吃瓜tv 路 ENGINEERING REFERENCE
Microvia Aspect Ratio: Calculator, Tolerances & Limits
Calculate depth divided by the agreed diameter, carry dimensional bounds through the ratio, and keep geometry review separate from process qualification.
Original engineering guidance 路 References checked September 27, 2026
Microvia aspect ratio is depth divided by a defined diameter
Microvia aspect ratio compares an actual via depth with a specified hole diameter using matching units. The arithmetic is simple; the measurement convention is the part that must be agreed. A tapered microvia can have an entry diameter, a target diameter and other locally measured dimensions. A ratio is ambiguous unless the drawing or process requirement identifies which diameter and depth reference it uses.
Do not divide the total board thickness by a microvia opening unless that thickness is genuinely the intended span. A short connection between two copper layers is a different feature from a hole through the complete board. This page owns geometry and tolerance arithmetic. Use the PCB microvia structure guide for entry/target capture and chain arrangements, and HDI PCB manufacturing for a construction review.
Define the numerator and denominator before calculating
Record the start and stop reference planes for depth. Then state the diameter location and measurement stage. For example, an opening before plating and a remaining lumen after plating describe different dimensions. A fabricator's process limit cannot be meaningfully compared with a ratio that uses a different convention, even if both values are displayed with the same number of decimal places.
| Input | Clarification needed | Common error |
|---|---|---|
| Via depth | Actual span and reference planes | Using total finished-board thickness |
| Hole diameter | Entry, target or another agreed convention | Substituting the smallest visible number without agreement |
| Measurement stage | Before or after relevant processing | Mixing drill data with a finished dimension |
| Tolerance | Bounds and whether they are independent | Treating nominal values as worst-case values |
| Review limit | Source, revision and applicable construction | Treating a calculator default as an IPC requirement |
| Result purpose | Geometric discussion or specified comparison | Calling a ratio result manufacturing approval |
Define the depth and diameter before calculating
Write the ratio as h divided by d, where h is the specified depth and d is the diameter chosen for the comparison. Both must use the same units. A tapered microvia can have different top and bottom diameters, and a drawing may also distinguish the formed opening from the finished plated geometry. Substituting one for another can change the ratio substantially without changing the physical hole.
For example, an illustrative depth of 75 micrometres divided by a 100 micrometre entrance diameter gives 0.75. Dividing that same depth by a 60 micrometre bottom diameter gives 1.25. Neither arithmetic result is inherently the correct capability comparison until the fabricator's definition is known. Label the denominator in the calculator output and on the drawing. Record entrance, bottom or finished diameter explicitly instead of an ambiguous 鈥渄iameter鈥� label.
Also define where depth begins and ends. A dielectric separation, a drilled feature depth and a distance that includes copper may not be interchangeable. Ask for a cross-section that shows the dimension's endpoints. The HDI manufacturing review should establish the relevant process definition. A useful tool result reads 鈥渉/d using stated entrance diameter and stated depth,鈥� followed by the dimensions and units. It should not show an unexplained green 鈥渕anufacturable鈥� label. Clear definitions make the result reproducible and allow a supplier to explain why its published capability uses a different reference. That explanation is more valuable than debating two ratios calculated from different physical quantities.
Calculate nominal and bounded aspect ratios
The worksheet calculates the nominal ratio and conservative bounds from your entered depth and diameter half-tolerances. It also compares the highest ratio with a limit that you supply and identifies the minimum diameter or maximum depth consistent with that limit. The tool does not supply a universal IPC threshold or a 吃瓜tv manufacturing capability.
Microvia Aspect Ratio and Tolerances
For the illustrative example, depth is 80 卤 10 碌m and the agreed diameter is 100 卤 10 碌m. The nominal ratio is 0.80. The highest bounded ratio is 90 / 90 = 1.00, while the lowest is 70 / 110, approximately 0.636. A nominal calculation alone therefore misses the full range permitted by those entered dimensional bounds.
The example limit of 1.00 is only an illustrative user input. It is not automatically taken from IPC, a supplier specification or 吃瓜tv capability data. In project mode, enter the limit and source confirmed for the actual process. If those are unresolved, ask for clarification before treating the interval comparison as a released design decision.
