Every claim a fabricator makes about a board is backed by a test, and IPC-TM-650 is the manual that defines how those tests are run. When two suppliers quote different values for the same laminate, or a customer asks why the measured Dk does not match the datasheet, the answer is usually in which test method was used and at what frequency.
This page maps the tests that matter on RF, high-speed and high-reliability programs to what they actually prove.
1. What IPC-TM-650 is
It is a collection of test methods, not a specification. It tells you how to measure something; it does not tell you what value is acceptable. Acceptance limits come from IPC-6012, IPC-4101 or the customer's drawing.
That distinction matters in a supplier discussion. "We meet IPC-TM-650" is meaningless on its own. The useful statement is "we measure Dk by IPC-TM-650 2.5.5.5 at 10 GHz, and the result is within the drawing limit".
2. Dielectric properties
| Property | Method | What it tells you |
|---|---|---|
| Dk and Df, clamped stripline | IPC-TM-650 2.5.5.5 | Baseline material Dk and Df, comparable across suppliers |
| Dk and Df, split post resonator | IPC-TM-650 2.5.5.6 | Fast measurement on small samples, 1 to 30 GHz |
| Dk, full sheet resonance | IPC-TM-650 2.5.5.7 | Large-area uniformity and glass weave effects |
| Volume and surface resistivity | IPC-TM-650 2.5.17 | Insulation quality after humidity |
Compare two materials only at the same method and frequency. A Df quoted at 1 GHz and one quoted at 10 GHz are not the same number and cannot be compared directly.
3. Thermal and mechanical
Tg by DSC or TMA. Two different methods give two different numbers for the same laminate, because they measure different transitions. A datasheet that does not state the method is not comparable.
Td, decomposition temperature. Relevant for lead-free assembly and for thick boards that see multiple reflows.
CTE, X/Y and Z. The Z-axis value drives plated hole reliability; the X/Y value drives board growth and connector alignment.
Peel strength. Copper-to-laminate adhesion, before and after thermal stress. Low peel strength shows up as pad lift during rework.
Thermal stress and IST. The test that finds barrel cracks. Thermal stress is a short, severe shock; interconnect stress testing cycles the board many times and is the better predictor of field life.
4. Moisture and insulation
Moisture absorption. Measured by weight change after conditioning. On RF materials this predicts frequency drift in humid environments; on high-voltage boards it predicts leakage.
Surface insulation resistance. Measured after humidity and bias, this is the test that catches flux residue and contamination. A board can pass ionic contamination and fail SIR.
CAF resistance. Conductive anodic filament growth between vias under bias and humidity. Relevant for fine-pitch, high-layer-count boards with small via spacing.
5. Plating and holes
Microsectioning. The method for measuring plating thickness, annular ring, etchback and voids. Everything about hole quality is decided here.
Plating adhesion and ductility. Relevant on thick boards and on high-CTE materials where the barrel is stretched.
Thermal cycling and thermal shock. Different ramp rates and dwell times, and they find different failure modes. Know which one the customer's qualification uses.
6. Solderability and finish
Solderability. Wetting balance or dip-and-look, after the specified ageing. This is what predicts whether the board assembles after six months in storage.
Finish thickness. X-ray fluorescence for gold and nickel, coulometric or XRF for tin and silver. Thickness has to be measured, not inferred from the process time.
7. How to use the manual in a supplier discussion
Ask which method, at what frequency, on what thickness, with what sample size, and what the acceptance limit is. Then ask for the raw data, not the summary table.
Three answers tell you whether the supplier understands the tests: they name the method without prompting, they state the frequency, and they distinguish between a material property and a board measurement.
8. Common confusions
Dk measured on a coupon is not the Dk the designer used, because the coupon includes plating and etch effects. Tg by DSC is not Tg by TMA. A thermal stress pass is not a thermal cycling pass. Ionic contamination pass is not SIR pass. A solderability test on fresh laminate does not predict shelf life.
Each of those has caused a program to accept a material that later failed, and each is avoidable by asking which test was run.
9. Summary
IPC-TM-650 tells you how to measure. The specification tells you what passes. When comparing suppliers or materials, match the method, the frequency and the sample condition, and ask for raw data. The numbers on a datasheet are only as useful as the method behind them.
Related reading
- IPC-6012 vs MIL-PRF-31032: Which Specification Applies
- J-STD-001 vs IPC-A-610: Process Standard or Product Standard?
- IPC-6012 Class 3 Compliance Checklist
- IPC Class 2 vs Class 3: What Actually Changes
- Microsection Inspection: What to Check and How to Read the Report
- High reliability PCB manufacturing
- PCBA assembly services
