ECSS Harness Manufacturing — High-Voltage Application (space-systems/ecss/q2030-hv)
Use when the task is the high-voltage provision of ECSS-Q-ST-20-30C
section 6.20, which binds the high-voltage section of the
IPC/WHMA-A-620 acceptance basis — sizing the clearance and creepage a
harness termination owes at its working voltage and operating pressure,
grading what was actually built, and setting the proof test that
demonstrates it.
Domain quick reference
- The provisions have an entry condition. Below the high-voltage
threshold a termination is graded by the ordinary acceptance criteria
and owes no clearance, creepage or proof test under this section;
applying them anyway adds stress and cost without adding evidence.
- The insulating medium follows the pressure, and it does so
non-monotonically. Well above the Paschen minimum the medium is air
and its breakdown gradient falls roughly with pressure, so a cabin
or a launch ascent is worse than the ground. Well below it the
residual gas cannot sustain an avalanche and the limiting mechanism
becomes flashover along the insulating surface, at a far higher
gradient. Between the two lies the band where the breakdown voltage
of air passes through its minimum.
- Inside that band no practical spacing holds off a high voltage. The
answer is not a larger gap: it is encapsulation, potting or a
pressurized enclosure, so that the band is never the medium the
feature is asked to work in. A feature sized by a gap there has been
sized against a curve it is not on.
- Clearance and creepage are different paths and are sized differently.
Clearance is through the medium and follows the breakdown gradient;
creepage is along an insulating surface and follows the tracking
performance of the material and how dirty the surface is allowed to
get. Creepage is never allowed to be shorter than the clearance it
parallels, because the surface path would otherwise be the weaker of
the two.
- The proof test is derived, not chosen. Its voltage comes from the
working voltage, its dwell has to be long enough for a weakness to
show, and the leakage current during the dwell is the measurement
that actually says whether the insulation is sound.
Workflow
- Validate the feature record: identifier, working voltage, operating
pressure, insulating material group, contamination category,
encapsulation flag, as-built distances and the proof-test record. A
non-positive voltage or pressure, or an unknown material group or
contamination category, is an input error, not a case to clamp.
- Decide whether the high-voltage provisions apply at this working
voltage and report a feature below the threshold as not applicable
rather than as compliant by accident.
- Pick the insulation regime from the operating pressure, then take
the breakdown gradient of that regime — pressure-derated air above
the band, surface flashover below it.
- Raise the encapsulation finding for an unencapsulated feature inside
the band, and do not size a gap for it: there is no gradient there
to size against.
- Size the clearance from the working voltage, the gradient and the
safety factor, with a floor no feature goes below.
- Size the creepage from the working voltage, the material group and
the contamination category, with its own floor and never below the
clearance.
- Grade the as-built distances, derive the proof voltage and grade the
applied voltage, dwell and leakage, absorbing representation error
at each limit with a named tolerance rather than by widening it.
Pitfalls
- Sizing a gap for a feature that operates inside the Paschen minimum
band. The curve has no usable gradient there, so a larger gap buys
nothing; encapsulation or pressurization is the only answer.
- Using the sea-level breakdown gradient for a feature that flies. The
gradient falls with pressure across the air regime, and the ascent
profile passes through the worst of it.
- Treating vacuum as free insulation. The gas stops breaking down, but
the surface does not, and a contaminated or tracked surface flashes
over well below the clean-surface gradient.
- Sizing creepage without the material and contamination inputs. The
same voltage on the same geometry needs very different surface paths
on a high-tracking-performance material in a controlled environment
and a poor one in a dirty one.
- Accepting a proof test that reached a voltage someone picked. The
level comes from the working voltage, and a short dwell or an
unrecorded leakage current leaves the test without its measurement.
Behavior contract (gate 3)
The feature validation, applicability threshold, pressure-regime
selection and gradient derivation, encapsulation finding, clearance and
creepage sizing with their floors and the creepage-not-below-clearance
rule, as-built grading and the proof-voltage, dwell and leakage checks
are exercised by the gate 3 contract test: scripts/test_q2030_hv.py
against scripts/q2030_hv_logic.py (stdlib unittest, offline). Run:
python3 scripts/test_q2030_hv.py
Compliance
- ECSS standards are freely downloadable (ESA); cite the source and
paraphrase per standards-map.yaml.
- compliance: STANDARDS-REF, gated: false.
