Lilienfeld 1925 Check Unintended Conductive Paths

# Check Unintended Conductive Paths: Separate the Intended Film Route from Control-Electrode Shorts

Step 18 determines whether fabrication created a direct electrical path that bypasses the intended field-controlled structure. The device has three conductors: terminal 11, terminal 12, and aluminum control foil 13. The copper–sulfur film is supposed to conduct between 11 and 12. Therefore, continuity on 11↔12 is not evidence of a short. The forbidden paths are direct or excessive conduction from 11↔13 and 12↔13, including bridges across the glass, along the fracture seam, through debris, around fixture surfaces, or through damaged film at the foil edge.

This step is deliberately diagnostic and low stress. It does not yet establish the full terminal-to-terminal current–voltage behavior of Step 19 or quantify normal control-electrode leakage under the final operating bias of Step 20. Instead it asks whether a gross or latent conductive defect is present before those higher-information tests risk heating or irreversibly changing the specimen.

## 1. Test a connection matrix, not an ambiguous “continuity” beep

Measure all three conductor pairs while holding the unused conductor in a declared state:

Driven pairExpected interpretationStep 18 purpose
11↔12finite resistance through film 15reference intended path; detect an anomalously metallic bypass only against Step 17 expectations
11↔13ideally blocked/high resistance at low stressdetect left-side bridge to the control foil
12↔13ideally blocked/high resistance at low stressdetect right-side bridge to the control foil

“Floating” is a circuit condition, not an absence of a condition. A floating third terminal may charge capacitively and change the measured transient. Repeat critical checks with the unused main terminal floating and then guarded or tied to the measuring low node, documenting both configurations. Never connect the control foil blindly to a handheld ohmmeter whose test voltage, polarity, compliance, and transient behavior are unknown.

Three-terminal path discrimination for Lilienfeld Step 18 A physical plan view distinguishes the intended terminal 11 to terminal 12 compound-film current path from forbidden terminal-to-control-foil bridges, with likely defect locations along the foil edge, fracture seam, debris, and fixture surface. THREE CONDUCTORS CREATE THREE PAIRS — ONLY ONE IS THE INTENDED PATH 11↔12 carries film current; 11↔13 and 12↔13 expose control-electrode bridges DEVICE PATH MAP 111213 INTENDED: compound-film route 11↔12 FORBIDDEN 11↔13FORBIDDEN 12↔13 edge beaddebrisseam path INTERPRETATION RULE finite 11↔12 is expected and checked against Step 17 excess 11↔13 or 12↔13 is a hold/reject signal, not field effect

## 2. Begin with a fixture blank and open-circuit baseline

At high resistance, the fixture can impersonate the defect. Clean, dry, and shield the test enclosure; immobilize low-noise triaxial cables; control humidity; and allow the instruments to warm up. With no device installed, execute the complete voltage sequence through the same switching paths. Then install an insulating blank or geometrically representative dummy and repeat. These measurements establish cable, switch, connector, standoff, and surface leakage.

The measured current is approximately

$$ I_{meas}=I_{DUT}+I_{fixture}+I_{offset}+C_{eq}\frac{dV}{dt}+I_{generated}, $$

where $I_{generated}$ includes triboelectric, piezoelectric, thermal, and electrochemical contributions. A stable nonzero reading can still be fixture leakage. A changing reading after a voltage step may be displacement current or dielectric absorption rather than a device short.

Use guarding to intercept unwanted leakage in the high-impedance measurement path. A driven guard is held near the sensitive node’s potential so little voltage appears across cable or fixture insulation; it is not the same as an earth-connected safety shield. Verify the guard configuration against the instrument manual, and keep all accessible conductive enclosure surfaces safely grounded.

## 3. Use a bounded voltage staircase with current compliance

Start at zero, measure the zero-bias offset, and ramp in small steps to a predeclared screening voltage below any expected irreversible electrochemical, thermal, or breakdown regime. At every step record the complete current transient and wait for a defined settling criterion or maximum time. Reverse polarity and repeat. Return to zero between sequences to test for residual charge and permanent change.

For a settled point, an apparent isolation resistance may be reported as

$$ R_{iso}(V)=\frac{|V|}{|I_{ss}(V)-I_0|}, $$

but only when the corrected current is resolved above uncertainty. If it is not resolved, report a lower bound

$$ R_{iso}>\frac{|V|}{I_{lim}}, $$

where $I_{lim}$ includes noise, drift, fixture subtraction uncertainty, instrument accuracy, and repeatability. Never report infinity.

Abort immediately on compliance, abrupt current increase, visible change, temperature rise, unstable oscillation, or post-step offset shift. A screening test should not create the short it seeks to find.

## 4. Diagnose time dependence and polarity, not only magnitude

A metallic bridge tends to produce prompt, approximately proportional current with weak settling-time dependence. Capacitive charging decays after each voltage step. Dielectric absorption can decay slowly. Ionic or moisture-assisted surface conduction may drift, show polarity asymmetry, or retain history. A nonlinear junction at the foil-film interface may be asymmetric without being a metallic short. Step 18 records these signatures and flags abnormal paths; it does not yet claim the normal gate-leakage model reserved for Step 20.

