On this page
Two ceramic plates labelled 96 percent alumina can require different circuit decisions. The remaining constituents, finished surface and dimensional condition belong to the material specification, while the circuit experiences the plate after printing, firing and assembly. Select an identified grade by the requirements it closes. A higher composition percentage is useful only when the associated properties and processed interfaces address the actual constraint.
For a drawing-specific part, review the Alumina thick film PCB construction, product evidence and quotation inputs alongside this method. Prepare the dfm checklist with your operating conditions.
Key design decisions
- Keep mandatory electrical, geometric and interface requirements separate from commercial preferences.
- Compare candidate grades at the delivered thickness and relevant measurement temperature.
- Allow a substitute only after evidence connects its exact finish and material identity to the intended printed circuit.
A composition percentage does not identify the whole ceramic
An alumina percentage describes composition on its stated basis; it does not disclose every feature of the fired microstructure. Grain boundaries, residual phases, density and surface texture are separate characteristics. Two suppliers can therefore use similar purity labels for materials that are not interchangeable at a resistor interface. Begin the comparison with the actual grade code and document revision. If the specification names only a percentage, identify which functional properties the original designer expected that shorthand to protect.
Avoid turning uncertainty about the original material into an automatic demand for the highest available purity. That change can add an unexamined surface and process condition while leaving the original acceptance question unanswered. Ask whether the existing circuit is constrained by printing response, heat transfer, dielectric behavior, dimensional stability or handling damage. The answer decides which grade evidence is relevant. Cosmetic similarity between incoming plates cannot establish equivalence in any of those functions.
Separate disqualifying conditions from preferences
Create a short list of conditions that every candidate must satisfy before price, availability or polishing convenience is compared. A circuit with a soldered terminal may require demonstrated pad attachment after its assembly cycle. A high-impedance network may require leakage measurements in a specified environment. A close-fitting carrier may impose a mechanical envelope. These requirements are not mutually exchangeable: exceptional flatness cannot compensate for a failed insulation allocation.
Write each condition as a measurable observation with a specimen state. The phrase good adhesion leaves open the pull geometry, thermal history and failure location. Similarly, low loss requires a frequency and configuration. Unknown evidence should remain a separate outcome, rather than receiving an average score that lets the candidate pass. This structure gives purchasing a clear reason to request missing information and prevents a grade substitution from becoming an argument over a single headline property.
Compare conductivity together with delivered thickness
A grade with a higher conductivity does not necessarily create a lower through-thickness thermal resistance when the delivered plate is thicker. For a uniform heat path, the relevant material comparison is the thickness divided by conductivity and active area. Use conductivity at a temperature appropriate to the operating region. If the source is smaller than the cooling footprint, spreading must be assessed separately; this simple comparison cannot predict the hotspot from the substrate outline alone.
For an illustrative substitution, assume a uniform active area of 100 square millimeters. Candidate A is 0.63 millimeters thick with an assumed hot-state conductivity of 16 watts per meter kelvin; candidate B is 0.72 millimeters thick at 18 watts per meter kelvin. Their calculated resistances are 0.394 and 0.400 kelvin per watt. At 8 watts the ideal drops are approximately 3.15 and 3.20 kelvin. The nominally better conductivity gives no thermal advantage in this constructed comparison. These are calculation inputs, not properties assigned to a named alumina grade.
θ_ceramic = t / (kA)
- θ_ceramic: one-dimensional thermal resistance in K/W
- t: delivered ceramic thickness in m
- k: conductivity at the evaluated temperature in W/(m·K)
- A: uniform heat-flow area in m²
Steady heat flow normal to a homogeneous plate; identical active area and contacts. Lateral spreading, interface resistance and temperature-dependent gradients are outside this estimate.
Make the printed feature decide which surface matters
A substrate surface that supports one conductor may not preserve the same resistor response or fine insulating gap. Put the most sensitive printed feature on the candidate comparison coupon. Keep paste conditioning and thermal processing equivalent so a difference in the substrate is not hidden by simultaneous recipe adjustments. Where process adjustment is needed to make the substitute usable, treat that as a separate material-and-process candidate rather than claiming direct interchangeability.