Calculate a bounded interval instead of only a nominal value
When dimensions have stated bounds, the highest ratio occurs at the largest depth and smallest positive diameter, assuming those bounds can occur together. The lowest ratio occurs at the smallest depth and largest diameter. This monotonic relationship makes a transparent worst-case interval possible: lower ratio equals h-min divided by d-max; upper ratio equals h-max divided by d-min. It is a bounding calculation, not a probability distribution.
Consider an illustrative depth of 75 plus or minus 8 micrometres and entrance diameter of 100 plus or minus 10 micrometres. The nominal ratio is 0.75. The bounded interval is 67/110 to 83/90, approximately 0.609 to 0.922. These inputs are invented for demonstration and are not a manufacturing specification. If a process discussion mentions a limit, compare it using the same definition and ask whether the quoted dimensions represent guaranteed finished bounds or preliminary targets.
Do not combine tolerances that refer to different stages without explanation. A dielectric material tolerance before lamination may not directly represent finished microvia depth. A laser target diameter may not be a finished inspection range. If the process data provide a directly measured ratio or a correlated geometry envelope, retain that evidence rather than forcing it into an independent-input model. The purpose of the interval is to reveal sensitivity and missing information. It should identify which input needs clarification, not create a false statistical confidence level. Record the source of every bound and keep the calculation with the exact stackup revision so later material changes cannot leave an obsolete result attached to the board.
Understand what the interval does and does not represent
The upper bound combines maximum depth with minimum diameter. The lower bound combines minimum depth with maximum diameter. The calculation assumes those extremes can be considered independently. It does not assign probabilities, fit a manufacturing distribution or estimate the percentage of boards that will meet a limit.
Both dimensions must remain positive throughout their entered ranges. The worksheet rejects a diameter tolerance equal to or greater than the nominal diameter because that would create a zero or negative denominator. It similarly rejects a depth interval reaching zero or below. These input checks protect the arithmetic; they do not validate the underlying measurement or drawing tolerance.
If the process team has justified correlated bounds, use the appropriate engineering method rather than interpreting this conservative interval as a statistical prediction. Preserve the distinction in reports. 鈥淗ighest ratio under entered independent bounds鈥� is more informative than an unexplained pass/fail color, because it tells the reviewer exactly which comparison has been performed.
Use the result to frame a specific design discussion
Suppose the example's maximum depth remains 90 碌m but the agreed project limit is 0.75. The minimum diameter consistent with that combination is 90 / 0.75 = 120 碌m. If the actual minimum diameter is 90 碌m instead, the maximum depth consistent with that limit is 0.75 脳 90 = 67.5 碌m. These are alternative geometric targets, not automatic recommendations to enlarge the hole or thin the dielectric.
Increasing the opening may affect capture lands, clearances and local escape space. Reducing the dielectric span may affect impedance, material availability and the lamination sequence. The correct next step is to compare those consequences with the fabricator and electrical designer. Changing whichever field makes the calculator turn favorable can create a different problem elsewhere in the board.
The HDI PCB design guide addresses local routing space. The controlled-impedance page addresses electrical geometry. Keep the aspect-ratio brief connected to those decisions when a change affects both the interconnect and the surrounding stack.
Inspect the rest of the geometry around the ratio
Two microvias with the same ratio can differ in taper, land dimensions, target geometry, copper thickness and local interfaces. The ratio compresses several design choices into one number and therefore cannot describe all relevant features. Review the capture and target lands, the relationship to surrounding copper and the available registration margin. Include the via's location in the package escape, where space constraints may drive a very different land arrangement from a simple coupon.
For a stacked connection, identify each individual via and the surface on which it lands. Do not add the depths of several distinct microvias and compare the sum to a capability statement intended for one formed hole. Conversely, do not use an acceptable individual ratio to conclude that the complete stacked interface is qualified. The HDI sequence guide explains why formation order and interfaces belong in the construction definition.
Filled, capped and open features should be distinguished according to the proposed route. A CAD symbol may not show the surface condition needed for a later connection or component pad. Specify the function rather than assuming the supplier will infer it from the package type. If the goal is to escape a dense BGA, provide representative geometry so the fabricator can review the interaction between hole, lands, traces and spacing. A ratio calculation is most useful when it is one line in that complete review. It can expose a depth or diameter problem, but it cannot replace the drawing, process proposal or evidence that the interfaces survive the required thermal history.
