PCB bow and twist measurement is useful only when the result includes the method, board condition, reference dimensions, and acceptance basis. Bow describes curvature primarily along a board’s length or width. Twist describes a departure from flatness in which the board’s corners or opposite regions do not lie in the same plane.
Neither a visible curve nor a percentage alone establishes a defect. A bare board can meet its agreed room-temperature requirement yet behave differently during reflow. Conversely, a board that appears uneven in an assembly fixture may need to be evaluated with the specified bare-board method before the supplier can assess conformity.
For design choices that influence flatness, see warpage control through PCB stackup and materials.
Bow, twist, and what the measurement represents
A bowed rectangular board has a broad curvature: imagine a shallow arc along its length, width, or both. The highest departure may be near the center, although the actual shape depends on the board and the reference plane.
A twisted board has a rotational or saddle-like departure from a plane. On a rectangular board, one corner may sit at a different height relative to the other corners, or opposite areas may rise and fall. A board can exhibit bow and twist simultaneously.
These descriptions help select where to look, but they do not replace a defined test. The reported value depends on:
- The surface or geometric plane used as the reference.
- The measured locations and direction of displacement.
- The reference length used for any percentage.
- Board orientation and test condition.
- Whether the sample is an individual board, a panel, or a rigid portion of a rigid-flex construction.
A maximum gap, a local height-map range, and a percentage calculated from a reference length are not automatically interchangeable. Record what each result actually represents rather than labeling every flatness number “bow and twist.”
Choose the method and acceptance basis first
IPC, IPC-TM-650 Test Method 2.4.22C, Bow and Twist, is a test method for specified bare rigid boards, panels, and rigid portions of rigid-flex boards. Revision C explicitly excludes special considerations for assembled boards. It also calls for care when evaluating nonrectangular samples.
Consult the full method and the revision invoked by the purchase documentation before testing. Do not infer the complete procedure from a summary, a supplier’s example, or the calculation in this article.
IPC, IPC-9641, High Temperature Printed Board Flatness Guideline, addresses high-temperature local flatness evaluation. Its table of contents distinguishes that subject from the room-temperature bare-board method in IPC-TM-650 2.4.22. IPC-9641 is therefore a different evaluation context, not an automatic replacement for the agreed room-temperature acceptance test.
A test method describes how a result is obtained. It does not, by itself, settle every contract question. Before release, the customer and supplier should identify:
- The applicable product specification and revision, if any.
- The drawing requirement and revision.
- The measurement method and revision.
- The acceptance limit, units, and sample population.
- The document-precedence rule and how conflicts will be resolved.
If the drawing and another invoked document appear to disagree, obtain a documented disposition before accepting or rejecting product. Do not substitute a published example threshold for the limit agreed on the order.
Measurement decisions by situation
| Situation | Measurement or reporting decision | Acceptance question |
|---|---|---|
| Rectangular bare board at incoming inspection | Identify the board, orientation, reference plane, measurement locations, and method revision. | What limit and sampling plan apply to this delivered board? |
| Manufacturing panel | Record the panel outline, rails and other relevant features, and whether the result is panel-wide or local to an individual board area. | Is the requirement for the panel, the separated board, or both? |
| Rigid portion of a rigid-flex board | Identify the rigid region evaluated and the method’s applicable provisions. | Does the drawing define a requirement for that region and its boundaries? |
| Nonrectangular board | Define the relevant outline, reference dimensions, and evaluation locations; review applicability with the supplier. | Has the interpretation been agreed for this geometry? |
| Populated board after reflow | Identify it as an assembled-board observation, with its own measurement setup and thermal history. | Is there a separate assembled-condition requirement or investigation plan? |
| High-temperature local flatness question | Define the temperature-dependent question and consult the applicable high-temperature guidance. | What locations, conditions, and decision criteria were agreed for this evaluation? |
Rectangular boards: define the plane and orientation
A reference plane is the basis from which displacement is assessed. Depending on the agreed method and setup, the record should identify the measuring surface or geometric reference, how the board is presented to it, and the locations used to determine departure.
For a rectangular sample, report the length and width directions and identify the tested face. If the method calls for more than one orientation or reading, retain each result rather than reporting only an unexplained maximum. A sketch or coordinate convention can make it clear which corner, edge, or board face corresponds to each observation.
