In Printed Circuit Board (PCB) manufacturing, quality control is extremely important. This article describes how to make the quality control process up to 20 times faster, with improved accuracy and reproducibility. This includes, but is not limited to, the quality of drilled holes, desmearing of drilled holes, the plating thickness of various metals and integrating of the final production board.
Test coupons are introduced in each production panel during the manufacture of printed circuit boards. This enables each panel to be tested without the actual production board being wasted. In order to confirm that product specifications have been met, the coupon is separated from the panel.
Continual monitoring of the process, rather than simply performing a final inspection, is vital as otherwise a great number of parts may have been produced and moved to the next process by the time the problem has been located. These parts would then need to be recalled.
The aim is the earliest possible detection of any deviations from required manufacturing standards, and to avoid adding value (more processing steps) to any product which is defective. For this reason, it is important to rapidly and accurately discover problems and implement corrections to the process as early as possible. If done, scrap rate is reduced, productivity increased and the operation becomes more cost-effective.
In these quality control tests, coupons from manufactured boards are extracted, mounted and ground/polished to the plane of interest. This article discusses numerous aspects of PCB coupon preparation, giving advice on how to attain the best accuracy and efficient during the process.
What is Micro-Sectioning of PCB’s?
A destructive technique for evaluating the quality of a PCB, micro-sectioning works by exposing a cross-sectional view of microstructure at a selective plane. This plane is normally the center of plated through holes or vias within +/-10% tolerance of accuracy of plated material thickness. As this is destructive analysis rather than an actual printed circuit, test coupons are normally used from the same board.
A specific coupon design has been introduced by IPC which can be included on each panel so that the coupon will experience the same process, as well as being representative of the remaining board. It is possible to remove this coupon for cross sectioning at any stage of production.
Figure 1. Examples of PC boards (a) without reference holes and (b) routed and drilled ready for use
In a great number of cases, it is possible to manufacture coupons into the board design and a test coupon in order to ‘press out’ of the board, allowing the quickest transfer of coupons for testing. If the test coupon is not designed for pressing out, or if another area of the board is needed, it is vital to consider that the cutting method ought not to introduce any defects in the sample.
If possible, aggressive methods such as the use of Bandsaws or Punch should be avoided as they inflict excessive damage in the sample. Routing or cutting with a Diamond Wafering Blade and coolant are the best methods for cutting the sample. If you require a sample from a populated board or production piece, it can be cut with the use of a silicon carbide (SiC) abrasive cutting wheel on a high speed rotary tool. It is important to ensure that the cutting is done away from the specific area of interest at all times.
Challenges in Preparing Vias and Plated Through Holes
- The final polishing quality should represent the true structure of the sample.
- There should be no scratches or edge rounding. This creates errors in measurement.
- During potting of the sample, there should not be any sample orientation error. This ensures the sample coupon does not have any tilt once it has been mounted (Figure 2a).
- It is crucial to hit the center of the target in order to avoid errors in dimensional measurement (Figure 2b).
Figure 2(a). Problems associated with tilt or target deviation
Figure 2(b). Change in apparent plating thickness due to improper targeting
Exceptional control of the grinding and polishing steps is necessary in order for each step to completely remove deformation from the previous step. True microstructure with an absence of edge rounding should be shown by the final polishing step, which should also find the target feature with accuracy and good alignment.
Several different mount shapes, sizes, and materials have been used conventionally. These depend on the application. Preparation is consequently inconsistent, and the quality of the result tends to vary.
In a lot of PCB manufacturing labs, numerous SiC papers are used to manually grind from course to fine grit size on a rotating table with water. Subsequently, water base Alumina suspension is placed on a napped cloth to perform a final polish. Regular stops to visually inspect the sample are required during this conventional procedure, in order to confirm the current location through the sample. A significant risk exists that the desired point is ground through, which may lead to a complete loss of that production sample.
Table 1 shows a typical conventional preparation method. Assuming no operator errors, each sample can take between 17 and 25 minutes or more to complete. Historically, grinding and polishing has been performed on a rotating plate using the manual speed setting as shown below.
Some applications make use of all the papers listed in sequence, and others would use between three and four, spaced through the range. Each paper would be used for around one minute.
Table 1. Conventional preparation method
||Base Speed (RPM)
|Plane back SiC papers with paper holding band: (Grit/FEPA Size):
120 (P120), 180 (P180), 240 (P280), 320 (P400), 600 (P1200), 800 (P1500), 1200 (P2500)
|MicroCloth with suspension MicroPolish Alumina powders with water from: 1.0, 0.3, 0.05
Typical polishing time: 3-4 minutes
|Typical grinding/polishing time per sample
|Processing time including cleaning between stages per set sample
(15 s for sample each step, 1 min for cloth)
The following limitations come with this approach:
- Sample throughput is low.
