1. SIR test principle

 

 

 

 

 

     The principle of SIR testing is very simple. It is to test two crossed cells placed in a set temperature and humidity aging environment.

 

resistance between the lower comb circuits (Figure 1). Residues may cause the surface insulation resistance value of the test sample to be low. If residues are left on the circuit board, it will affect the reliability of the product in the application environment.

 

 

 

 

 

2. Factors affecting SIR testing

 

 

 

 

 

1. The influence of surface coating

 

 

 

The test method is usually to certify the performance of the flux on the bare copper board, but the residues of the NiAu and HASL processes have a certain impact on the leakage on the PCB surface. The ethylene glycol in the flux used in the HASL process will be absorbed by the epoxy substrate woven fabric, enhancing the water absorption of the substrate. There have been some cases showing that the metal layer of NiAu will enhance the tendency of electrochemical migration. The impact of surface coating depends on the flux and chemicals used in the surface coating. Using RMA flux, the impact on the surface coating is minimal (Figure 2), but if no-clean flux is used, the residue will enhance the melting of Ni, resulting in a much lower SIR value than other samples.

 

 

 

2. Influence of test voltage

 

 

 

     The voltage applied to the comb circuit is a key influencing factor in SIR testing. Currently, high voltages around 50V are commonly used in test standards (the voltage on most samples is higher than 100V/mm). In the past, using such high voltages allowed the use of less sensitive resistance testers. It is no longer necessary to use such high voltages, and newer test circuits can test at 5V or lower.

Using a higher test voltage will not make the test difficult, but using a high voltage will destroy the dendrites before they grow enough to pass the test. Using no-clean flux and 5V, a significant decrease in SIR values due to dendrite growth can be seen (Figure 4).

The same flux was tested at 50V, showing the growth trend of dendrites in the early stages of the test, but high voltage will cause the SIR recorded value to recover in the later stages of the test (Figure 3). Therefore, it is very important that the test voltage should be the working voltage in the actual application.

 

 

 

3. The importance of monitoring frequency

 

 

 

    Current testing standards require that the SIR value of each test sample be measured three times within 7 days. From the above SIR time curve, it is obvious that such sparse data frequency is not enough to display the required details. For example, for the above-mentioned no-clean flux test, if the first test is performed at 48 hours, the test value is likely to be as high as 109 ohms, thereby missing the impact of initial dendrite formation. Visual inspection relies on the quality of the operator. Frequent electrical monitoring of dendrite growth is the most direct and certain detection method.

 

 

 

4. Effect of temperature

 

 

 

   For conventional RMA fluxes, high temperatures reduce SIR values, most likely due to increased moisture absorption (Figure 5). No-clean flux will evaporate at high temperatures, causing the SIR value to increase. Below 65°C, the SIR readings were very low using 15 and 30 ul of no-clean flux, but above this temperature the flux residue decomposed. Pay special attention to the relationship between flux dosage and measurement results. The amount of flux used in the test should reflect the amount of flux actually applied on the board.

 

 

 

3. SIR test method

 

              

 

 

 

1. Test preparation

 

 

 

A large number of chemical processes and chemicals are used in the modern PCB manufacturing process, such as solder resist materials, flux (including solder paste and rework tin wire) and conformal coatings. SIR is often used to prove the electrical insulation and electrochemical corrosion resistance status of products. The following SIR test is a test of various material combinations used in no-clean flux assembly of PCA. The work was tried to be consistent with the actual manufacturing process. The surface mount SIR sample used is also similar to the one used in the previous part (Figure 1), There are 4 comb circuits on this test board, but their wiring can be adapted to SMT flat package device mounting.

 

 

The test sample uses real process parameter settings. This is not just testing a bare board. The flux residue and thermal effects caused by the device will also be truly revealed. Some chemicals have good reliability on their own, but their application together can cause SIR values to decrease. In order to detect this phenomenon, SIR testing must be performed in the presence of the chemicals used in the actual process. Therefore, during the test process, two surface coatings, HASL and NiAu, were used, and the influence of the solder wire used for repair was also considered. This project was conducted under typical process conditions and was intended to provide a comprehensive set of materials compatible with typical PCA manufacturing situations.

 

 

 

2. Sample preparation

 

 

 

●HASL surface coating

 

Dip tin for 5 seconds at 255°C, standard flux.

 

 

 

●NiAu surface coating

 

Using palladium-activated Ni/Au from ATOTech, the bath temperature is 88°C and 6g/l Ni. Soaking time is 25 minutes (4-6um, Ni). The temperature of the gold plating bath is controlled at 84°C, and the infiltration time is 12 minutes (0.06um Au).

 

 

 

●Permanent solder mask

 

The sample is first cleaned with pumice stone, and the solder mask is printed with a 32.82 silk screen. Solder mask is applied to the entire sample. Dry at 85°C for 30 minutes before exposure to 450mJ/cm2. Make sure that only one QFP area on the coupon is exposed and half of the coupon has no solder mask after development.

 

The unexposed solder mask is removed with 1% anhydrous K2CO3 at 35℃ and 2.5Bar air pressure. Finally, the board is baked at 150°C for 1 hour. The thickness of the solder mask on the empty board is 26um.

 

 

 

● Flux

 

Manually spray flux onto the sample with a spray gun, then place the sample upright on the absorbent material for 30 seconds to remove excess flux. Wave soldering uses 63/37SnPb alloy, the soldering temperature is 250°C, and it is soldered in the air.

