A shield can held only by solder joints may detach or shift during multiple reflow passes, hot-air gun rework, or when the product is dropped. Adhesive reinforcement adds an extra safeguard, but where the adhesive dots go and whether solder paste is printed on every pad should be decided by experiment. A common approach is to first build the “strongest” combination (solder paste on all pads + adhesive), validate that it passes reflow cycles, hot-air gun heating, and drop tests, and only then move step by step toward “less adhesive, less labor.”
For how to choose the shield can itself (one-piece, two-piece, shield clips, material and flatness), see “How to Choose a Shield Can“; for whether heavy components will fall off during double-sided reflow, see “How to Assess Component Drop in Double-Sided Reflow“. The figures in this article come from a production-line experiment record (with company names, customer names, and part numbers removed); they are a single case, not a general specification.
1. When to Consider Adhesive
Q: In what situations should a shield can get adhesive?
A: Three situations are most common: the product goes through multiple reflow passes (double-sided boards, or rework later in the process), components under the shield can may later need hot-air rework, or the end product has drop or vibration requirements. If the shield can goes through the oven only once and the product has no drop specification, getting the pad design and solder volume right is usually enough, and adhesive is not necessarily needed.
2. List the Combinations to Compare: How Solder Paste Is Printed × Where the Adhesive Goes
Q: Which approaches should the experiment compare?
A: Split “solder paste” and “adhesive” into two or three options each and cross them into several groups. The experiment we read was 2 × 2: solder paste was either “printed on all pads” or “not printed on some pads,” and adhesive was dispensed in two positions; a material-saving approach of “adhesive only at the four corners, none along the side walls” was added afterward.
Adhesive can be applied with an automatic dispenser, or by hand at the prototyping stage with a dispensing syringe; the adhesive’s curing conditions (temperature and time) should follow the adhesive maker’s data and be the same for every group — otherwise you are comparing degrees of cure rather than dispensing positions.
3. Validation Conditions: Reflow Cycles, Hot-Air Gun, Drop
Q: What conditions should be used for testing?
A: Set them according to what the product will actually encounter later; three are common: multiple reflow passes, hot-air gun heating (simulating rework), and drop testing. The conditions used in this experiment were:
Reflow: two levels, 3 passes and 6 passes.
Hot-air gun: 260 °C, two levels of 60 seconds and 120 seconds of heating.
Drop: 60 cm and 120 cm, 3 drops each.
3 units per group; all combinations crossed, 8 groups in total.
The judgment looks at whether the shield can detached or shifted, and whether components underneath shifted or shorted as a result.
4. Start with the Strongest, Then Move Toward Savings Once Everything Passes
Q: With so many combinations, which group should come first?
A: Start with the most conservative “solder paste on all pads + adhesive”; only once it passes is there a basis for trying less adhesive or less solder paste. If even the strongest approach fails, the problem is usually not the adhesive — go back and look at the pads, the solder volume, or the flatness of the shield can.
The results of this experiment: in the first round, all 8 groups passed; the second round tried “solder paste on all pads, adhesive only at the four corners,” with reflow 3/6 passes × hot-air gun 60/120 seconds and a 120 cm drop, 4 groups and 12 units in total, and all passed as well — so four-corner dispensing became an acceptable material-saving approach. This conclusion applies only to that case’s shield can size, adhesive, and conditions; a different product needs the test repeated.
5. Fix the Metal Lid Before Rework: Another Time to Use Adhesive
Q: When an entire module has to be removed for rework, what if the shield can (metal lid) comes off first?
A: One production-line practice is to dispense red glue around the metal lid before removal and heat it to cure, bonding the lid and the module into one piece, then heat from above and below to remove the module. One rework procedure we read had five steps: dispense red glue around the metal lid → heat with a hot-air gun to cure the red glue → fix the motherboard on the rework station and heat from above and below to remove the module → clean the pads → position the new module and heat again to reattach it. A separate failure analysis also cautions that the heating used to simulate removal can itself cause BGA ball shorts, so when a removed module is to be analyzed, first distinguish whether a defect was already there or was caused by the removal.
6. Records to Keep After Mass Production
A dispensing position map (how many dots, and where) and upper and lower limits for adhesive volume or dot diameter.
Adhesive lot number, opening and warm-up times, and curing conditions.
Validation report: combinations, conditions, sample size, and judgment.
Cleaning of the stencils or fixtures used for adhesive: stencils used to print red glue can be cleaned with a stencil cleaner, to keep adhesive from drying in the apertures.
Limitations of This Article
The experiment figures come from three versions of one production-line experiment record, with 3 units per group; this is a small-sample engineering judgment, not a reliability certification.
The full values for adhesive type and curing conditions were not fully read in the source document; the article only says “follow the maker’s data.”
The red-glue fixing process before wave soldering (dispensing red glue, curing, push-force testing) is outside the scope of this article.
Frequently Asked Questions
Q: Once adhesive is applied, can the shield can still be removed later?
A: It will be harder to remove, so first confirm whether components underneath will need rework later; if they will, consider a two-piece shield can — see “How to Choose a Shield Can“.
Q: Are 3 units per group enough?
A: Enough to screen out approaches that clearly do not work; before releasing to mass production, it is advisable to confirm once more with a larger sample and write the conditions into the validation report.