Record which face was printed. Supplier descriptions can distinguish faces or finishing states, and an operator may unknowingly reverse a plate that appears symmetrical. Relate surface measurements to the direction and scale of the functional feature. An average roughness number describes only part of the topography and does not detect every isolated pit. Compare edge definition, continuity and the interface after firing, because a clean wet print is not the final material response that the circuit designer needs.
Turn a candidate list into an evidence comparison
The comparison table should reveal the reason a candidate remains open. Retain raw results and distinguish a supplier typical value from a measured receiving characteristic or a finished-circuit test. Do not place these three categories in one numeric column without identifying their meaning. A typical flexural value describes the supplier specimen and method; it does not establish an allowable concentrated clamp force on a machined circuit plate.
For dimensions, use the same datum and support convention for both candidates. For electrical checks, preserve temperature, test configuration and recovery state. A small reported difference is uninformative if method uncertainty is larger than the difference. The useful selection record can therefore contain a request for better discrimination, rather than a forced winner. The table below describes evidence categories and decision consequences; it does not supply acceptance thresholds for a product that has not been specified.
| Property or condition | Evidence needed | Selection consequence |
|---|---|---|
| Composition and grade | Identified specification and delivered-lot records | Controls the material being compared |
| Thickness and thermal data | Actual section with temperature-matched conductivity | Determines the ceramic contribution to heat transfer |
| Printing face | Surface characterization and representative fired feature | Checks whether the same circuit pattern remains usable |
| Mechanical fit | Drawing-based dimensions, support and edge inspection | Rejects an incompatible assembly envelope |
| Processed interface | Attachment and electrical results after assembly exposure | Determines whether the substitute is functionally acceptable |
Use the failure pattern to locate the missing requirement
Resistance moving across an entire substituted lot can implicate a changed substrate interaction, but it does not prove a composition cause. First compare film geometry, face assignment and thermal records. A defect concentrated along machining edges suggests a different investigation from a uniform electrical shift. A bond pad that lifts only after assembly can expose an interface weakness that incoming ceramic measurements never addressed.
Preserve the observed failure surface and its location before reworking the specimen. Ceramic fragments attached to a removed pad indicate a different failure plane from a clean separation at the film boundary. A shift that follows plate orientation may point toward texture or handling. A response that follows furnace position may have little to do with grade. These signatures are diagnostic leads, not automatic conclusions. Their purpose is to add the missing condition to the selection comparison without blaming purity for every change.
Validate a substitute against retained original material
Use retained original-grade material as a comparison whenever its identity and history are trustworthy. Build both groups through the circuit operations that can alter the decisive characteristic, and preserve substrate and print position in the record. If only old finished circuits remain, recognize that their age and process history complicate the comparison. A fresh substrate trial cannot be treated as a controlled substitute test against an undocumented historical assembly.
Qualification should challenge the requirement that made the grade choice important. For a thermal concern, measure the assembled temperature response under the agreed cooling condition. For a resistor concern, examine the electrical distribution through the relevant processing sequence. For a mechanical concern, include the actual support and attachment. The decision is stronger when independent lots confirm the observed behavior. Sample quantity and acceptance statistics belong to the agreed risk and evidence plan; this guide does not invent a universal qualification count.
Release a material identity that purchasing can preserve
The resulting material callout should connect an approved grade to the face condition, dimensions and evidence that supported the choice. Keep alternative grades individually identified. A generic note allowing equal or better purity can reopen all the hidden questions the review just closed. If commercial flexibility is required, define the functional equivalence tests and the approval responsibility before a purchasing substitution occurs.
Also identify changes that require reconsideration: a new surface treatment, altered thickness range, revised grade formulation or a different ceramic manufacturing location can affect the comparison basis. The request need not dictate a supplier factory recipe. It should describe the circuit feature and response that must survive the change. Supplying the original substrate specification, comparison results and failed-feature photographs lets a drawing-led review evaluate the real substitution task instead of interpreting the percentage on a purchase order as a complete engineering requirement.
Send the grade comparison requirements
Provide the existing material identity and the specific function that a proposed substitute must preserve.
- Original and proposed alumina grade documents, composition basis, delivered lots and any permitted substitutions.
- Drawing thickness, active heat-flow area, mounting contact and temperature conditions used in the thermal comparison.
- Functional face, finish measurements, sensitive printed features and the named paste and assembly sequence.
- Hard acceptance criteria, comparison specimens, observed failure locations and required validation evidence.
The drawing-upload form loads as you reach this section.