
Component masking (also written masking before cleaning) is the process step that isolates components which cannot tolerate the cleaning fluid, so the rest of the assembly can still be washed. It is a process decision with decisions of its own, and it is not a consumable purchase. When a customer specification reads clean the whole assembly and the bill of materials holds a part the fluid would ruin, masking is the route that keeps both statements true.
The eight decisions below are drawn from how a masked cycle behaves on the CHUANGQI CQ-C9610 Inline PCBA Water Cleaning Machine, whose published sheet caps the workpiece envelope at 580 × 450 × 80 mm, lets the conveyor run anywhere between 0 and 1500 mm/min, and specifies DI rinse water above 15 MΩ. Those figures matter because a mask stack eats height and dwell time. The machine facts are published; the decision sequence is CHUANGQI’s own process guidance, and our application engineers recommend requesting a written washability statement from the component supplier for every masked part — nothing in this sheet substitutes for it.
Key takeaways
- Masking is a process decision with eight gates, not a consumable.
- The 80 mm workpiece limit on CQ-C9610 bounds every mask stack.
- Ask the component supplier for a washability statement, in writing.
- Removing the mask damages more boards than the wash does.
- Log the mask stack in the first-article report, not in email.
Eight decisions that have to close before the first board enters the washer
The list is ordered by what blocks the next step, not by importance. Decision one gates everything else, because a part with no washability statement cannot be planned around. The sequence applies to any conveyor-fed spray washer; the numeric bounds quoted here come from the CQ-C9610 sheet and will shift on a batch machine with a fixed basket height.
1. Which components are actually non-washable
Start from the bill of materials and mark every part whose supplier documentation limits contact with water, detergent or solvent. Relays with open contacts, electret microphones, unsealed pressure sensors and some buzzers appear on such lists often, but the list that counts is the one your suppliers publish for the parts you buy. Where no statement exists, the part is unknown, not permitted.
2. Who owns the deviation
Masking asserts that part of the assembly will receive less cleaning than the specification asks for. That assertion needs an owner: usually the customer, in writing, or your own quality function against the customer’s intent. Decide this before choosing a masking method, because a refusal here changes the whole project into a selective-cleaning job.
3. What the mask has to withstand
A mask sees detergent, mechanical spray, DI rinse and hot-air dry in one pass. The four conditions act together, so a material approved for a splash is not thereby approved for a spray cycle.
4. Which masking method fits the geometry
Peelable liquid mask, die-cut tape, silicone caps, machined covers and dedicated carrier pockets all exist. Each occupies space, and each removal method carries its own risk to neighbouring parts.
5. How the stack fits the envelope
This is where the machine specification bites. The CQ-C9610 sheet lists a maximum workpiece of 580 × 450 × 80 mm, and that 80 mm is the sum of board thickness, component height, mask material and anything the carrier adds. A tall relay with a cap over it can consume most of the allowance on its own, and the limit depends on the tallest point anywhere on the panel rather than on the average.
6. Where the board sits in the spray field
The conveyor runs from 0 to 1500 mm/min on this model, and that range sets how long any surface stays under the headers. A masked area that shadows a neighbouring part also changes what that part receives, so the spray pattern around the mask belongs in the process plan. What the part actually receives depends on nozzle arrangement, pump pressure, board width and the mask’s own profile, so treat the speed range as an envelope rather than as a setting.
7. How removal is validated
Removal is the step that fails quietly. Adhesive left behind on a connector, a peeled corner lifting the solder mask, a cap dragging a wire: three defects the wash never caused. Prove removal on scrap assemblies first, using the lighting and magnification your line already inspects under.
8. What evidence ships with the lot
The masked lot needs proof that the parts you protected were protected and the rest was cleaned. That usually means a first-article record showing the mask stack, plus whatever cleanliness evidence the customer specification names. Under IPC J-STD-001 Revision H (2026 Q1), which added its own chapter devoted to cleaning and residue requirements, acceptance is built per product rather than taken from one universal number; the standard also dropped its single ionic threshold in 2018 (Rev G, Amendment 1). Quote the customer’s own acceptance route rather than a remembered figure (per IPC, ipc.org).