Surface features can complicate interpretation. Copper patterns, finishes, markings, and other features may affect where a measurement is taken or what a probe contacts. Describe relevant contact locations and follow the specified method rather than silently changing the reference.
Do not add a fixture, restraint, or correction simply to obtain a convenient reading. Any support or restraint used should be consistent with the applicable method or an explicitly agreed alternative and should be recorded.
Panels are not just large individual boards
A panel can have an overall shape that differs from the local shape of any board within it. Rails, array geometry, routed features, and the locations of individual circuits can affect the result. Measuring only the panel outline does not necessarily establish the flatness of every separated board.
The release documentation should say which item is being accepted:
- The delivered panel before separation.
- A defined board area within the panel.
- An individual board after separation.
- More than one of these, with separate methods or criteria where needed.
Identify the panel side, array position, and relevant rail or routing condition in the record. If comparison after separation is required, preserve the identity of the originating panel and board position. The PCB panelization design guide covers related panel decisions; the key here is making the measured item unambiguous.
Percentage arithmetic: an illustration, not a test procedure
Suppose an agreed calculation uses a measured displacement of 0.40 mm and a reference length of 100 mm:
(0.40 mm / 100 mm) × 100 = 0.40%
The units cancel, producing a dimensionless percentage. This example assumes that the displacement and reference length have already been selected appropriately for the calculation.
It is not an IPC measurement procedure, a statement that 100 mm is the correct denominator for a particular board, or a pass/fail determination. The applicable method and agreement determine how the displacement and reference length are defined. Without those definitions, two percentages may look comparable while describing different measurements.
Build a measurement record someone else can use
A useful report lets the receiving engineer or supplier understand what was tested and reproduce the comparison under the agreed conditions. The fields below are a starting point; the applicable method and purchase documents may require additional information.
| Record field | What to capture |
|---|---|
| Product identity | Part number, drawing revision, order or lot identifier, and sample identifier |
| Item tested | Bare individual board, panel, or identified rigid portion; panel position where relevant |
| Stage and condition | Incoming bare board, after separation, after assembly, or another clearly named stage |
| Method and acceptance | Test method and revision, applicable drawing or specification, limit, and document-precedence reference |
| Sample selection | Lot population, quantity measured, selection approach, and sampling plan or agreement |
| Geometry and orientation | Board or panel dimensions, tested face, axes, corners, and relevant outline features |
| Reference and setup | Reference plane or surface, measurement locations, supports or fixtures used, and equipment identification as applicable |
| Environmental and history information | Relevant measurement conditions, time or stage of testing, and known handling or thermal history |
| Raw results | Individual displacement readings, units, locations, and orientations |
| Calculated results | Formula, reference length, units, and resulting percentage if required |
| Disposition | Applicable criterion, result against that criterion, unresolved questions, and approving authority |
Record raw readings as well as calculated values. If the acceptance decision is pending a clarification, say so; do not mark an item conforming or nonconforming against an assumed limit.
Sampling deserves particular attention. A measurement on one board is evidence about that sample under the stated conditions, not automatically about the entire lot. The customer and supplier should agree on the sample quantity, selection, and response to a nonconforming result before using the inspection outcome for lot disposition.
Incoming bare board versus post-reflow observations
The most common comparison error is to put a bare-board acceptance result and an assembled-board observation in the same column without explaining how they differ.
| Question | Incoming bare-board check | Post-reflow or assembly investigation |
|---|---|---|
| What is being assessed? | The specified bare board, panel, or applicable rigid portion at the defined incoming stage. | An assembly or board affected by additional process steps, components, and handling. |
| Which basis applies? | The invoked bare-board method, revision, and acceptance documents. | A separately defined measurement and decision basis appropriate to the assembly question. |
| What history matters? | Known manufacturing, packaging, transport, and handling history relevant to the agreement. | The above history plus available assembly-process and thermal-cycle information. |
| What can the result show? | Whether the sampled item meets the agreed bare-board criterion. | What shape or displacement was observed under the stated assembly-stage conditions. |
| What can it not prove alone? | That every location will remain flat at reflow temperature. | That the bare-board supplier caused the observed condition or that a room-temperature bare-board limit was exceeded. |
For a useful investigation, retain traceability between the incoming sample and the later assembly where practical. Record whether the later observation was made at room temperature or during a defined thermal condition, what equipment and reference were used, and whether the board remained in a panel or had been separated.
