repeatability, rotary geometric, and thermal test structures for acceptance planning. | Standards define methods; you still need fixture and sampling plans for your exact machine. | ISO standard pages | Reviewed 2026-05-23 |
| Geometric vs thermal scope boundary in ISO 230 | ISO 230-7 explicitly states it does not cover angular positioning accuracy, repeatability of rotary axes, or thermal effects, which are addressed in other parts. | Passing one standard family test does not prove full arc-second capability under load and temperature. | ISO 230-7 scope statement | Reviewed 2026-05-23 |
Resolution vs Accuracy Boundary
Decision-critical gapThis gap is one of the most common planning errors in arc-second RFQs: a chain can advertise fine counts-per-rev while still publishing accuracy bands that are tens of times larger than a 1/3 arc-second target.
| Item | Published value | Decision impact | Boundary | Source | Date |
|---|---|---|---|---|---|
| Target quantization need for 1/3 arc-second | 3,888,000 counts/rev theoretical minimum | This is only a step-size baseline. It does not include backlash, compliance, or thermal drift. | Do not convert counts-per-revolution directly into accuracy claims. | Derived from NIST SP 330/SP 811 angle identities | Model reviewed 2026-06-17 |
| Target quantization need for 1 arc-second | 1,296,000 counts/rev theoretical minimum | This is the minimum count spacing needed to represent 1 arc second, before uncertainty floors are considered. | A controller can command this increment and still fail installed-axis accuracy if backlash, encoder error, or thermal drift is larger. | Derived from NIST SP 330/SP 811 angle identities | Model reviewed 2026-06-17 |
| HEIDENHAIN ECN 100 | ±20" positioning accuracy; 33,554,432 position values/rev | Resolution is ~0.0386"/count, but published accuracy band is ~60x of a 1/3" target. | Fine resolution does not erase installation error chain. | HEIDENHAIN ECN 100 product page | Accessed 2026-05-23 |
| HEIDENHAIN ECN 2000 | ±10" system accuracy; 33,554,432 position values/rev (25-bit) | Resolution is ~0.0386"/count, but published accuracy band is still ~30x of a 1/3" target. | Product materials also state mounting and thermal constraints that must be maintained. | HEIDENHAIN ECN 2000 product information | Accessed 2026-05-23 |
| ECN 2000 with FANUC serial interface | 8,388,608 position values/rev (23-bit interface mode) | Quantization becomes ~0.1545"/count, around 4x coarser than the 25-bit value. | Interface-level limits can dominate effective resolution even within the same encoder family. | HEIDENHAIN ECN 2000 product information PDF | Document date 2019-06; reviewed 2026-05-23 |
| HEIDENHAIN RCN 2001 / RCN 5001 reference class | RCN 2001: ±2", 67,108,864 values/rev; RCN 5001: ±4", 268,435,456 values/rev | Even this class is ~6x (±2") to ~12x (±4") wider than a 1/3" target band, while quantization is ~0.0193"/count to ~0.0048"/count. | High bit-depth and premium class hardware still require full installed-axis uncertainty closure. | HEIDENHAIN RCN 2001 / RCN 5001 product page | Accessed 2026-05-23 |
| Gearbox backlash references (HPF / PSN) | HPF: reduced backlash <3 arc-min classes; PSN: reduced backlash classes as low as <1 arc-min by variant | 1 arc-min = 60 arc-sec, so reducer backlash alone can exceed sub-arc-second targets by orders of magnitude. | Catalog reducer data is component-level and cannot replace installed-axis validation. | Harmonic Drive HPF and Neugart PSN pages | Accessed 2026-05-23 |
Interface and Mounting Limits (Counterexamples)
Implementation realityThese are common cases where teams over-read catalog specs. Treat each row as a pre-RFQ checklist: if the minimum check is missing, keep the claim in conditional status.
