Bare Board
Bake Calculator

A bare PCB sitting in ambient air is continuously absorbing moisture into its laminate. Under normal storage conditions this is a slow process, and the amount absorbed is small enough that most boards can go straight to assembly without any issue. The problem arises when a board has absorbed more moisture than the laminate can safely release during the rapid temperature ramp of a soldering process.

What happens during soldering: reflow soldering drives a board from room temperature to peak temperatures of 230 to 260°C in a matter of minutes, depending on the profile. At these temperatures, moisture trapped inside the laminate converts to steam. Steam expands with considerable force. If the vapor pressure inside the laminate exceeds the mechanical strength of the resin-to-glass fiber bond, the laminate delaminates. In less severe cases, the result is measling, crazing, or blistering visible on the board surface. In the worst case, inner layer separation compromises electrical integrity and cannot be repaired. The same risk applies to wave soldering, where the board underside contacts liquid solder at around 250 to 260°C, and to selective soldering. Hand soldering with a well-controlled iron on individual joints is generally not a significant risk for bulk laminate moisture.

The risk is proportional to moisture content and heating rate: a board at 1500 PPM going through a well-tuned reflow profile with a gradual preheat will typically survive without damage. A board at 4000 PPM going through a fast profile with a steep ramp is at real risk. This is why the Aquaboard study tested multiple PPM thresholds rather than a single pass/fail limit. The 800 PPM target recommended by IPC-HDBK-001 is a conservative safe level for standard production. The 2400 PPM limit found acceptable in three successive reflow passes in the same study shows that there is engineering margin, not a cliff edge.

When baking is actually warranted: baking should be a triggered response to a specific condition, not a standing procedure applied to every board that arrives. Conditions that justify baking include: boards that have been stored without dry packaging for longer than their floor life allows, boards returned from the field or pulled from long-term storage, boards with known high-absorption materials such as polyimide flex or paper-phenol substrates, or boards where a previous assembly attempt produced delamination or blistering and moisture is the suspected cause. If boards arrive from a qualified fabricator in sealed moisture-barrier packaging with desiccant, and are used within a reasonable time after opening, baking adds no value and only introduces the risk of solderability degradation from the heat cycle.

Baking is a corrective measure, not a process step: every bake cycle applies thermal stress to the surface finish and the laminate. HASL and immersion tin finishes degrade measurably with each bake. OSP coatings are directly damaged by baking. The correct strategy is to prevent moisture absorption through proper dry packaging from the fabricator through to the point of assembly, so that baking is never necessary in the first place. A process that requires routine pre-bake of every board has a storage or procurement problem that should be solved at the source.

The same logic applies with even more force to populated assemblies. Once components are soldered to the board, each unnecessary bake cycle ages the solder joints, stresses component packages, degrades any conformal coating already applied, and accelerates intermetallic growth at solder interfaces. Electrolytic capacitors are particularly sensitive to prolonged heat exposure. If a populated assembly genuinely needs moisture removal before a subsequent process step, such as before conformal coating application or before encapsulation, there is a real technical justification and the bake should be done at the lowest effective temperature for the shortest validated time. But if the bake is applied routinely to all populated boards at various stages without a documented failure mode it is preventing, it is process ritual with a real cost: it consumes oven time, adds handling risk, and slowly degrades the assemblies it is supposed to protect. The question to ask in any such process review is not "why should we remove this step" but "why does this step exist and what is the evidence that it works."

Diffusion is the movement of molecules from an area of high concentration to an area of low concentration, without any external force driving them. A familiar example: a sugar cube placed in coffee will dissolve and spread evenly through the cup on its own, even without stirring. The same principle governs how moisture moves in and out of a PCB laminate.

The epoxy resin in an FR-4 board is a porous polymer that binds moisture molecules to itself at the molecular level, much like wood absorbs water. When a dry board is placed in humid air, water molecules slowly work their way inward from all surfaces. When the same board is placed in a hot, dry oven, the process runs in reverse: moisture migrates from the interior toward the surfaces and evaporates off.

Why temperature is the critical variable: the speed at which molecules move through the resin increases exponentially with temperature, following the Arrhenius relationship. At 40°C, the diffusion rate is so slow that drying a standard FR-4 board to below 800 PPM takes around 7 days, and a high-absorption board with long storage takes up to 21 days. At 120°C, the same process is several hundred times faster, and a standard FR-4 board can be dried to below 800 PPM in approximately four hours. Lowering the surrounding humidity to zero does not change this rate. The moisture is already inside the resin; temperature is what gets it out.

This is also why a condensing drying cabinet at 40°C is an excellent storage device for already-dry boards, but is not capable of drying a board that has absorbed moisture during ambient storage. The cabinet prevents further absorption effectively. It does not undo absorption that has already occurred.

The three material groups are defined by how much moisture the board absorbs after 28 days at 85°C / 85% RH (the standard accelerated moisture soak condition).

Note: copper does not absorb moisture. A board with dense copper planes on outer layers has less available resin volume per unit area, which lowers its effective absorption rate. This is why some FR-4 designs fall into Group 1 despite being made from standard epoxy resin.

One Group 2 exception worth noting: Aramid/Epoxy (PCB II.11 in the Aquaboard study) reached 800 PPM in only 70 hours at ambient conditions, faster than typical FR-4. The aramid reinforcement is hygroscopic. If your boards use Aramid reinforcement, treat them conservatively within Group 2.

Surface finish affects whether and how you can bake a board. IPC-1601 Table 3-1 lists the following guidelines:

Every bake cycle degrades solderability to some degree. Baking is a corrective measure, not a routine process step. The right strategy is proper dry packaging from the board fabricator through to assembly, so that baking is never needed.

The calculation model was established for PCB thickness below 2.6 mm. For thicker boards, longer bake times will be required because moisture must diffuse a greater distance to reach the surface. As a rule of thumb, time scales approximately with the square of thickness.

The study population was predominantly E-glass/Epoxy (FR-4) and E-glass/Polyimide boards. Applying the model to less common materials within each group should be treated as a starting estimate, not a certified result.

Stacking boards in the oven significantly reduces baking efficiency. The Aquaboard study confirmed that air circulation between boards is required for the time estimates to hold. Use proper rack fixtures to ensure airflow on all board surfaces.

For process qualification or customer-traceable results, verify moisture content using IPC-TM-650 Method 2.6.28 (weight-loss method). Weigh the board before baking, bake, weigh again at regular intervals. The board is dry when weight stabilizes. An analytical balance accurate to 0.0001 g is required.

The floor life values shown are averages from the Aquaboard study. Individual board designs may behave differently, particularly boards with unusual copper density, unusual aspect ratios, or large exposed edge areas from routing.

Source: W. Horaud, S. Leroux, H. Fremont, D. Navarro. "PCB Materials Behaviours towards Humidity and Impact of the Design, Finishes, Baking and Assembly Processes on Assembly Quality and Solder Joint Reliability." Solectron / ACB / IXL. Presented at IPC APEX 2004.