1---2name: q2030-hv3description: "Compute the high-voltage provisions an electrical harness owes under ECSS-Q-ST-20-30C section 6.20 and its IPC section 20 basis. Use when the task is sizing clearance and creepage at a harness termination and grading the as-built distances: derating the breakdown gradient of air with operating pressure, refusing air as an insulator inside the Paschen minimum band unless the feature is encapsulated, taking the surface-flashover gradient below that band, scaling creepage with the insulating material group and contamination category, and deriving the proof voltage, dwell and leakage limit of the high-voltage test. Trigger: ecss, q-st-20-30c, harness-high-voltage-clearance, harness-creepage-distance, paschen-minimum-band, vacuum-surface-flashover, harness-proof-voltage-test."4license: Apache-2.05compliance: STANDARDS-REF6standards:7 - id: ecss8 reference-only: true9gated: false10domain: space-systems11pack: space-systems12compatibility: "agentskills.io SKILL.md; any SKILL.md host (Claude Code, Hermes, OpenClaw)"13metadata:14 source: https://skillmd.com/skills/ashfordeou/q2030-hv15 domain: space-systems16 subdomain: ecss17 tags: [ecss, q-st-20-30-harness-scope, q2030-hv, harness-high-voltage-clearance, harness-creepage-distance, paschen-minimum-band, vacuum-surface-flashover, harness-proof-voltage-test]18 version: 0.1.019 author: Aero Agent Skills20---2122# ECSS Harness Manufacturing — High-Voltage Application (space-systems/ecss/q2030-hv)2324Use when the task is the high-voltage provision of ECSS-Q-ST-20-30C25section 6.20, which binds the high-voltage section of the26IPC/WHMA-A-620 acceptance basis — sizing the clearance and creepage a27harness termination owes at its working voltage and operating pressure,28grading what was actually built, and setting the proof test that29demonstrates it.3031## Domain quick reference3233- The provisions have an entry condition. Below the high-voltage34 threshold a termination is graded by the ordinary acceptance criteria35 and owes no clearance, creepage or proof test under this section;36 applying them anyway adds stress and cost without adding evidence.37- The insulating medium follows the pressure, and it does so38 non-monotonically. Well above the Paschen minimum the medium is air39 and its breakdown gradient falls roughly with pressure, so a cabin40 or a launch ascent is worse than the ground. Well below it the41 residual gas cannot sustain an avalanche and the limiting mechanism42 becomes flashover along the insulating surface, at a far higher43 gradient. Between the two lies the band where the breakdown voltage44 of air passes through its minimum.45- Inside that band no practical spacing holds off a high voltage. The46 answer is not a larger gap: it is encapsulation, potting or a47 pressurized enclosure, so that the band is never the medium the48 feature is asked to work in. A feature sized by a gap there has been49 sized against a curve it is not on.50- Clearance and creepage are different paths and are sized differently.51 Clearance is through the medium and follows the breakdown gradient;52 creepage is along an insulating surface and follows the tracking53 performance of the material and how dirty the surface is allowed to54 get. Creepage is never allowed to be shorter than the clearance it55 parallels, because the surface path would otherwise be the weaker of56 the two.57- The proof test is derived, not chosen. Its voltage comes from the58 working voltage, its dwell has to be long enough for a weakness to59 show, and the leakage current during the dwell is the measurement60 that actually says whether the insulation is sound.6162## Workflow63641. Validate the feature record: identifier, working voltage, operating65 pressure, insulating material group, contamination category,66 encapsulation flag, as-built distances and the proof-test record. A67 non-positive voltage or pressure, or an unknown material group or68 contamination category, is an input error, not a case to clamp.692. Decide whether the high-voltage provisions apply at this working70 voltage and report a feature below the threshold as not applicable71 rather than as compliant by accident.723. Pick the insulation regime from the operating pressure, then take73 the breakdown gradient of that regime — pressure-derated air above74 the band, surface flashover below it.754. Raise the encapsulation finding for an unencapsulated feature inside76 the band, and do not size a gap for it: there is no gradient there77 to size against.785. Size the clearance from the working voltage, the gradient and the79 safety factor, with a floor no feature goes below.806. Size the creepage from the working voltage, the material group and81 the contamination category, with its own floor and never below the82 clearance.837. Grade the as-built distances, derive the proof voltage and grade the84 applied voltage, dwell and leakage, absorbing representation error85 at each limit with a named tolerance rather than by widening it.8687## Pitfalls8889- Sizing a gap for a feature that operates inside the Paschen minimum90 band. The curve has no usable gradient there, so a larger gap buys91 nothing; encapsulation or pressurization is the only answer.92- Using the sea-level breakdown gradient for a feature that flies. The93 gradient falls with pressure across the air regime, and the ascent94 profile passes through the worst of it.95- Treating vacuum as free insulation. The gas stops breaking down, but96 the surface does not, and a contaminated or tracked surface flashes97 over well below the clean-surface gradient.98- Sizing creepage without the material and contamination inputs. The99 same voltage on the same geometry needs very different surface paths100 on a high-tracking-performance material in a controlled environment101 and a poor one in a dirty one.102- Accepting a proof test that reached a voltage someone picked. The103 level comes from the working voltage, and a short dwell or an104 unrecorded leakage current leaves the test without its measurement.105106## Behavior contract (gate 3)107108The feature validation, applicability threshold, pressure-regime109selection and gradient derivation, encapsulation finding, clearance and110creepage sizing with their floors and the creepage-not-below-clearance111rule, as-built grading and the proof-voltage, dwell and leakage checks112are exercised by the gate 3 contract test: scripts/test_q2030_hv.py113against scripts/q2030_hv_logic.py (stdlib unittest, offline). Run:114python3 scripts/test_q2030_hv.py115116## Compliance117118- ECSS standards are freely downloadable (ESA); cite the source and119 paraphrase per standards-map.yaml.120- compliance: STANDARDS-REF, gated: false.