Water spots on a washed PCBA are dried dissolved solids, not leftover flux. Air knife drying decides where those solids land, and five variables set the outcome: rinse quality, drag-out, air geometry, board topography and dwell time.
The evidence is a document audit, run on 26 October 2026 for process engineers at EMS plants. Of the 22 product pages on chuangqitech.com, 21 carry a technical parameter table, and every field name in those tables was tallied. One publishes a drying temperature: the CQ-C743 offline PCBA cleaning machine, up to 99 °C. One names a method in words: the CQ-C758 circuit board cleaning machine, forced-air drying. Three publish a conveyor speed. Two publish a DI water resistivity figure, both above 15 MΩ. No sheet publishes a drying-air figure — no knife pressure, no air temperature tied to a belt speed, nothing a buyer could accept or reject a machine against.
Key takeaways
- A water spot is a dried solute deposit, not visible flux residue.
- Of 22 datasheets, one publishes a drying temperature; none publishes knife settings.
- Pair every drying figure with a conveyor speed, or the number cannot be tested.
- Board height and hole density move drying time more than a few degrees do.
- Not every spot is a defect: mask texture and metal finish read differently.
Five variables that decide whether a board exits wet or spotted
1. Solute load in the final rinse. What dries on the board is what was dissolved in the water that stayed behind. A rinse reading above 15 MΩ carries very little ionic material, which is why the same dryer can spot one shift and not the next — the water changed, not the machine. Resistivity is published for two of the water-wash models (the CQ-C9610 and the CQ-C743); check it at the point of use, not at the DI plant outlet.
2. Drag-out and rinse staging. Every board carries a film of detergent out of the wash zone. The CQ-C9610 quotes a detergent figure of 1.0 to 2.5 litres per hour, and that number is a proxy for how much chemistry is being introduced and therefore how hard the rinse stages have to work. A single rinse stage with heavy drag-out will pass water that reads clean and still deposit residue.
3. Drying-air geometry. Where the air hits, at what angle, across what gap, and how many knives act on the panel: these decide whether water is pushed off the edge or driven into a via and left there. No datasheet in the set publishes any of it. That is the single largest blind spot in the purchase documents we audited, and it is also the variable most often changed on site without anyone writing it down.
4. Board topography. A flat test coupon and a production panel do not dry at the same rate. The CQ-C9610 inline PCBA water cleaning machine takes workpieces up to 580 mm wide, 450 mm long and 80 mm tall; the height limit is where trouble starts: a tall connector or a shield can creates a shadow that air cannot reach, so the panel exits with a wet patch behind it. Heavy copper and dense through-hole fields hold heat and water differently from thin inner layers.
5. Time between the last rinse and the air. On CQ-C9610 the belt spans 0–1500 mm/min, while CQ-C758 and CQ-C760 move at 0.05–1.8 m/min. At the slow end, a board sits in the tunnel long enough for water to spread and pool into a larger droplet; at the fast end it reaches the air stage still shedding. The same machine produces different spot patterns at different belt speeds, which is why a drying temperature quoted on its own tells you almost nothing.
What 22 datasheets actually publish about drying
The table below compares the four water-wash families on the fields a buyer can actually put into an RFQ. Read the empty cells as the real finding: the documents describe throughput and water supply far more consistently than they describe the stage that produces the defect.
| Model | Drying row | Throughput row | Water row |
|---|---|---|---|
| CQ-C9610 inline | Not published | Conveyor 0–1500 mm/min | Inflow 1.0 m³/h; DI water above 15 MΩ |
| CQ-C743 offline | Up to 99 °C | Two-tier basket 810 × 560 × 100 mm | DI above 15 MΩ; 0.45 µm filtration |
| CQ-C758 board | Forced air, no figure | Conveyor 0.05–1.8 m/min | Not published; tank max 20 L |
| CQ-C760 brushing | Not published | Conveyor 0.05–1.8 m/min | Alcohol-based detergent |
Source: chuangqitech.com product pages, technical parameter tables read on 26 October 2026. The counts in this article come from the same 22 pages and can be re-run by reading the same field names.
How to write drying acceptance into the specification
A drying claim that cannot be tested is a marketing sentence. Five lines turn it into a specification item:
- State the criterion, not the temperature. “No visible deposit on masked and unmasked surfaces, inspected under the agreed light, at the agreed belt speed” is a criterion. “Dries at 99 °C” is a setting.
- Name the board. Attach the actual panel: layer count, copper weight, tallest component, via count. A dryer qualified on a coupon is not qualified on that panel.
- Pin the belt speed to the drying figure. Ask for the drying time at the speed you will run production at, and treat a number given without a speed as unverified.
- Define the inspection method. Which light, which angle, who signs, and what counts as a spot versus mask texture. Two people looking at the same panel under different light will disagree.
- Ask what the machine can hand over. A drying stage that logs nothing leaves you with a defect nobody can trace to a batch, a shift or a speed.
Where this framework stops working
Not suitable for solvent processes. If your residue is not water-soluble, the rinse-load logic above does not transfer: the carrier is a solvent, the drying mechanism is evaporation, and the variables reorder.
Not suitable for boards that cannot take the temperature. Where a component or a coating sets a ceiling below the dryer’s capability, you are buying time and airflow, not heat, and the checklist should be rewritten around dwell length instead.
Not a defect call on its own. A faint edge mark on solder mask is often cosmetic; a ring around a via that later blooms under coating is not. This article tells you where spots come from. Deciding which ones matter is a customer drawing or an internal workmanship standard, and we cannot write that for you.
Why do water spots appear only after the DI rinse?
Because pure water is a poor carrier but a good solvent: it lifts ionic material off the board and then leaves it behind when it evaporates. Spots that appear only after the rinse point to solute in the final stage, not to detergent carry-over.
Can I fix spotting by raising the drying temperature alone?
Rarely. Temperature accelerates evaporation, but a droplet in a shadow or on a cold heavy-copper field stays long enough to dry in place. Raising heat without changing airflow moves the spot rather than removing it.
Does conveyor speed really change drying?
Yes, and it is the variable most often left out of the quote. Published belt ranges on our water-wash models span 0 to 1500 mm/min and 0.05 to 1.8 m/min, so dwell time between rinse and air varies by an order of magnitude across a shift.
Are spots on solder mask a real defect?
Not automatically. Ask what the spot does downstream: if a coating must adhere over it, treat it as a defect; if the surface is cosmetic and uncoated, agree the limit with your customer before you chase it.
What should I send to get a drying recommendation?
Send the panel description, the current belt speed, the rinse water reading at the point of use, and one photo of the defect under your inspection light. A sample test is free of charge with no purchase obligation, and we reply within 24 hours with before-and-after video and a cleaning report.
Specification values come from chuangqitech.com product pages as published in October 2026 and describe the machines as built, not a process guarantee for your plant. The field counts are a document audit of those same pages.
Chasing spots on one board family? Send the panel and the belt speed through the cleaning application form and you get a quote within 24 hours. Start with the machine: inline PCBA water cleaning machine (CQ-C9610). For the drying stage on heavy assemblies, see rinse and dry parameters for heavy copper; for the residue that is not water, see white residue after soldering.