What goes wrong when masking is improvised
Improvised masking fails in four recognisable ways, and only the first one is about cleaning.
| Failure | How it appears | What prevents it |
|---|---|---|
| Wrong part masked | Residue remains under a shielded component | Mask map signed against the BOM |
| Mask lifts mid-cycle | Fluid reaches a sealed part | Material approved for spray plus dry |
| Height overrun | Board jams or scrapes the header | Stack measured against 80 mm |
| Removal damage | Adhesive or lifted mask on a joint | Removal trial on scrap boards |
Data source: the CQ-C9610 product specification sheet on chuangqitech.com, October 2026, for the height limit; the failure modes are process observation rather than measured data.
The height row is the one that surprises people, because it is arithmetic rather than chemistry. A 3 mm board, a 25 mm component, a 1.5 mm cap and a carrier rail can reach the ceiling of the envelope without anyone ordering anything unusual.
How masking changes what the washer has to prove
A masked assembly shifts the burden of proof from the wash to the record. Three items move.
- The cleanliness claim narrows. Your evidence covers the unmasked area, and the masked area carries its own justification.
- The dry stage matters more. Fluid trapped under a cap has no route out except the air knife, so dry-zone behaviour has to be confirmed on the masked configuration specifically.
- The ionic check needs a defined board. Every CHUANGQI unit leaves the factory only after passing an ionic cleanliness check on a test board, and that board is not your masked assembly. Treat the factory result as a machine acceptance, then run your own evidence on product.
Where a customer is moving to selective coating, the masking discussion tends to travel with the deposition strategy; that side is covered in the notes on coating masking before the coat. Mixed-technology boards add a further layer of constraint, described in the mixed-technology cleaning strategy.
Not every assembly should be masked
This sheet is not suitable for an assembly where several tall parts across most of the panel need protection — at that point the mask stack consumes the working envelope and the honest options are a selective cleaning head, a different flux chemistry, or a design change that moves the incompatible part to a second operation. It is also the wrong route when the specification forbids any deviation from full-board cleaning, and there the decision belongs to design review rather than to the process engineer.
One limit of the method itself: the eight decision gates do not measure anything. They force a record. If you need a number for the masked configuration — residue under a cap, drying completeness at a corner — that number comes from your own first-article test and not from this checklist.
Send the board drawing, the component list with washability statements and the customer specification, and the engineering team will say which of the eight gates your assembly will struggle with, with a quote within 24 hours. The CQ-C9610 Inline PCBA Water Cleaning Machine page states the envelope the stack has to stay within.
Frequently asked questions
Can a board with a non-washable component still be cleaned?
Yes, and it is routine. The washer cleans the exposed area, the mask keeps fluid off the part that cannot take it, and the record explains which area was covered by which claim. What changes is the wording of the cleanliness evidence: it now covers the unmasked surface, while the masked area is justified by the supplier statement and the customer deviation instead of by a cleaning result.
Is peelable mask safe for an 80 mm tall assembly?
The 80 mm figure is a ceiling on the whole stack, not a clearance above it. Board thickness, the tallest component, the mask layer and any carrier rail add up, so a tall part plus a cap plus a rail can exceed the limit while every individual item looks small. Measure the finished stack at its highest point on the panel, and if the number lands close to the ceiling, drop the carrier rail before you reconsider the mask.
Does the customer specification allow partial masking?
Usually it does not say, which is the problem. A specification written as clean the whole assembly has no clause for areas that receive less cleaning, so the deviation has to be raised explicitly and answered in writing. Ask before the process is fixed, because a refusal turns the job into selective cleaning or a design change, and both are cheaper to plan than to retrofit after the first article.
What proof do I send with the first masked lot?
Send the mask map against the bill of materials, the stack height measured at the tallest point, the removal trial results from scrap boards, and the cleanliness evidence your customer specification names for the unmasked area. Add the supplier washability statement for each masked part. Four documents answer every question an auditor asks about a masked lot, and none of them is an email thread.