How to Decide Which Type of Flux to Use and How to Use Flux! - Workbench Wednesdays 59

Thank you for all the suggestions. Since I have more damaged sensors, I will try both approaches:

1) trying baking with lower temperature

2) desoldering the old sensor, cleaning, soldering with getting some paste

As shabaz pointed out, you can probably just rely on the flux in the paste and not worry about cleaning.

If you have a concern, then I would see if the sensor supports a water-based washing.

If so, use a solder paste with a water soluble flux. (Which I would imagine any of the SAC-based pastes is already using anyway.)

I would add to do a good cleaning with IPA before applying the paste. That way no old residue is left over. And in this case, the flux in the paste should be enough to provide a good solder joint.

Hi,

Personally I would just use paste, i.e. no additional flux beyond what is already in the paste, and not wash it off. Ideally with a stencil or just a tiny dab on each pad if the stencil is not practical or possible. In other words, for rework, still just do what would happen to the board if it was being produced rather than being reworked.

With connectors, switches etc., with rework, I wipe off the flux with IPA, but avoid the connector/switch area (this is easy to do with wipes). This is for flux types that do not need to be removed ("no clean"), but if I want it removed for cosmetic reasons.

Pure isopropyl alcohol does not leave a residue. If something has been dissolved in it or it has picked up contaminants, those will remain after the alcohol evaporates. Alcohol could presumably get trapped in the sensor so the bake period is to get it to evaporate even though there is not much air in the sensor. Any elevated temperature would speed evaporation - it probably doesn't need to be 100 degrees. 100 degrees sounds like they are trying to get water to turn to steam. Some alcohol is diluted with water, but even water will evaporate well below 100 degrees.

Try warming it up to about 50 degrees C (which is pretty warm but not too hot to touch) for an hour or two - to see if your readings improve.

If you do replace it, find out how the manufacturer recommends protecting it from cleaning fluids (which are used in almost every assembly operation).

I've seen adhesive labels and tape used to cover ports during cleaning.

The sensor gets a non-linear offset when exposed to alcohol. I exposed it to too much alcohol apparently and have two options. Bake the sensor for 10 hours at 100-105 Centigrades (and damage the rest of the PCB) or replace it. If I replace it, it is a good idea to use flux, so it needs to be cleaned after soldering, which gets me to my starting point - there is a new sensor damaged by too much alcohol.

I should probably try to remove it without flux and place it back also without flux, or using some solder paste. (I haven't done this before, I always use flux.)

I'm not sure I understand. The performance of many types of sensor is adversely affected by ingress of many liquids and particulates. Such sensors are generally masked (with a sealed cover) if the circuit board is to be exposed to such liquids during the manufacturing and assembly processes. There should be no flux in the sensor, so flux cleaner doesn't need to go in there either. 

It is a general, but know very well known effect harmful to all capacitive sensors. I haven't heard about it until last year, when I bought a new humidity sensor (manufactured and calibrated in the same facility as another one) and there was a perfect match of temperature measurements, but difference in humidity measurement). The difference was higher at low relative humidity levels and lower at high relative humidity levels. As an example, have a look at this document, section Exposure to chemicals: https://sensirion.com/media/documents/6D95AA80/616446E3/Sensirion_AppNotes_Humidity_Sensors_Handling_Instructions.pdf

The only way to get out of it is to bake the sensor at 100-105 Centrigrades for 10 hours - which might damage connectors / displays and other components on the PCB.

Here is an example of my measurements, the azure color belongs to a new sensor, purple to the contaminated one.

I am curious to know how alcohol caused a failure.

Manufacturer's often use an aqueous wash stage after soldering, It needs a proper drying stage as well.

There are other chemicals that can clean flux, but I'm not sure they are more benign than alcohol.

I would appreciate an advice on choosing flux / cleaning after soldering. I have a sensor board with a capacitive sensor. This sensor got damaged by an exposure to alcohol. I need to replace the sensor, getting if off the board will be simple, but how to solder a new one? I should use flux, but I can't clean it with alcohol after that. Or is the only option using a precisely placed solder paste? The sensor package is DFN6 / DFN8.