Reflow Zone Temperature Verification: Beyond the Profile
A profile is a measurement of one assembly on one day. The zone controllers report their own temperature, the board passes through, and the thermocouples produce a thermal profile that is filed with the product record. What that curve cannot show is whether the oven itself is still doing what the controller believes, and two ovens running the same recipe can produce the same curve while their zone air temperatures differ by ten degrees.
Zone verification closes that gap. It measures the oven rather than the product: the actual air temperature in each zone, the time the board spends in each zone, and the way the zones interact with each other. The result is a machine that can be compared over time and compared with a sister line, which is what makes a transfer of a product between lines a controlled event rather than an experiment.
What a Setpoint Does Not Tell You
A setpoint is a target for a controller that measures the air temperature at one point, usually at the side of the zone or at the heater outlet. If the sensor is offset by five degrees, the zone runs five degrees hotter or cooler than the number on the display, and every recipe on that machine is shifted. The controller does its job perfectly and the process is wrong.
Two machines can also differ in the way the air is distributed. One may deliver most of its heat through impingement from the top and the other from the bottom, and the same setpoint then produces a different board temperature. The differences are usually small enough to be invisible in a profile taken on a heavy test board and large enough to matter on a thin one.
Zone Air versus Board Temperature
Zone air temperature responds to the controller, and board temperature responds to the air plus the thermal mass of the assembly. The gap between them is the driving force for heat transfer, and it narrows as the board approaches the zone temperature. A board with a large ground plane and a heavy connector needs a larger gap to reach the same peak, which is why the recipe for a dense assembly cannot be transferred to a sparse one.
The measurement plan should therefore include both. The reflow zone temperature is characterised with a fixed probe inside each zone, and the board is characterised separately with the thermocouples described in the notes on <a href="https://www.gopcba.com/reflow-profile-thermocouple-placement/” title=”thermocouple placement”>thermocouple placement. The two records together describe the machine and the product, and either one alone leaves the other unverified.
Thermocouple Attachment and Error
An attached thermocouple measures the temperature of the joint between the wire and the surface, and it does not measure it accurately unless the attachment is right. A bead held on the surface by a drop of adhesive gives a reading that lags the board, and a bead buried in a deep solder fillet reads the fillet rather than the pad. High temperature tape over the bead, with the wire routed flat against the board, is the usual compromise.
Wire gauge matters for the same reason. A heavy wire conducts heat along itself and away from the bead, and a long run of wire under a component changes the local heat balance. The error from a poor attachment is typically several degrees and it can be more than twenty, which is larger than the tolerance the profile is trying to hold. A profile that is not repeatable between two runs is often an attachment problem rather than an oven problem.
Conveyor Speed and Dwell
The conveyor speed sets the dwell time in each zone, and the dwell time multiplied by the driving force sets the heat transferred. The speed is set by the recipe and is then assumed to be constant, but a chain that stretches, a drive that slips under load or a rail that runs at a different speed from its pair changes the dwell. Measuring the transit time of a marked board through the tunnel is the direct check.
The speed has to be verified with the production load on the belt, because a loaded conveyor behaves differently from an empty one. Where the oven has several boards in the tunnel at once, the loading pattern matters as well, since a dense train of boards shields the zones from each other and changes the air temperature profile that the last board sees. The warpage that results from unequal heat is covered in the notes on warpage control.
Zone-to-Zone Interaction
Zones are not isolated. Heat moves from a hot zone into its neighbours by conduction along the tunnel walls and by air exchange at the openings, and the controller of a cooler zone compensates by heating more. That is normal, and it becomes a problem when the compensation reaches the limit of the heater or when the interaction changes, for example after a baffle is disturbed during maintenance.
The signature of a changed interaction is a profile in which the ramp between two zones has changed shape while the zone displays are unchanged. A step in the curve where there was a smooth ramp means the transition between zones has changed, and the cause is mechanical: a damaged baffle, a fan running at the wrong speed, or a tunnel that is no longer sealed at the entrance.
Closed-Loop and Open-Loop Control
Most ovens run closed loop on air temperature and open loop on the product. The controller holds the zones, and the product is expected to follow. Where a product is critical, a board mounted thermocouple can be used to close the loop on the board, which is more accurate for that product and less general for the machine.
The practical arrangement is to verify the machine periodically and to control the product by recipe. The verification sets the machine baseline, and any change to the baseline is investigated before a recipe is adjusted. Without that discipline, a drifting sensor produces a series of recipe changes that compensate for the drift and leave the oven in a state that no longer matches any documented setting.
Nitrogen and Airflow Effects
An inerted oven changes the verification in two ways. Nitrogen has a lower heat capacity than air, so the same impingement velocity transfers less heat and the zones need a higher setpoint for the same board temperature. The flow pattern also changes, because the gas is introduced and exhausted at defined points and the balance between them affects the temperature uniformity across the tunnel width.
The oxygen concentration should be recorded with the verification, since a change in the nitrogen supply or in the balance of the machine changes both the oxygen level and the airflow. A joint appearance that changes while the profile stays constant is often an atmosphere change rather than a thermal one, and the two records together make the cause obvious.
Verification Frequency and Records
A reasonable schedule verifies the zone air temperatures monthly with a calibrated probe, the conveyor speed weekly, and the board profile whenever the product or the recipe changes. Any maintenance that touches a heater, a fan, a baffle or the conveyor triggers a verification before production resumes, and the verification is repeated after the first production shift to confirm stability.
The records should be a table of zone temperatures, the transit time and the profile, each with the date, the machine, the recipe revision and the instrument used. Trending the zone temperatures against the calibration history shows a sensor that is drifting long before it fails, and that trend is the cheapest predictive maintenance available on a reflow oven.
Responding to a Drift
When a zone is found to be outside its band, the first question is whether the sensor or the heater has changed. A cross check with a second probe at the same point settles it, and the answer decides whether the fix is a calibration offset or a repair. Adjusting a recipe to compensate for a bad sensor hides the fault and moves it into every product on the machine.
Where the drift is small and within the band, it is still worth recording, because the useful signal is the direction and the rate. A zone that has moved two degrees in three months is telling the maintenance planner something, and an oven that is verified only when a defect appears has no such early warning. The soak zone behaviour behind these measurements is discussed further in the notes on soak and spike profiles.
FAQ
Can a profiling run replace zone verification? No. The profile shows what the board experienced, which is the right control for the product, and it cannot show whether a zone sensor is offset.
How accurate should a zone measurement be? The probe and its meter should be calibrated to better than one degree, and the attachment should be stable enough that two runs agree within two degrees.
Should the recipe be adjusted after a zone is recalibrated? Yes, and the adjustment should be reviewed against the last profile for every product on that machine, because a sensor offset affects all of them at once.