A single before-and-after number is not enough if the setups differ. Different reference planes, board faces, measurement locations, and sample conditions can create an apparent change that is not directly attributable to reflow.
Connect flatness to assembly risk without overstating it
Board shape can affect how reliably an assembly is processed and inspected. A panel that does not sit as expected in tooling may behave differently during printing or placement. Local board shape and component coplanarity can also affect solder-joint formation, particularly where joint access and inspection are difficult.
That does not mean bow or twist alone proves a soldering defect. Paste deposition, component condition, placement, thermal profile, support, and inspection method may also matter. A suspected BGA issue, for example, needs evidence from the assembly and solder-joint investigation, not just a bare-board percentage. See BGA soldering defects and the overview of PCB inspection for related inspection considerations.
Ask the assembly team which specific operation or location is at risk. A panel-wide room-temperature result may be relevant to handling, while a local high-temperature question near a component may require a different evaluation. Define that question before selecting the measurement.
Isolate the cause and hand off the evidence
When a result is unexpected, separate conformity from cause:
- Confirm the requirement. Retrieve the drawing, purchase order, applicable specification, method revision, and any written clarification.
- Confirm the sample and stage. Identify whether the measurement concerns the delivered panel, a separated bare board, or an assembly.
- Review the measurement. Check orientation, locations, reference, raw readings, calculation, and whether the agreed method was followed.
- Compare like with like. Examine other identified samples from the same lot or panel using the agreed plan; do not assume an unplanned sample represents the lot.
- Preserve traceability. Retain identifiers, available measurement records, and relevant process history for the investigation.
- Agree on disposition. The customer’s designated authority and supplier should document acceptance, further investigation, or another agreed action.
Potential contributors span board construction, panel geometry, fabrication history, assembly heat exposure, and handling. Those possibilities guide investigation; they do not assign fault. Do not apply a universal rework, cure, or bake instruction to a board with an unexplained flatness result. Any proposed action needs technical review and agreement for the specific product.
Release and RFQ fields
For a new order, include the flatness requirement in controlled release documents rather than relying on an informal note. For an inquiry, the following information helps the supplier identify questions before the order is placed:
- Part number, revision, outline, and whether delivery is as individual boards or panels.
- Applicable drawing, product specification, and document-precedence rules.
- Required test method and revision, including how any nonrectangular or rigid-flex region is to be evaluated.
- Acceptance criterion, units, measurement stage, and whether it applies to the panel, board, or both.
- Required sample quantity, selection basis, reporting fields, and disposition process.
- Relevant assembly concern, such as a known fixture interface or location requiring further evaluation.
- A designated technical contact and approval path for discrepancies.
The customer is responsible for stating the intended requirement and acceptance authority; the supplier should review the request, identify ambiguities, and perform and report agreed work within the project scope. Assembly-stage evaluation and decisions should be assigned to the responsible assembly party rather than assumed to be part of bare-board acceptance.
Discuss requirements through HilPCB’s PCB manufacturing service, submit the controlled details with a quote request, or contact HilPCB to clarify the measurement and handoff before release. Specific testing and reporting remain subject to the project agreement.
FAQ
Is bow the same as twist?
No. Bow is broad curvature along a board direction; twist is a departure in which corners or opposite regions do not share a plane. A board can have both.
Does IPC-TM-650 2.4.22C set the acceptance limit?
It is a test method. Determine the applicable acceptance limit from the invoked product documents and project agreement, including any document-precedence rule.
Can a panel result represent every separated board?
Not automatically. Panel-wide shape and local board shape are different measurements. The requirement should identify the item and stage being accepted.
Can a passing bare-board result guarantee flatness during reflow?
No. A room-temperature bare-board result does not establish local flatness throughout a thermal cycle. Define a separate high-temperature or assembly-stage evaluation if needed.
What should be sent with a flatness complaint?
Send the part and lot identification, applicable requirement, sample stage, method and orientation, raw readings, calculation, and relevant photographs or measurement records if available. Identify any unresolved comparison conditions.