Flux by definition is an acidic material.

Any amount left behind is too much as the residual salts are always susceptible to the absorption of moisture, thereby corrosion. Chemically it is corrosive, electronically with the appropriate amount of moisture dendritic growth can and will occur and this will impact electrical functionality.

The only flux which is advertise to be left on the surface of the board is the low solid no clean flux. This flux must also have reached soldering temperatures to drive off the activators and remain benign and if it has not been heated to soldering temperature, they are still corrosive and those residues have to be addressed and removed.

The limits of excess flux is typically identified with a solvent extract test of which defines the amount of salt residues left behind per given size of the board. Personally this test is only good for rosin base fluxes as the other fluxes may not be soluble in this solution and false readings will occur. The best test is the Surface Insulation Resistance Test as can be found in the IPC test methods. This test will determine whether or not the residues left behind will be detrimental to the product or not if left behind.

This is an involved topic and it is difficult to provide sufficient information in such a short forum, but if more question exist please feel free to contact me at any time.

The primary function of flux is to remove oxidation from the surfaces to be soldered therefore improving wetting of the liquid solder.

Salts are produced upon reaction with the oxide and are ionic in nature. They have been related to dentritic growth. No clean fluxes have been designed to leave non-corrosive residues and therefore anything left behind should not be detrimental to the PCB. I am not an expert on fluxes, however.

From a protection point of view, excess residues may affect further processing. If a conformal coating or encapsulation resin is to be applied then excess residues can cause issues with adhesion, especially in the case of conformal coatings.

Although no-clean fluxes do not necessarily require cleaning, if you were to try and remove any excess then it may be rather more difficult if there are significant amounts of residues on the board. I am not aware of any particular limits of residue that are allowed however if you keep the amount used as close to the minimum required as possible, then it reduces the possibility of any issues.

In general, it is preferred that technicians should always clean after hand-soldering, and there should be no visible residues around the repaired solder joint. This is particularly important, oddly, when using no-clean fluxes.

There are many ways to measure cleanliness, but the three most common are:

1. Visual inspection of the solder joint for solder balls, flux residues or white residues (few or no residues is the preferred result);
2. Ionic contamination measurements using super-pure deionized water and measuring the amount of salts lifted off the board by the deionized water (lower is better); or
3. Surface Insulation Resistance measurements, usually over a 28-day period, where contamination on the board will cause a slight but measurable current flow between leads in the PCB (higher resistance readings are better).

Visual inspection is erratic and subject to operator fatigue, so it is only acceptable on low-volume, low-density designs.

Ionograph testing (using machines such as “Omegameter” brand cleanliness testers) can be very rigorous, reasonably fast and very inexpensive, but there is a gap between knowing how clean the board is and how clean it needs to be to have a long and happy life. This gap can be eliminated if repeated testing shows a correlation between a certain degree of contamination and predictable board failures, so that’s a good design process if you’re making long runs of one certain board.

SIR testing is the most precise, and defines a precise link between a cleaning failure and noise on a circuit or the failure of a board. But it takes a long time and a lot of money to perform.

So the short answer is to define a rigorous cleaning process and ensure that your people stick to it.

But why do you even need to clean no-clean fluxes?

The answer is found in the chemistry of the no-clean flux process.

Most no clean fluxes become “no-clean” through two sequential steps. First, the flux manufacturer reduces the quantity of corrosive activators in the flux. These activators are typically salts or halites, and we all know that salt will corrode metals. By reducing the quantity of activators in the flux they reduce the product’s ability to corrode a PCB, thereby giving it a longer life without cleaning.

The problem is that low-halite fluxes are very hard to solder – without activators, you have a relatively weak flux, which is why no-clean fluxes are so often hard to solder well.

Next, the flux manufacturers add materials which encapsulate the activators in a protective bubble of inert material, so if the salts happen to bump up against a copper substrate they will be rendered harmless and unable to corrode the material. This is an excellent and reliable process that works great in automated soldering systems where temperatures can be precisely controlled.

But in hand soldering, temperatures vary all over the place, from right next to the soldering tip to cooler spots centimeters away. So in a manual soldering environment not all the encapsulants are heated sufficiently, which leaves unencapsulated activators free on the circuit board, ready to do their damage.