Spray nozzle clogging on a PCBA washer is a flow problem wearing a pressure problem’s clothes. The pump gauge sits where it always sat, the temperature reads correctly, and one corner of the board still comes out with a film on it. Most lines replace a pump that was never broken. The eight checks below close the question in order, and each one is read without opening the cabinet first.
All readings below are taken on the CHUANGQI CQ-C9610, an inline PCBA washer whose product page documents five figures the whole sheet depends on: 10–16 L/min drawn at the rinse, 1.0–2.5 L/h of detergent, 1.0 m³/h of DI feed at above 15 MΩ, and a belt that runs at 0–1500 mm/min. Treat them as what the machine is rated for, not as settings to impose.
Key takeaways
- A clogged nozzle shows up as a flow drop, not a pressure drop.
- Rinse draw on CQ-C9610 is rated at 10–16 L/min; read it live.
- Detergent demand of 1.0–2.5 L/h is the second meter to read.
- One dirty corner usually means one blocked fan, not a worn pump.
- Log the nozzle position, or the same one clogs again next month.
What a clogged nozzle looks like from outside the cabinet
Start with the method, because the instinct on a wash line is to open the door and look. Spray nozzles are hidden behind headers and inside a wet enclosure, so the visible evidence is on the board and on the instrumentation, not on the nozzle itself.
Three signals travel together when a fan starts to block. The residue is positional rather than random, repeating on the same board location run after run. The wet-flow readings drift down while the pump pressure holds. And the defect appears after the rinse more often than after the wash, because a starved rinse stage cannot carry away what the wash released.
Read them in that order. A positional defect with stable flow points somewhere else entirely, and the eight checks below will say so by check three.
The eight checks, in the order they close a question
Each check rules something out. Stop when a check rules in the fault, because the remaining checks measure the same cabinet from other angles and will not change the answer.
1. Is the defect positional or random?
Run the same board twice and mark the residue. A defect that stays in one place maps to the spray field above it; a defect that moves maps to the chemistry, the belt speed or the dry stage. This check costs one board and rules out the chemistry and dry-stage causes before you touch the cabinet. If the board rotates between passes, fix the orientation for the test, otherwise the mapping is lost.
2. Is the wet flow down at both ends?
The rinse circuit and the detergent circuit each carry a meter, and both are documented: a DI draw in the ten-to-sixteen litre range, and a detergent figure that runs up to several litres an hour. If both readings sit at the bottom of their bands on the same board mix that used to sit mid-band, the restriction is upstream of the nozzles and shared by both circuits. Filter loading is the usual cause and the pump is not.
3. Does pump pressure still hold at its commissioning figure?
Pressure and flow are not the same measurement, and this is the heart of the confusion. A pump pushing against a partly blocked header can hold pressure while delivering less volume, because the restriction has moved the operating point. Compare against your own commissioning record rather than against the specification band, since the band describes the machine and your record describes your line.
4. Which fan is starved?
If the defect is positional, it points at one fan of nozzles rather than the whole header. Note which side of the board reads clean. A single fan starving on a four-fan header is a nozzle or a local filter issue; two adjacent fans starving points at a manifold or a valve. This is the check that decides whether the fix is a cleaning job or a maintenance intervention.
5. Is the restriction debris or scale?
Open the header only after the four checks above, and look at what came out. Soft debris that smears is flux solids and filter media that bypassed the element. Hard, crusted deposits are scale from water hardness or from detergent left to dry in the line. The two have different fixes, and a scale problem will return within weeks if the only action taken is to rod out the nozzle.
6. Is the rinse water pushing the problem back in?
Resistivity above 15 MΩ describes the feed, not the tank. Read at the rinse outlet while the line runs: a rinse that stops holding its resistivity is loading dissolved solids onto the board, and those solids dry into exactly the kind of film that looks like a blocked spray. This is the check that most often explains a “clog” on a machine whose nozzles turn out to be clean.
7. Is the dry stage moving the residue instead of removing it?
Air knives do not clean, but they do decide where the last film lands. On a board with a tall component, the shadow downwind of that component dries last and shows residue first, and the position is as repeatable as a blocked nozzle would produce. Confirm the dry-zone behaviour before ordering parts: a belt speed change or a knife adjustment is free and a nozzle set is not.
8. Is the same nozzle blocking repeatedly?
Repeat clogging in one position is a system problem, not a cleaning problem. It points at a filter that is a stage too coarse for the soil on your boards, a manifold with a dead leg where solids settle, or a line that sits wet between shifts. Fix the cause; otherwise the next cleaning cycle restarts the same clock.
