What a PCBA cleaning station has to control is not the single result of “clean or not,” but five conditions that make the result drift: how long boards wait after reflow before cleaning, whether the DI water is pure enough, where rinsing stops, how long the chemistry has been in use, and whether ionic residue is creeping up. Write these five into the check sheet and cleaning quality becomes stable; once residue is left behind, conformal coating adhesion, test leakage and long-term corrosion problems follow.
1. From reflow to cleaning: set a maximum dwell time
The longer flux residue sits, the harder it is to remove; after cooling and absorbing moisture, it sticks more firmly under components. Common practice on the line:
- Set a maximum dwell time from reflow to cleaning (for example, 8 hours); boards past it are handled separately.
- After a product change, shift change or machine repair, clean one panel and verify it before starting.
- Record cleaning parameters (temperature, time, DI water) on the check sheet, and inspect at 10–20× after cleaning for no residue.
- Schedule cleaning after AOI and visual inspection, so boards are not touched by hand after cleaning.
Q: How soon after reflow should a PCBA be cleaned? A: As soon as possible; lines commonly set the limit within 8 hours and write it on the route card. Boards past the limit should not go straight into the normal program; first evaluate whether to lengthen cleaning or change parameters.
2. DI water: manage it with an “alert value plus stop value”
When the resistivity of the DI water used for cleaning and rinsing drops, there are more ions in the water, and boards come out carrying ions. Common practice:
- Two limits: at the alert value, act first (change resin, check piping); at the stop value, stop the line.
- Check resistivity every shift, and sample other items (particles, bacteria, total dissolved solids) weekly or monthly; place the measurement point after the final filter and before the point of use.
- A common cleaning machine specification: DI water resistivity at the machine inlet ≥ 10 MΩ·cm, tank temperature 45–65 °C; the machine does not accept boards when water quality is out of limit. (The DI water system’s own alert and stop values are a separate set of figures; see “Proving a board is clean.”)
- Sterilize or clean sampling bottles, or you will be measuring the bottle’s contamination.
Q: What DI water resistivity is needed for cleaning? A: Set it by the cleaning process and product sensitivity; a common lower limit on the line is 10 MΩ·cm, with alert and stop values on top. The closer to sensitive steps such as wire bonding or encapsulation, the stricter the limits.
3. Rinse endpoint: go by conductivity, not just time
A fixed 5-minute rinse under-rinses boards with heavy residue and wastes water on lightly soiled ones. Batch (offline) aqueous cleaners commonly run “clean → multiple DI water rinses → dry” in one chamber, ending the rinse when conductivity reaches the set value. In-line aqueous cleaners:
- Fresh DI water enters the final section alone, then cascades counterflow to earlier sections, so the last section is cleanest and water is saved.
- Fit resistivity monitoring at the rinse inlet; air knives remove water before hot-air drying.
- When cleaning no-clean or rosin residue, add a “chemical isolation” section, using air knives to blow cleaning solution back to the wash tank and reduce carry-over into the rinse.
- Boards with many small parts are sensitive to high-pressure sprays, so adjust spray pressure by section; the conveyor type (mesh belt, rollers) also affects pad damage, so evaluate it when choosing a machine.
Q: How long should rinsing take? A: Judging the endpoint by conductivity or resistivity is more reliable than a fixed time: set the cleanliness value to reach and stop when it is reached; also record the actual rinse time, since a sudden increase signals a problem in the wash section or chemistry.
4. Chemistry life: manage by time, with automatic lot lock at expiry
Chemistry concentration falls with use and time. Lines commonly write chemistry life as time (for example, two levels of 4 and 48 hours), with the system automatically locking the lot number at expiry rather than relying on memory; always confirm concentration after a change, and some machines replenish or monitor concentration automatically. When residue under parts won’t come off, check in order: is agitation enough (spray pressure, ultrasonics) → does the standoff gap under the part let liquid in → chemistry concentration, time and temperature.
Q: How often should cleaning chemistry be changed? A: Set it from the chemistry supplier’s recommendation and actual production volume; lines commonly manage it by a time limit with concentration checks; change it on time rather than waiting until boards stop coming out clean.
5. Ionic contamination testing (ROSE): manage it as a trend, not the sole pass line
ROSE (resistivity of solvent extract) immerses the board in solvent and measures how much the solution’s resistivity drops, converted to NaCl equivalent. IPC-TM-650 2.3.25 is the test procedure for this method (IPC-TM-650 test methods list).
But a ROSE value cannot be taken directly as proof of “safe.” An IPC expert panel summary notes that ROSE was originally a process control tool, not a general product acceptance criterion; the 2018 revision of J-STD-001 (Revision G, Amendment 1) removed the long-used 1.56 µg/cm² NaCl equivalent figure, basing acceptance instead on SIR testing, historical evidence of no residue-related failures, or temperature-humidity-bias (THB) testing (IPC: How Clean Is Clean Enough?). The SIR test method is IPC-TM-650 2.6.3.7.
Common practice on the line follows this direction:
- Sample ionic contamination daily by process step, with an in-house reject value written as NaCl equivalent per area (for example, 2.2 µg NaCl/in²), and handle exceedances under the quality procedure.
- No-clean processes can be measured too, but as a trend: use the same product’s history as the baseline, and if values creep up, check the process; parts, PCBs, handling, machines and carriers can all be ion sources, and SEM, FTIR or ion chromatography can trace the cause if needed.
- Chemical reliability (corrosion, electromigration) is confirmed separately by dedicated tests such as SIR.
Q: What ionic contamination result counts as a pass? A: The current J-STD-001 no longer uses a single 1.56 µg/cm² as the pass line; acceptance must be shown by SIR, failure history or THB; ROSE stays on the line as daily trend control, with an in-house reject value and attention to changes.
Q: Do no-clean solder pastes still need ionic testing? A: You can test, but the purpose is to see whether the process has changed, not to judge whether residue is safe. A sudden rise means a source (material, board, hands, fixtures) has changed.
Cleaning station check sheet (adapt it to your own version)
| Item | Control | Common value / frequency |
|---|---|---|
| Dwell time | Reflow exit to cleaning | Maximum 8 hours, written on the route card |
| First article | After product change, shift change, machine repair | Clean one panel first, inspect at 10–20× |
| DI water | Two resistivity limits | Every shift; commonly ≥ 10 MΩ·cm |
| Rinse | Conductivity / resistivity endpoint | Record actual time for every lot |
| Chemistry | Time limit + concentration | For example 4 / 48 hours, system lot lock |
| Tank temperature | Cleaning temperature | Record every shift; commonly 45–65 °C |
| Ions | ROSE sampling for trends | Daily; in-house reject value + historical baseline |
How JIAHON can help
You can discuss with JIAHON online how to write dwell time, DI water limits and chemistry life into your check sheet. Cleaning station equipment and consumables:
- In-line PCBA cleaning machines
- Ultrapure water systems (DI water systems)
- Flux cleaners
Have a similar problem? Write to JIAHON to see whether there is a matching solution.
Related reading
- “Proving a board is clean: four documents that pass a customer audit”
- “How to measure and read SIR (surface insulation resistance)”