On this page
Conductive films, moisture or contaminated media can create leakage between sensor nodes that were intended to be isolated. The consequence depends on where the path appears. Leakage from wiper to return pulls the output one way; leakage to excitation pulls it another; leakage between independent channels may defeat an assumption of separation. A useful margin calculation therefore uses a measured or bounded resistance at named nodes and solves the complete loaded circuit. It never converts one generic insulation value into a universal harsh-environment claim.
Key design decisions
- Identify leakage endpoints and polarity from the physical assembly.
- Solve normal and adverse output with the receiver connected.
- Compare uncertainty-bounded outputs with system-owned valid and diagnostic thresholds.
Name the surfaces and nodes joined by the unwanted path
Map conductors, printed resistors, terminals, substrate edges, protective layers, seals and housing references. State whether the path is surface, bulk, through residue or across a connector. A single insulation reading between shorted conductors may not represent wiper-to-return leakage at a mid-travel position. Multiple simultaneous paths require nodal analysis. Record bias polarity because ionic conduction can be direction and time dependent. The exposure state must include medium, concentration, temperature, duration and recovery condition.
Calculate the node shift created by a bounded shunt
At one wiper position, reduce the intended track to its Thevenin voltage and resistance, then place the leakage and receiver at their actual nodes. For a leakage path from output to return, receiver and leakage are parallel. Other endpoints need a different equation.
R_p=(R_L R_leak)/(R_L+R_leak); V_o=V_th R_p/(R_th+R_p); M=V_o-V_limit
- V_th and R_th describe the intended sensor network at one position.
- R_L is connected receiver resistance.
- R_leak is the measured or bounded output-to-return leakage resistance.
- M is signed separation from the applicable receiver limit V_limit.
Steady linear resistances in one named exposure state; transient polarization and multiple leakage endpoints are solved separately.
Calculate the loss of margin at a low-output station
Consider illustrative values V_th=0.80 V, R_th=2.0 kΩ and R_L=100 kΩ. Without leakage, output is about 0.784 V. A 30 kΩ leakage to return makes the parallel load approximately 23.1 kΩ and output about 0.737 V. If the system-owned lower valid threshold is 0.700 V, remaining signed margin is only 37 mV before measurement and component uncertainty. These numbers are not capability or acceptance values; they show why a resistance that appears large beside the track can still consume a significant threshold allowance.
Sweep leakage sensitivity through the full valid travel
The Thevenin voltage and resistance change with wiper position, so repeat the solution at endpoints, curve knots and high-impedance regions. Evaluate leakage to excitation and return separately, using polarity that moves output toward each threshold. For dual channels, include cross-channel paths with both instantaneous channel voltages. Present normal output, adverse output, uncertainty and remaining margin by position. A single worst insulation resistance without node identity cannot generate this map.
| Input | Evidence | Decision use |
|---|---|---|
| Leakage endpoints | Physical and electrical node map | Choose circuit topology |
| Resistance versus state | Controlled exposure measurement | Set adverse shunt |
| Receiver and bias | As-connected schematic | Solve actual output |
| Threshold and uncertainty | System interface requirement | Calculate remaining margin |
Treat polarization and wetting time as time-dependent evidence
Leakage may fall during wetting, rise with contamination concentration, or recover after drying. Measure current or resistance at defined times and bias rather than reporting only a convenient final reading. Receiver sampling can observe a transient that a slow insulation meter misses. If capacitance matters, model settling and filter response separately from steady resistance. Reversing bias can reveal asymmetry but may change the surface state. Preserve the sequence so comparisons are interpretable and safe handling is governed by the responsible laboratory.
Distinguish media leakage from terminal and track changes
A shift that follows added shunt resistance in a dry circuit supports the predicted topology. A common reference movement may instead come from terminal or return resistance. A position-local residual persisting under high insulation can involve the printed curve or contact. Measure accessible node currents, excitation and return while holding mechanical position independently. Guarded fixtures and blank coupons can help locate leakage, but fixture contamination can create its own parallel path. Capture the as-found wet state before cleaning or disassembly destroys evidence.
Validate with controlled shunts and the actual exposure sequence
First insert known safe resistors between the specified nodes and verify measured output against the circuit model across selected positions. Then run exposed specimens and unexposed controls using the approved media procedure. Compare leakage measurement and connected output on a common timeline, including recovery. Use multiple specimens when a population decision is required. Chemical safety, pressure and energized-media risks require facility-owned controls. A short test demonstrates only its defined condition and cannot establish field lifetime without a justified model.
Release leakage paths and margin ownership explicitly
The interface record should define protected nodes, minimum insulation or maximum leakage for named states, measurement voltage, polarity, timing, receiver circuit and threshold owner. Drawing data should show conductor spacing, glaze or protective coverage, edges, terminals and seal interfaces. ChipSimple can review a card against supplied materials and geometry; enclosure sealing and complete application qualification remain system responsibilities. Reopen the analysis after changes to media, seal, edge, coating, terminal, bias, track, receiver or diagnostic logic.
Close the remaining implementation and validation risks
Calculate leakage beside source tolerance, receiver variation, contact resistance, terminal drop, conversion uncertainty and temperature effects. Do not consume the same threshold allowance twice through a shared reference. Show signed high-side and low-side margins because one path can improve separation from one limit while approaching the other. Cross-channel leakage can move paired outputs plausibly together and requires architecture review beyond a single-channel calculation. Reverse-calculate the resistance that would reach each threshold under adverse normal tolerances, then compare it with exposure evidence. Extending glaze or sealant may lengthen a surface path but can create a contact edge, trap liquid, interfere with joining or concentrate stress. Evaluate coverage registration, thickness, terminations and substrate edges on actual artwork. Any process change needs compatibility evidence and renewed electrical characterization; dry insulation alone does not prove wet performance.
Complete the page-specific release closure
For each proposed mitigation, repeat the whole margin map. Guarding may redirect leakage but add capacitance; protective coverage may alter contact clearance; a stronger pull resistor can improve leakage tolerance while increasing contact current and loading error. State those trade-offs in circuit terms. Verify normal start-up, steady operation and diagnostic switching because node potentials change by mode. Preserve leakage current traces rather than only pass or fail, and compare exposed specimens with fixture blanks. If recovery occurs, record its time dependence instead of treating the last dry value as proof that no wet-state risk existed.
Provide the leakage topology and receiver thresholds
Readout-margin review needs measured exposure resistance tied to exact electrical nodes.
- Sensor schematic, terminal map, track values, excitation and receiver modes.
- Media, exposure, bias, temperature, duration and recovery definition.
- Leakage endpoints, measured bounds and time-dependent records.
- Valid or diagnostic thresholds, uncertainty allocation and validation owner.
The drawing-upload form loads as you reach this section.