 

 

 

● Solder paste

 

Solder paste printing uses DEK 260 printing system with ProFloTM solder box. The thickness of the stainless steel drain template is 6mil, and solder paste is printed to the peripheral comb circuit to facilitate QFP mounting. Although both are filled with solder paste, the QFP is only placed on the SIR comb position surrounded by solder mask and placed by hand. The peak temperature of reflow soldering is 222℃.

 

 

 

●Hybrid assembly technology

 

In order to simulate the mixed assembly production situation, the sample was sprayed with flux and then passed through wave soldering, then solder paste printing and reflow soldering.

 

 

 

● Solder wire for rework

 

Use a small amount of solder wire to solder 4 places on each middle comb circuit. The solder flows on the lead to avoid soldering, and the temperature of the soldering iron tip is 350°C.

 

 

 

3. SIR test

 

 

 

1) The sample is placed on a 256-channel test rack and connected to the Auto-SIRTM surface insulation resistance tester. All test specimen labels are metal to avoid introducing possible contamination. Always wear gloves when handling specimens.

 

    

 

    2) To install 64 specimens on the test rack, 256 (4×64) connecting cables are required (Figure 2). This is achieved through the AUTO-SIRTM test system’s 34 cables grounded inside and outside the system. In this example 16 connections are used, each including 16 signal lines and 16 ground lines.

 

 

3) These precautions are to eliminate the influence of frictional static electricity on small current recording. The test specimen must be placed in the test box to ensure smooth surface airflow (see ISO/PW19455-17 trial draft for details).

 

 

 

   4) The test box is slowly heated to 40°C and the relative humidity is 93% (according to IPC-68-2-20). The purpose of using such a low test temperature (different from the traditional 85%/85% requirement) is to ensure that the temperature-sensitive components of the no-clean flux (mainly organic acids) can remain on the test board during the entire test process. This is becoming the accepted test condition for no-clean fluxes, with a test time of 7 days and a test voltage of 50V (equivalent to a voltage gradient of 125v/mm on a 0.4mm pad spacing).

 

 

 

   5) Measure the leakage current on each test sample every 10 minutes and calculate the logarithm (Log) value of the resistance. At the end of the 7-day test, the humidity was first reduced and then the temperature was reduced to prevent condensation on the test specimens.

 

 

 

4. Test sample matrix table

 

 

 

This list requires 58 samples because there are two surface coatings: HASL and NiAu. There are 4 comb circuits on each test specimen. Circuits A and D have solder masks and are tested with soldering and mounting of components; circuits B and C do not use solder masks. See the list for the process sequence. The code name AR of the conformal coating indicates the acrylic type, and PU indicates the polyurethane type. AR2 has a higher Tg value than AR1.

 

 

 

5. SIR test results

 

 

 

The test results are related to the combination of chemicals used in making the specimen. In the appendix of IPC J-STD-001B D-5.3 regarding the use of SIR for material and process compatibility testing, it is required that "the minimum value of SIR in each test must be greater than 1x108 ohms". We quote this result as the basis for judging the qualification of each material combination. /Fail (PASS/FAIL) criteria. Generally we find that material combinations greater than 1x1010 ohms ensure that the process is safe.

 

 

 

From Figure 3, we can see the impact of attaching or not attaching the device on the test results. The decrease in SIR value can be attributed to the fact that residues from the soldering process are absorbed into the area below the device by capillary action or diffusion. Moreover, the peak temperature of this part of the space will be relatively low due to the heat absorption of the device, which may result in the activator not being completely decomposed. The slight decrease in SIR value in the first 30 hours or so did not occur on the sample without the device. We can also see that the SIR value of HASL is roughly lower than that of NiAu (Figures 4 and 5). The no-clean flux will affect the SIR value of the two surface coating samples. However, in some material combinations, the SIR value of NiAu will be lower (see PU coating sample curve). A surprising finding is that PU coating will increase the SIR value in a high-humidity environment. When the test box was in the later stage of the test, the resistance of the test box dropped after the temperature of the discharged moisture dropped. The drop in humidity has nothing to do with the drop in temperature.

 

 

 

     If the sample is coated with a temporary ammonia-based solder mask of Latex, the SIR value will be significantly reduced, regardless of other chemicals (Figure 6). Using solder wire may cause the SIR value to increase or decrease under the same process conditions, which may be due to inconsistent flux dosage and heating applied in this way. This should be the actual situation of the production process, but the SIR value of all test samples is not lower than 109 ohms. The applied solder may also cause changes in the dimensions of the comb circuit, affecting test results.

 

 

 

6. Others

 

   

 

● In order to reliably detect dendrite growth, frequent monitoring is necessary;

 

 

 

● The impact of permanent solder mask on SIR is minimal;

 

 

 

● When the device is mounted on the board, the SIR value will be reduced due to the entrapment of residues. Although the NiAu board is sometimes affected by the combined effects of other substances such as coating, its SIR value is lower than that of the HASL board;

 

 

 

● But it is obvious that the flux used in HASL will reduce the SIR value;

 

 

 

● Temporary ammoniacal Latex solder mask will significantly reduce the SIR value;

 

 

 

● No-clean solder paste and flux have good compatibility;

 

 

 

● Except for the PU conformal coating which will significantly increase the SIR value, other coatings have little impact on the SIR;

 

 

 

● The compatibility of solder resist, flux, solder, tin wire and coating used in the current process can be evaluated using SIR values.