| Constraint | Published data point | Risk if misread | Minimum check | Source | Date |
|---|---|---|---|---|---|
| HEIDENHAIN ECN 2000 interface-dependent position values | EnDat interface: 33,554,432 values/rev (25-bit); FANUC serial interface: 8,388,608 values/rev (23-bit). | Copying the higher bit-depth value into every control stack can overstate usable quantization by 4x. | Lock interface, interpolation chain, and controller scaling before using counts/rev in RFQs. | HEIDENHAIN ECN 2000 product information | Document date 2019-06; reviewed 2026-05-23 |
| HEIDENHAIN RCN 2001 / RCN 5001 high-accuracy reference | RCN 2001: ±2" system accuracy, 2^26 values/rev; RCN 5001: ±4" system accuracy, 2^28 values/rev. | Very high position values per rev still do not guarantee sub-arc-second system accuracy. | Use these references as benchmark classes, then verify installed-axis uncertainty under your load/thermal envelope. | HEIDENHAIN RCN 2001/5001 product page | Accessed 2026-05-23 |
| Encoder mounting thermal compatibility requirement | ECN 2000 documentation requires mounting material thermal expansion coefficient in the range 10×10^-6 to 16×10^-6 K^-1 (for mounting dimensions >100 mm). | Ignoring this constraint can introduce thermal mismatch error even when encoder nominal specs are strong. | Match mounting material CTE and validate warm-state drift before tolerance sign-off. | HEIDENHAIN ECN 2000 mounting instructions | Document date 2019-06; reviewed 2026-05-23 |
Methodology and Source Map
The method is transparent: convert first, then compare your uncertainty floor with a safety-adjusted target. Anything outside source-backed facts is labeled as model assumption.
Core formulas
arcsec = value * unitFactor unitFactor: arcsec=1, arcmin=60, degree=3600, radian=648000/π degree = arcsec / 3600 radian = arcsec * π / 648000 combinedFloorArcsec = sqrt(backlash² + encoder² + thermal²) * profileFactor * controlFactor marginRatio = combinedFloorArcsec / (targetArcsec / safetyFactor)
This keeps unit conversion deterministic and makes risk interpretation explicit. It prevents raw conversion outputs from being mistaken as guaranteed machine capability.
Coverage-factor boundary: NIST TN 1297 uses U = k × uc and reports ~95% at k=2 / ~99% at k=3 under normal assumptions. This tool does not declare a formal k; the safety-factor field is a screening control, not a certified uncertainty statement.
Method flow
Tool layer resolves immediate tasks, report layer validates whether the result is actionable for real procurement and acceptance.
| Source | Support role | Timestamp |
|---|---|---|
| NIST Special Publication 330 (SI Brochure companion) | Unit definition consistency used for angle conversion identities on this page. | Accessed 2026-05-23 |
| NIST SP 811 (2008 Edition) | Reference conversion factor for arc-second to radian and angle unit relationships. | Reviewed 2026-05-23 |
| NIST TN 1297 §5 + Appendix D | Uncertainty-combination method (RSS) and expanded-uncertainty framing used in model boundaries. | Updated by NIST 2023-03-01; reviewed 2026-05-23 |
| NIST TN 1297 §6 (Expanded uncertainty) | Coverage-factor boundary used in supplier comparison checks (k=2 vs k=3 confidence framing). | Updated by NIST 2021-06-02; reviewed 2026-05-23 |
| JCGM 100:2008 (GUM) | General metrology framework for evaluating and expressing measurement uncertainty. | Reviewed 2026-05-23 |
| JCGM 200:2012 (VIM) | Formal definitions for resolution, accuracy, precision, and repeatability used to separate concept boundaries. | Published 2012; reviewed 2026-05-23 |
| HEIDENHAIN ECN 2000 | Encoder class reference around ±10 arc-sec for selected configurations. | Accessed 2026-05-23 |
| HEIDENHAIN ECN 2000 product information PDF | Interface-dependent position values (25-bit vs 23-bit) and mounting thermal-expansion constraints. | Document date 2019-06; reviewed 2026-05-23 |
| HEIDENHAIN ECN 100 | Encoder class reference around ±20 arc-sec. | Accessed 2026-05-23 |
| HEIDENHAIN RCN 2001 / RCN 5001 | High-end angular encoder reference (±2"/±4" with 26/28-bit class) used as counterexample boundary. | Accessed 2026-05-23 |
| Renishaw VIONiC + REXM | High-accuracy installed-system references up to ±1 arc-sec (conditions apply). | Accessed 2026-05-23 |
| Harmonic Drive HPF reference page | Backlash class examples used to explain why arc-minute stacks may miss arc-second goals. | Accessed 2026-05-23 |
| Neugart PSN reference | Backlash class spread by frame/stage used for comparison boundaries. | Accessed 2026-05-23 |
| ISO 230-2:2014 | Positioning accuracy and repeatability acceptance framework. | Reviewed 2026-05-23 |
| ISO 230-7:2015 | Rotary-axis geometric evaluation framework and scope boundaries. | Reviewed 2026-05-23 |
| ISO 230-3:2020 | Thermal effects validation framework. | Reviewed 2026-05-23 |
Model Assumptions: Source-Backed vs Heuristic
No hidden constantsThe calculator mixes exact unit math, metrology methods, and practical screening assumptions. This table shows which values are source-backed and which must be replaced by measured data before acceptance.
