On this page
A nominal sheet-resistance value gives one nominal aspect ratio, but it does not show whether that geometry can tolerate the stated material and end-effect variation. A feasibility calculation produces an interval of acceptable ratios or an explicit empty result. That result helps reject an incompatible candidate before spending effort on artwork detail or trying to recover the value through excessive adjustment.
Key design decisions
- Distinguish a nominal solution from a ratio that works for every declared bounded input.
- Intersect the electrical interval with actual length and width constraints.
- Test dimensional tolerance around the chosen geometry before accepting the interval result.
Decide whether every allowed combination must fit
Begin with a finished resistance band from Rlow to Rhigh at one specified reference condition. The calculation asks whether one active aspect ratio N can meet that band for all allowed sheet-resistance and end-contribution values. This is a robust bounded question. It is more demanding than asking whether some favorable material value can produce an acceptable resistor at the same geometry.
Use intervals that describe the actual processed material and construction intended for the part. A supplier nominal value is not by itself a lower and upper production bound. If the available information is statistical, retain its distribution and confidence basis rather than relabeling a sample minimum and maximum as absolute bounds. The interval method remains useful for screening hypothetical requirements, provided those assumed inputs are clearly separated from measured production evidence.
Write the active-body and end terms separately
Represent the untrimmed rectangular element by R = Rs N + E, where N is positive, Rs is the processed sheet resistance and E is the combined signed end correction or physical end contribution under the declared length convention. Define Rs between Slo and Shi and E between Elo and Ehi. A physically positive resistance interval requires positive sheet-resistance bounds.
For independent bounded inputs in this monotone model, the smallest predicted resistance is Slo N + Elo and the largest is Shi N + Ehi. Even if the extreme inputs cannot occur together, this rectangular interval gives a conservative screen. Narrowing it through correlation requires a supported joint model. Do not simply pair low sheet resistance with a convenient high end term to make a candidate appear feasible without evidence that such compensation is dependable.
Solve the lower and upper inequalities
To keep the lowest prediction above Rlow, require N at least (Rlow − Elo)/Slo. To keep the highest prediction below Rhigh, require N no greater than (Rhigh − Ehi)/Shi. Intersect these inequalities with positive geometry and any pre-existing aspect-ratio restriction. If the calculated lower endpoint exceeds the upper endpoint, no single ratio meets the stated robust requirement.
An empty interval is an engineering result. It can arise from excessive sheet-resistance spread, large end variation or an unusually tight final band. Possible changes include reducing a justified input interval, changing the material candidate, enlarging the allowed resistance band or using a separately qualified adjustment route. The interval calculation does not decide which change the product may accept; it identifies the conflicting constraints so that decision can be made explicitly.
Nlow = (Rlow − Elo)/Slo; Nhigh = (Rhigh − Ehi)/Shi; feasible N ∈ [Nlow,Nhigh] ∩ (0,∞)
- Rlow and Rhigh define the required resistance band in ohms.
- Slo and Shi are positive lower and upper processed sheet-resistance bounds in ohms per square.
- Elo and Ehi bound the combined end term in ohms using the same active-length definition.
- N = Lactive/Wactive is the dimensionless rectangular aspect ratio.
The linear body-plus-ends model applies throughout the specified interval; N is positive, all extreme combinations are conservatively allowed, and trim and operating-condition shifts are outside this reference-state calculation.
Screen two material intervals for a one-kilohm requirement
Consider an illustrative required resistance band of 950 to 1,050 Ω. Assume the combined end term lies between 8 and 12 Ω. Candidate A has a processed sheet-resistance interval of 98 to 102 Ω per square. Its lower aspect-ratio endpoint is (950 − 8)/98 = 9.612245. Its upper endpoint is (1,050 − 12)/102 = 10.176471. A positive interval therefore exists before geometric fit is checked.