A geometric limit belongs to a process and a definition
Published articles often present a single ratio as a rule for every microvia. That shortcut loses the construction, measurement basis and revision behind the claim. For a real project, ask which process and dimensional convention the limit applies to, whether it refers to a particular material or structure, and what additional requirements remain outside the ratio comparison.
The public identifies an HDI design-standard scope; it does not justify extracting an unverified acceptance number for this calculator. Use the licensed applicable standard and the project agreement for normative requirements. A historical revision reference should not be presented as the current required revision without checking the contract and publisher information.
Equally, a supplier's demonstrated example does not automatically establish a general capability for every stackup. Materials, target geometry, fill requirements and subsequent processing can differ. Ask for a construction-specific response and retain its conditions. The result of that review should guide the drawing, not disappear into a generic note that all vias satisfy a chosen ratio.
Keep interface evidence separate from the ratio
A favorable depth-to-diameter relationship cannot establish target-interface cleanliness, adhesion, copper continuity or fill condition. It also cannot qualify several connected microvias simply because each individual hole has the same nominal ratio. A chain introduces relationships and interfaces that are absent from the one-hole calculation.
Review the complete chain and its intended thermal exposures. The microvia qualification and evidence checklist focuses on construction identity, test purpose and records. It complements the arithmetic here rather than repeating it. Suitable evidence needs to address the concern under review and correspond to the structure actually being ordered.
A room-temperature measurement and a prepared section can each answer useful questions, but their scope should remain visible. Do not turn the presence of a report into proof of every reliability requirement. Define the specimen, conditioning, method and decision criteria with the responsible engineering team. The IPC-TM-650 methods guide helps separate a test procedure from an acceptance requirement.
Ask for construction-specific qualification evidence
A published capability range describes a process under stated conditions. It is not necessarily a qualification statement for every material, via arrangement or assembly thermal sequence. Ask which geometry definition the supplier uses, which materials and copper construction are represented, and whether the proposed stacked or staggered arrangement is covered. Record any differences between the evidence and the intended build.
Separate the geometric check from evidence about interfaces and thermal performance. A board may pass a room-temperature electrical test while still requiring additional qualification for the application. The microvia reliability evidence guide provides a framework for discussing coupon construction, thermal exposure and monitoring. It does not imply that every microvia needs the same test or that all stacked microvias are unreliable.
For a prototype, agree what is being learned. A development build can establish routing feasibility, assembly behavior or a specific qualification question, but one successful sample does not automatically establish production performance. Identify specimen count, construction revision and the actual conditions observed. If a geometry change is made to reduce the ratio, review the effect on land size, package escape and electrical performance rather than assuming it is a free improvement. Similarly, making a dielectric thinner can affect impedance and stackup balance. A useful engineering conversation links the ratio to the full design. The outcome may be an accepted construction, a revised geometry or a defined test need. Any of those is more defensible than a tool assigning approval from a single threshold with no process context.
The is historical evidence for asking about interface and thermal-test coverage; it does not establish a universal aspect-ratio limit.
Avoid preventable unit and precision mistakes
The worksheet uses micrometers for both dimensions so the ratio is dimensionless. If a drawing uses millimeters, convert both depth and diameter consistently before entering them. For example, 0.08 mm equals 80 碌m and 0.10 mm equals 100 碌m. Converting only one dimension creates a result wrong by a factor of one thousand.
Extra decimal places do not improve an uncertain measurement. Report precision appropriate to the drawing and measurement basis, and preserve the tolerance rather than rounding it away. A displayed value close to an entered limit should prompt review of the source dimensions, not a search for a rounding convention that makes the comparison appear favorable.
If different teams use 鈥渄rill size,鈥� 鈥渓aser opening鈥� and 鈥渇inished hole鈥� loosely, resolve the vocabulary before comparing spreadsheets. The calculator's diameter-basis field forces that discussion into the brief. A clear sentence about the convention is often more valuable than another decimal place in the result.