| Model item | Page use | Evidence level | Decision boundary | Source |
|---|---|---|---|---|
| Exact angle conversion | 1 arc second = 1/3600 degree = π/648000 radian, and all calculator unit outputs start here. | Deterministic unit identity | Safe for conversion answers, not sufficient for machine capability claims. | NIST SP 330 / NIST SP 811 angle-unit conventions |
| Root-sum-square floor combination | Combines backlash, encoder, and thermal terms before profile/control screening factors are applied. | Metrology method reference | Valid only when uncertainty components and distributions are defensible; acceptance still needs a formal uncertainty statement. | NIST TN 1297 Section 5 and Appendix A |
| Safety factor field | Tightens the effective target by dividing target arc-sec by the selected safety factor. | Screening heuristic | Not a formal coverage factor. For certification, replace with an explicit k value, confidence level, and uncertainty budget. | NIST TN 1297 Section 6 for coverage-factor boundary |
| Profile multiplier | Metrology/light dynamic = 0.8x, general indexing = 1.0x, heavy duty/high inertia = 1.45x. | ServoRotary planning heuristic | No reliable public dataset was found for a universal multiplier; replace this screening value with measured load, thermal, and duty-cycle evidence. | Explicit model assumption in this page |
| Control stack multiplier | Direct encoder = 0.7x, calibrated gearbox = 1.0x, open-loop estimate = 1.55x. | ServoRotary planning heuristic | No reliable public dataset was found for a universal multiplier; use this only for early screening before installed-axis tests. | Explicit model assumption in this page |
Radius Sensitivity: Why 1 Arc Second Is Not One Fixed Length
s = theta x rNIST SP 330 states the plane-angle relationship as angle equals circular arc length divided by radius. That means the same angular error becomes a larger linear error as fixture radius increases.
Arc vs chord boundary
At 100 mm radius, the arc-length projection for 1 arc second is 0.484814 µm. The chord value is 0.484814 µm, effectively identical at this scale; fixture offsets and Abbe geometry can still require a full kinematic model.
| Radius | 1 arc-sec | 1/3 arc-sec | 10 arc-sec | Decision note |
|---|---|---|---|---|
| 50 mm | 0.242 µm | 0.081 µm | 2.424 µm | Useful for quick tolerance translation, but still does not include backlash or thermal drift. |
| 100 mm | 0.485 µm | 0.162 µm | 4.848 µm | Useful for quick tolerance translation, but still does not include backlash or thermal drift. |
| 250 mm | 1.212 µm | 0.404 µm | 12.120 µm | Useful for quick tolerance translation, but still does not include backlash or thermal drift. |
| 500 mm | 2.424 µm | 0.808 µm | 24.241 µm | Large fixtures turn small angles into micron-level endpoint shifts; verify whether arc, chord, or full fixture transform is the measured quantity. |
| 1000 mm | 4.848 µm | 1.616 µm | 48.481 µm | Large fixtures turn small angles into micron-level endpoint shifts; verify whether arc, chord, or full fixture transform is the measured quantity. |
Standards Scope and Non-Covered Conditions
Standard references are useful only when scope boundaries are explicit. Use this matrix to avoid assuming one passed test closes all arc-second risks.