Where the eight checks sit, and what each one rules out
The table pairs every check with the reading that decides it and the fault it eliminates. The third column is the one worth keeping: a maintenance log that records what was ruled out stops the same investigation being repeated by the next shift.
| Check | What you read | What it rules out |
|---|---|---|
| 1 Position of defect | Repeatability on the same board location | Chemistry and dry-stage causes |
| 2 Wet flow, both meters | DI draw 10–16 L/min; detergent 1.0–2.5 L/h | Pump wear as the primary cause |
| 3 Pump pressure | Gauge against your commissioning baseline | A dead pump |
| 4 Which fan | Clean side versus dirty side of the panel | A whole-circuit fault |
| 5 Debris or scale | What comes out of the header | A cleaning-only fix |
| 6 Rinse outlet | Resistivity at the tank outlet, line running | Nozzles as the source of the film |
| 7 Dry stage | Residue position relative to a tall part | Spray starvation |
| 8 Recurrence | Same position over weeks | A one-off event |
Data source: CHUANGQI product documentation for the flow and resistivity figures, read in October 2026. The check sequence is the author’s diagnostic method, not a machine procedure.
Two of those readings are worth holding in mind as ranges rather than numbers. Board width, how many rinses the machine carries, and the detergent strength in the tank all move the 10–16 L/min figure, so a low reading only means something when the board mix and the setting are unchanged. The detergent figure swings the same way.
What to log so the same nozzle does not clog twice
A nozzle cleaning that ends with a clean board and an empty memory guarantees a repeat. Write four things into the maintenance record while the header is still open.
- The position. Fan number and side, not “the nozzles were cleaned”.
- The reading before and after. The flow figure that recovered is the evidence the fix worked.
- What came out. Debris or scale, and whether the filter element was changed at the same time.
- The trigger for the next inspection. Event-based, tied to the filter interval you actually run.
For the broader sequence this sheet sits inside, the PCBA washer troubleshooting tree covers the faults that are not spray-related, and the maintenance interval plan sets the cadence for the consumable side.
Where this diagnostic sheet stops working
The eight checks assume a conveyor-fed spray washer with a published flow figure for the rinse and the detergent circuit. They are not suitable for an ultrasonic batch tank, where the comparable reading is cavitation activity rather than nozzle flow, and they do not carry over to dry-ice blasting, where there is no liquid circuit to read at all.
They also stop short of a root-cause analysis on the chemistry. If the checks point at residue that the rinse cannot carry away, the answer sits in detergent concentration, bath life and drag-out rather than in the spray field, and a different investigation applies. And where a machine has no flow instrumentation at all, checks two and three cannot be run as written: the practical substitute is to time a measured draw at the rinse outlet, which is cruder but still positional.
Send the spray header layout plus the position of the dirty area on the board through the cleaning application form. You will get back the two checks most likely to close it on your machine, plus a quote within 24 hours. Every figure quoted here sits on the CQ-C9610 Inline PCBA Water Cleaning Machine specification sheet, which is what to check before ordering a nozzle set.
Frequently asked questions
Why does one corner of the board stay dirty when pressure looks normal?
Because pressure and flow are separate measurements, and a partial restriction downstream of the gauge moves the operating point without dropping the reading. One corner staying dirty points at a single fan of nozzles or at the dry stage downwind of a tall component. Check the wet-flow meters first, then map the defect position before opening the cabinet.
How often should spray nozzles be cleaned on a conveyor washer?
Set the interval from your filter changes rather than from the calendar. Every time a filter element loads faster than usual, the material that bypassed it has to go somewhere, and the nozzles are the first place it stops. On most lines that means inspection tied to the filtration interval, plus an immediate check whenever a positional defect appears.
Can a clogged nozzle damage the boards?
It does not damage them directly, but it leaves residue in a fixed place, and residue that survives the rinse is what later fails an ionic check or a coating adhesion test. The risk is a systematic escape rather than a random one, which is why a positional defect should be treated as a release question and not only as a maintenance task.
Do I need spare nozzles on the shelf?
Yes, and by position rather than by quantity. One fan that clogs repeatedly will keep clogging until the upstream cause is fixed, and having that specific nozzle type on the shelf turns a stopped line into a short job. Stocking an even number of every nozzle on the header is the more expensive way to reach the same certainty.
Every figure here is a published CQ-C9610 specification value. It describes what the machine is rated to do; it is not a setting to copy onto your line.