Candidate B is assumed to lie between 188 and 212 Ω per square with the same end interval. Its endpoints are approximately 5.010638 and 4.896226, respectively. The lower exceeds the upper, so this candidate has no robust solution. Its nominal values could still suggest N = (1,000 − 10)/200 = 4.95. That attractive nominal number does not repair the contradictory extreme inequalities. All intervals in this example are chosen calculation inputs rather than published paste performance.
| Candidate input | Required minimum N | Permitted maximum N | Electrical outcome |
|---|---|---|---|
| A: Rs = 98–102 Ω per square; E = 8–12 Ω | 9.612245 | 10.176471 | Nonempty interval |
| B: Rs = 188–212 Ω per square; E = 8–12 Ω | 5.010638 | 4.896226 | Empty interval |
| B evaluated only at nominal Rs = 200 and E = 10 | Nominal N = 4.95 | No bounded proof | Nominal target is insufficient |
Turn an acceptable ratio into a rectangle that fits
Suppose the illustrative active body may have length from 4 to 10 mm and width from 0.8 to 1.2 mm. With no additional coupled constraint, the available aspect-ratio range is 4/1.2 to 10/0.8, or approximately 3.3333 to 12.5. Candidate A's complete electrical interval lies inside this geometric range. These dimensions describe the example envelope and are not minimum-feature or manufacturing claims.
For a chosen N, obtain the allowable width interval by intersecting the width bounds with Lmin/N to Lmax/N. At N = 9.9, this gives widths from 0.8 to approximately 1.0101 mm. Choosing a 1.0 mm width gives a 9.9 mm active length. Reserve conductor overlap, terminal approaches and edge keep-outs outside that active rectangle. Using the complete substrate outline as Lmax would overstate usable length when those regions also need space.
Fit the entire dimensional envelope inside the ratio window
A nominal ratio inside the feasible interval is not sufficient when length and width vary. For the illustrative 9.9 mm by 1.0 mm body, assume length can vary by ±0.05 mm and width by ±0.01 mm. The conservative minimum ratio is 9.85/1.01 = 9.752475; the maximum is 9.95/0.99 = 10.050505. Both remain inside Candidate A's electrical interval, so this dimensional envelope passes the chosen bounded screen.
The corresponding extreme resistance predictions are approximately 963.743 Ω and 1,037.152 Ω after applying the sheet and end bounds. A different end-length definition or a correlation between printing width and film thickness could change the input model. Preserve dimensional and electrical uncertainty separately from the assumed production intervals. Near an interval endpoint, uncertainty in the inspected geometry may determine whether the part can actually be shown to satisfy the intended boundary.
Validate the corners of the input space
Build a comparison plan around the combinations that set the two resistance boundaries. Measure actual processed lengths, widths and resistance with the same terminal convention as the end correction. Include representative material and process conditions supporting the claimed sheet-resistance interval, plus geometries near the selected ratio limits. A single central sample verifies neither the low prediction nor the high prediction.
Compare the measured terminal resistance with Rs N + E and examine residuals versus body length, width and location. Persistent shape-dependent residuals indicate that the assumed constant end term or sheet-resistance model needs refinement. Sheet-resistance characterization itself requires a measurement geometry and method suited to the sample; methods developed for continuous semiconductor films do not automatically establish an electrical property for a short printed resistor with strong end interactions. Retain product-shaped correlation as part of the validation.
Diagnose a window that disappears after processing
A nominally suitable ratio that fails on both resistance extremes suggests that the material interval or geometric variation was understated. A failure concentrated in short bodies points toward the end term; a proportional shift across several ratios points toward sheet resistance or common width change. If a candidate becomes feasible only by averaging two unrelated paste grades, verify that the proposed material route actually exists and has controlled processing before using that interpolated interval.
Trimming can change the question by introducing an upward adjustment range and a new geometry. It must not be inserted as an unexplained margin into an untrimmed feasibility result. Likewise, TCR, voltage dependence, noise and powered stability remain companion requirements even when the reference-state interval closes. Deliver the surviving ratio interval, chosen dimensions, tolerance envelope and explicit rejected candidates. This makes sheet-resistance selection a reproducible constraint decision rather than a search for one convenient nominal square count.
Provide the aspect-ratio feasibility inputs
Send the resistance band and bounded process inputs needed to calculate a usable geometry interval.
- Required resistance band, reference state and whether every bounded combination or a stated statistical criterion must pass.
- Candidate processed sheet-resistance intervals and combined end-term intervals with their geometry and measurement definitions.
- Active length and width bounds, keep-outs, dimensional tolerances and any coupled area or layout restrictions.
- Representative geometry and resistance data, model residuals, adjustment route if any and acceptance ownership.
The drawing-upload form loads as you reach this section.