| Standard | What it covers | What it does not cover | Minimum action | Source |
|---|---|---|---|---|
| ISO 230-2:2014 (confirmed 2025) | Direct measurement methods for positioning accuracy and repeatability; applies to both linear and rotary axes. | Does not by itself close geometric or thermal risk; must be paired with other test codes. | Use as baseline for bidirectional repeatability acceptance criteria. | ISO 230-2 abstract and publication page |
| ISO 230-7:2015 (confirmed 2026) | Geometric accuracy and error-motion characteristics of rotary axes, including speed-influence checks. | Explicitly does not cover angular positioning accuracy/repeatability or thermal effects. | Use together with ISO 230-2 and ISO 230-3 before tolerance sign-off. | ISO 230-7 scope statement |
| ISO 230-3:2020 (confirmed 2026) | Thermal test structures including ambient variation, spindle heating, axis motion, and machine-operation effects. | Thermal method does not remove the need for mechanical backlash characterization. | Run warm-state drift checks with production-equivalent duty cycles. | ISO 230-3 abstract and publication page |
| NIST TN 1297 + JCGM 100 | Combined and expanded uncertainty framing to avoid single-number over-claims. | No substitute for real machine data; model inputs still require measured evidence. | Document assumptions and coverage factors in every supplier comparison. | NIST TN 1297 and JCGM 100 |
Scenario Snapshots
Decision examplesEach scenario includes precondition, process, and result so the page can be executed as a workflow, not read as generic advice.
Scenario A: alias query from search
Precondition: Visitor asks: “what is 1 arc second?” or “what is 1/3 arc second in degrees?”
Process: Tool presets run either 1 arc second or 1/3 arc second and immediately return degree/radian plus uncertainty margin.
Result: User gets direct conversion answer and an explicit warning if their floor exceeds the effective target.
Scenario B: RFQ screening
Precondition: Supplier claims “arc-second class” but provides only catalog backlash and no thermal test.
Process: Report tables map catalog class to target class and highlight missing evidence items.
Result: Team rejects ambiguous claims or requests standardized validation evidence before shortlist.
Scenario C: architecture escalation decision
Precondition: Conditional tool result after tuning attempts.
Process: Comparison matrix evaluates open-loop, closed-loop, servo, and direct-drive tradeoffs.
Result: Decision shifts from parameter tweaking to architecture-level risk reduction.
Scenario D: thermal surprise prevention
Precondition: Bench tests pass but production shifts fail intermittently.
Process: Risk matrix and validation gates enforce thermal drift measurement before release.
Result: Hidden warm-state drift is detected earlier, reducing downstream quality escapes.
Validation Gates Before Claiming Arc-Second Precision
Use this gate table when the tool gives a conditional result. Validation should be staged before committing vendor selection or published tolerance claims.
| Validation gate | Execution method | Reference standard |
|---|---|---|
| Repeatability gate | Run bidirectional indexing cycles and compute dispersion across multiple thermal states, not only best-case single runs. | ISO 230-2:2014 (confirmed 2025), positioning accuracy and repeatability test code. |
| Rotary geometric gate | Measure axis-of-rotation geometric behavior and speed-related deviation under production-equivalent loading. | ISO 230-7:2015 (confirmed 2026), geometric accuracy for axes of rotation. |
| Thermal drift gate | Perform warm-up and duty-cycle tests to quantify rotary and structural thermal distortion impacts. | ISO 230-3:2020 (confirmed 2026), thermal effects determination framework. |
RFQ Evidence Packet for 1 Arc Second Claims
Supplier comparableA supplier-ready request needs comparable proof, not just a smaller number. Use this packet when the tool returns pass or conditional; otherwise, redesign the mechanical stack first.