Reject invalid inputs and misleading precision
Use a positive depth and a positive diameter for this geometric comparison. A lower diameter bound that reaches zero makes the upper ratio undefined for this model; no finite upper ratio can be calculated from that interval. If a tolerance exceeds the nominal dimension, correct or clarify the invalid interval before calculating. Leave unknown dimensions open for review instead of treating an example default as a supplier recommendation.
Units need equal attention. Convert both dimensions to the same unit before division. A depth in micrometres divided by a diameter in millimetres without conversion creates a thousandfold error. Display the entered units and converted dimensions in the exported record. For mil inputs, use the defined conversion of 25.4 micrometres per mil. The dimensionless ratio should not carry a length unit, and a notation such as 0.75:1 should be explained consistently.
Use appropriate precision. Three significant figures may be useful for a calculation, but the underlying dimensions and measurement resolution determine what the result means. A value of 0.7499 should not be treated differently from 0.7501 against an unspecified nominal capability. The supplier must establish the acceptance definition and relevant uncertainty. When you change depth, diameter, units or tolerance, recalculate and replace any earlier exported conclusion that no longer matches. Keep the original dimensions with the output so another engineer can reproduce the arithmetic. These checks keep a mathematically correct formula from answering the wrong geometric question.
Send the dimensions and the unresolved question together
For an aspect-ratio review, provide the board revision, actual layer span, opening dimensions, tolerance basis and proposed process. Include a section drawing and identify any stacked or buried-via relationships. State the source of the review limit and whether the material and stackup are already approved or still open to alternatives.
The downloaded brief records only the current worksheet state. Editing an input clears the old result and disables the old download until the calculation is rebuilt, reducing the risk of attaching a stale comparison. Contact remains available without a calculation. Attach the controlled manufacturing data separately; no project values are automatically submitted to a form or analytics system.
Export the dimensions and the question together
The handoff should include via group identifier, start and end layers, depth definition, diameter definition, nominal values, bounds and units. Add the material and stackup revision, fill/cap requirements and the relevant package region. Attach a labelled cross-section or manufacturing drawing. A screenshot of a ratio alone leaves the most important assumptions invisible.
Use the worksheet to state the unresolved question precisely. Examples include whether the stated diameter range is the correct process reference, whether a thinner dielectric is compatible with impedance targets, or whether the proposed stacked interface is represented by existing evidence. These questions are specific enough for a supplier to answer without guessing the customer's intent. If an internal limit is used for preliminary screening, identify its source and keep it separate from the supplier's manufacturing acceptance.
Submit the package through manufacturing review while retaining the HDI context. Use the geometry calculator as a planning aid and retain its stated assumptions with the result. A geometry result should remain separate from the manufacturing decision and its evidence. Before release, reconcile accepted dimensions into the actual CAD and manufacturing files and rerun the check against that revision. The final record should make it possible to reconstruct both the arithmetic and the decision: what was calculated, which definition was used, what evidence was considered and what remains open. That is the practical value of a microvia aspect-ratio tool for an engineer or purchaser.
Aspect-ratio questions before release
Should the calculator always use the smallest hole diameter?
Use the diameter convention required by the agreed process criterion. A tapered hole can have several meaningful diameters. Choosing the smallest without documenting the convention may create an incompatible comparison rather than a more useful one.
Is a 1:1 ratio universally acceptable?
No universal approval is provided here. The example value is an editable teaching input. Confirm the applicable definition, process and criterion, then review capture geometry, interfaces and reliability separately.
Can reducing depth solve every microvia problem?
No. It changes one geometric relationship and may alter the stackup or electrical requirements. The proposed change still needs material, routing, process and evidence review. A ratio does not describe every feature that determines a successful interconnect.
Does the tolerance interval predict manufacturing yield?
No. It combines independent dimensional bounds without probabilities or correlations. It is a conservative geometry worksheet, not a statistical process model, capability analysis or guarantee of delivered yield.
Bring the requirement and the drawing together
Send the board revision, construction, quantity and the standard or engineering question you need reviewed. Identify unresolved requirements so the response can address them.
Send requirements for reviewOptional: attach the downloaded worksheet manually. This guidance does not certify a design, approve a shipment or replace the applicable authorized standards.