For a linear candidate bridge, fit both polarities with

$$ I(V)=G_{bridge}V+I_{off}, $$

and examine residuals. The slope is more robust than one $V/I$ point, while the intercept exposes offset. Require that the extracted conductance exceed the combined blank and uncertainty bounds before calling a device path detected.

Guarded leakage measurement and diagnostic current transients A guarded source-measure circuit separates device current from fixture leakage, while current-versus-time and current-versus-voltage plots distinguish charging, fixture leakage, and a conductive bridge. A SHORT TEST IS ONLY AS GOOD AS ITS LEAKAGE FLOOR Blank the fixture, guard the sensitive node, preserve settling transients, then compare both polarities GUARDED PAIR MEASUREMENT SMUforce V / read I DUT PAIR11↔13or 12↔13 HILO / ammeter return driven guard intercepts fixture leakage earth-grounded safety shield Run open fixture + blank through the identical switch and voltage sequence. AFTER EACH VOLTAGE STEP time after stepmeasured current capacitive / absorption decay prompt sustained bridge current blank / fixture floor DISPOSITION LOGIC PASSboth control pairs remain below uncertainty-aware leakage limit through the screening rangeHOLDunresolved above blank, slow drift, humidity sensitivity, or disagreement between configurationsREJECTrepeatable sustained excess current, compliance event, permanent offset change, or localized bridgeTRACEcorrelate failing side and polarity with Step 15–17 geometry, coverage, thickness, and composition maps

## 5. Localize a detected path without destroying the evidence

First repeat the failing pair after swapping instrument channels, cables, and fixture positions. Re-run the open and blank. Clean only according to a predeclared protocol; a failure that disappears after uncontrolled cleaning is lost evidence. Vary humidity within a safe controlled range when surface conduction is suspected. Use low-stress scanning probes, guarded segmented fixtures, emission-free thermography only if sensitivity permits, or optical reinspection at the mapped minimum-clearance and film-edge coordinates.

Correlate the failing side with earlier maps:

  • Step 15: minimum foil clearance, proud foil, fracture extension, debris, terminal lift;
  • Step 16: film overreach, island, edge bead, lifted film, or unresolved foil crossing;
  • Step 17: local thickness maximum, composition anomaly, low sheet resistance, or measurement damage.

Do not scrape, cut, or burn open the bridge before nondestructive evidence and coordinates are secured. If destructive cross-sectioning is required, use a sacrificial failed sample and preserve its genealogy.

## 6. Set acceptance limits from the next tests, not convenience

Define maximum permitted screening current $I_{max}(V,t,T,RH)$ and minimum isolation resistance before testing. The limit must be comfortably below the current that would corrupt Step 20’s control-leakage measurement or create apparent modulation in Step 22. Include instrument uncertainty and blank variability:

$$ I_{DUT,UB}=|I_{meas}-I_{blank}|+U_{meas}+U_{blank}. $$

Pass only when $I_{DUT,UB}\le I_{max}$ at every required voltage, polarity, and control pair. If the result is limited by the fixture floor, improve the fixture or report an unresolved bound; do not convert insufficient sensitivity into a pass.

## 7. Minimum Step 18 record

Retain:

  • device genealogy and registered Step 15–17 maps;
  • complete 11↔12, 11↔13, and 12↔13 connection matrix, including unused-terminal state;
  • instrument models, calibration state, integration time, ranges, voltage burden, compliance, and guard topology;
  • fixture-open and insulating-blank traces through identical cables and switches;
  • voltage staircase, polarity, dwell and settling criteria, full current transients, and zero returns;
  • temperature, humidity, light, elapsed time, cleaning state, and cable-motion controls;
  • corrected currents, uncertainty bounds, $R_{iso}$ values or lower bounds, and linear-fit residuals;
  • every abort, compliance event, offset shift, hysteresis signature, and localization attempt;
  • pass, hold, or reject disposition for both forbidden control pairs.

The historical geometry is grounded in [J. E. Lilienfeld, US Patent 1,745,175](https://patents.google.com/patent/US1745175A/en): the copper terminal coatings are out of direct contact with the aluminum foil while compound film covers the foil edge. The modern guarded-measurement practice follows the [Keithley Low Level Measurements Handbook](https://www.tek.com/en/documents/product-article/keithley-low-level-measurements-handbook---8th-edition), which documents fixture leakage, humidity, settling, shielding, and driven guarding as central controls in high-resistance and low-current measurements.

## Check Unintended Conductive Paths’ Place in the Process Lineage

Steps 15–17 established geometry, coverage, and material properties. Step 18 now proves that the intended 11↔12 film path has not been accompanied by a direct 11↔13 or 12↔13 bridge, within a declared low-stress detection bound. Only an isolated device proceeds to Step 19, where the main terminal current–voltage characteristic is measured systematically. Step 20 then measures normal control-electrode leakage under its dedicated protocol rather than inheriting an undiagnosed short.

Take lilienfeld 1925 check unintended conductive paths further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app