| Evidence item | Why it matters | Minimum proof | Reject / flag if |
|---|---|---|---|
| Metric identity | Prevents comparing resolution, accuracy, repeatability, and backlash as if they were the same claim. | Label each number with the measured quantity, units, plus/minus convention, and whether it is component-level or installed-axis. | Reject or mark conditional when a quote says only “1 arc-sec precision” without metric definition. |
| Test method and positions | ISO 230-2 style positioning checks depend on direct measurement of individual axes, tested positions, direction, and repeats. | Dated report with axis tested, positions, approach direction, repeats, instrument, and uncertainty statement. | Do not compare vendors when one gives catalog values and another gives installed repeatability logs. |
| Thermal envelope | A cold bench result can diverge from production after warm-up, ambient change, or duty-cycle heating. | Warm-state drift or ISO 230-3 style thermal test notes for the expected duty cycle and ambient range. | Treat as high risk when only room-temperature no-load data is provided. |
| Load and inertia condition | Payload inertia and radial/axial loads change compliance, settling, and backlash behavior. | Test payload, inertia estimate, fixture radius, speed/accel profile, dwell time, and orientation. | Do not transfer supplier bench data to heavy-duty production without matching load conditions. |
| Coverage factor or confidence statement | NIST TN 1297 shows expanded uncertainty depends on coverage factor k and confidence assumptions. | Coverage factor, confidence level, combined standard uncertainty, and list of included terms. | Mark as not directly comparable when accuracy values omit k/confidence/test condition disclosure. |
Architecture Comparison
Comparison is anchored to uncertainty budgeting logic, not only upfront BOM. Use it when the result is conditional or borderline.
| Option | Quant context | Strength | Limitation | When to choose |
|---|---|---|---|---|
| Open-loop stepper + reducer | Can provide fine command increments, but uncertainty floors are often dominated by backlash and disturbance sensitivity. | Lowest upfront cost in many projects | Weak for sub-arc-second claims unless mechanical and disturbance controls are exceptional. | Arc-minute to low arc-second class is sufficient and cycle profile is gentle. |
| Closed-loop stepper + precision gearbox | Better fault detection and correction logic than open loop, while still inheriting mechanical limits. | Balanced cost vs control observability | Cannot erase poor mechanical stack-up or thermal drift. | Need better robustness but still optimizing budget versus full servo retrofit. |
| Servo + high-resolution feedback | Common path for tighter dynamic control with stronger disturbance handling. | Higher dynamic precision envelope | Higher integration effort and commissioning burden. | Cycle-time and repeatability requirements exceed practical stepper envelope. |
| Direct-drive + high-accuracy encoder | Removes gearbox backlash source and can align better with very tight arc-second budgets. | Best path for strict angular precision classes | Higher hardware and thermal-control demands. | Program-level cost of positional error is high enough to justify premium architecture. |
Risk and Trade-off Controls
Misuse risk
Treating unit conversion as proof of axis capability causes inflated tolerance claims and unstable supplier decisions.
Cost risk
Underestimating uncertainty can trigger late redesign, repeated commissioning, and scrap costs.
Scenario mismatch
Cold-state demo results often diverge from warm-state production behavior if thermal gates are skipped.
| Risk trigger | Quant boundary | Decision impact | Minimum mitigation |
|---|---|---|---|
| Confusing conversion correctness with machine capability | Conversion exactness = 100%, capability confidence = variable | Teams may claim 1/3 arc second readiness before validating backlash and thermal floors. | Require floor/target ratio and validation gate completion before external claims. |
| Backlash floor above target class | Backlash > effective target arc-sec | Controller improvements cannot close the gap alone. | Prioritize mechanical redesign, preload strategy, or architecture upgrade. |
| Encoder floor misunderstood as system floor | Catalog sensor spec copied directly into system claim | Quoted accuracy may fail in installation because alignment and thermal terms were omitted. | Use installed-axis calibration and uncertainty budget breakdown. |
| Thermal effects ignored | No hot-state drift measurement under realistic duty cycle | Cold-bench pass can fail during continuous operation. | Run ISO 230-3 style thermal drift checks before acceptance. |
Action A: Margin > 2.0
Freeze claims, reduce mechanical floor, and rerun tool before discussing controller-level optimization.
Action B: Margin 1.0-2.0
Keep architecture provisional and run validation gates with explicit pass/fail criteria.
Action C: Margin <= 1.0
Move to supplier comparison with standardized uncertainty and thermal evidence requirements.
Known Unknowns (Avoid Over-Claiming)
These items are intentionally flagged as uncertain so teams can plan validation work instead of assuming transferability.
Cross-vendor 24/7 proof at sub-arc-second in full production load
Status: No reliable public dataset currently available
Most public sources describe component capabilities, not long-horizon assembled-system drift under identical duty cycles.
Minimum action: Require supplier logs for payload inertia, thermal envelope, and cycle profile matching your use case.
Coverage-factor disclosure in public encoder/axis brochures
Status: Frequently missing or not directly comparable
Many brochures publish accuracy values without clearly stating whether intervals map to k=2, k=3, or another confidence convention.
Minimum action: Request explicit uncertainty model, coverage factor, and confidence statement before vendor ranking.
Transferability from lab fixture to installed machine
Status: Needs site-specific confirmation
Fixture stiffness, mounting geometry, and ambient control can materially shift uncertainty budgets.
Minimum action: Repeat acceptance on the installed axis, not only on a supplier bench.
Cost of ownership delta between conditional and pass architecture
Status: Depends on local maintenance model
BOM delta is visible, but downtime and tuning burden vary significantly by team maturity and duty cycle.
Minimum action: Model lifecycle cost (scrap, downtime, rework) before choosing lower upfront cost options.
FAQ
What is 1 arc second?
1 arc second is 1/3600 degree, π/648000 radian, approximately 4.848136811e-6 radian, and about 0.484814 um of circular arc length at a 100 mm radius.
Is “1 arc second” the same intent as “arc second”?
Yes. For this site, 1 arc second is an alias inside the canonical /learn/arc-second page because the query asks for the same unit meaning, conversion, and precision decision workflow.
What is 1/3 arc second in degrees?
1/3 arc second equals 0.333333 arc-second, 9.259e-5 degree, and 1.616e-6 radian. This page computes the same conversion in the tool panel.
Is “1/3 arc second” the same intent as “arc second”?
For this site, yes. We treat 1/3 arc second as an alias intent inside the canonical /learn/arc-second page to avoid duplicate pages and keep one decision context.
What does “1 3 arc second resolution” mean in search queries?
It usually refers to the same value as 1/3 arc second resolution (0.333333"). This page handles both spellings and maps them to one canonical decision workflow.
Why does exact conversion not guarantee machine accuracy?
Conversion only translates units. Accuracy claims depend on mechanical backlash, encoder uncertainty, thermal drift, and control behavior under load.
What is the fastest quality gate from the tool output?
Check floor/target margin ratio. If it is above 1.0 after safety factor, treat the claim as conditional and require validation evidence.
How do I choose a safety factor?
For early feasibility screening, many teams start around 1.2 to 1.5, then tune by measured variance and acceptance risk. This range is a modeling heuristic, not a published ISO/NIST default.
What is the practical difference between resolution and accuracy?
Resolution is the smallest detectable indicated change (VIM 4.14), while accuracy is closeness to true value and is not itself a numeric quantity (VIM 2.13). In practice, fine counts-per-rev can coexist with weak installed-axis accuracy.
Why can the same encoder family show 25-bit and 23-bit values?
Interface and electronics chains can expose different position values per revolution. For example, ECN 2000 documentation lists 25-bit values on EnDat and 23-bit values on FANUC serial interfaces, so quote resolution with interface context.
Can closed-loop control always solve a bad gearbox?
No. Feedback improves observability and correction but cannot fully remove structural backlash and compliance limits.
What if my target is larger than one arc-minute?
The tool still works, but this page is optimized for precision discussions in arc-second classes where uncertainty budgeting is critical.
Which standards are practical for acceptance planning?
ISO 230-2 for repeatability and positioning logic, ISO 230-7 for rotary geometric checks, and ISO 230-3 for thermal drift framing.
Does an encoder spec equal installed system spec?
No. Installed accuracy also depends on alignment, coupling behavior, mounting quality, and thermal control.
How should I compare suppliers using this page?
Ask each supplier for the same uncertainty terms: backlash, encoder floor, thermal drift, duty-cycle evidence, and validation method.
When should I escalate to servo or direct-drive?
Escalate when margin remains conditional after reasonable mechanical optimization and tuning, or when cycle-time sensitivity is high.
What is the minimum dataset before purchase?
At minimum: bidirectional repeatability results, thermal drift behavior, backlash measurements on assembled axis, and acceptance criteria tied to your duty cycle.
Need supplier-ready tolerance screening?
Share your target arc-second class, backlash floor estimate, and duty cycle. We can convert this page output into an RFQ comparison checklist.