- It is inconvenient to switch different grit sizes of SiC papers during the grinding process.
- Quality is inconsistent as a result of operator variability.
- Cleaning can cause cross-contamination if not handled properly.
- Smearing can be caused by an uncontrolled load, which can hide crucial features like micro-cracks, delamination, or voids.
- Polishing relief can be caused by uncontrolled times. This makes both focused and accurate plating thickness measurement difficult.
- Frequent stops are required in order to inspect the sample during processing. This is necessary to ensure accurate targeting.
- A high degree of operator skill is necessary in order to maintain the plane of grind and hit target.
Production delays may be caused by running one sample at a time with poor preparation. This may also lead to a poor interpretation of results and too little statistical data for analysis, and ultimately costs will be high and business may be lost.
In order to overcome these problems associated with the conventional method, Buehler has created accessory kits for semi-automatic grinder-polisher equipment. These are called the PC-Met High Volume accessory, and the PWB-Met Small Hole accessory.
PC-Met Precision High Volume Printed Wiring Board Accessory™
This option is perfect for users who have a high volume of samples and can accommodate as many as 36 coupons (at 0.125” thickness). It can target the center of features accurately, down to 0.008” (200 μm). As many as six coupons are placed in each sample position, which are held in place by two indexing pins with a 0.094” diameter. All of these samples are simultaneously ground and polished to target.
This option is designed for users who have smaller target feature sizes. The center of features can be accurately targeted, down to 0.004” (100 μm) and as many as 18 coupons can be accommodated at once (up to a thickness of 0.125”). There are six specimen positions, each having three slots.
Each coupon occupies a slot, and is held in place with two short pins of diameter 0.094”. As each coupon is pinned separately, the accuracy of targeting the hole’s center is far higher than the PC-Met accessory.
Accurate micro sectioning of coupons within the target requirement is produced by both of the Buehler Accessories for Printed Circuit Boards. True microstructing at the area of interest is also produced. These occur within four preparation steps.
These accessories also have the benefit of excellent reproducibility, as the final result is not dependent on an operator, and every sample maintains the correct orientation and reaches the target area.
A standard coupon is needed in order to use automation to target features. This establishes the distance between the locating pins and the plane to which one needs to grind.
Theoretically, this distance can be altered manually by setting up the distance appropriately, however the following standard coupon dimensions are recommended (see Figure 3).
Figure 3. Dimensions of test coupon for use with PC-Met and PWB-Met
Once the coupons have been mounted into their holders, the operation of PC-Met and PWB-Met is essentially the same.
Set Dial to Zero Position
As mentioned above, it is necessary to set the total grinding distance from the feature of interest to a reference pin on which the sample is mounted. As shown in Figure 3, in the standard equipment the distance is set to 0.150”. By grinding to diamond stops in the fixture, the perfect target is reached.
It is crucial to set the diamond stops as accurately as possible in order to ensure accurate targeting. In order to do this, a Dial Gage is used. When a calibrated reference pin is used to set the Dial Gage to zero, the target plane is set at the center of the holes. A reference pin of 0.1972” diameter is required for a standard distance of 0.150”. Different reference pins can be used for other distances. The calculations for these pins are as follows:
Distance from Reference holes to Target holes + 0.047” (half the value of 0.094” diameter index pin).
Diamond Stops Setting
There are six diamond stops on the Sample Holders of both accessory kits – three “Short” and three “Long” Diamond Stops. The purpose of the long diamond stop is to set the position for the first grinding step, while the position of the second grinding step is set by the short diamond stop.
The Long Stop’s recommended settings are 0.007” (177.8 μm) and the Short Stop’s recommended settings are 0.003” (76.2 μm). By adjusting the Sample Holder fixture’s position relative to the Dial Gage, these distances can be set (Figure 4). This ensures that the grinding surface plane will not pass beyond the center and remains above the target plane.
Figure 4. Setting the position for (a) diamond ‘Long Stop’ and (b) diamond ‘Short Stop’
The coupons are able to be held firmly in place in the Sample Holder as they are mounted on to pins. Prior to mounting, the coupons are placed on to the pins with the help of a positioning tool (Figure 5). The High Volume PC-MET system utilizes numerous coupons placed on two long pins (60-5053).
Regular spacing is maintained by the pin loading tool. It is important to keep space between each coupon in order to make sure of the best penetration of the mounting media. The high accuracy PWB-MET system utilizes coupons loaded individually onto two short pins (60-5186).
Figure 5. Coupons are loaded on to the pins with the aid of the Pin Loader
Resins used traditionally, like polyesters, can function as health hazards and are liable to produce inferior preparation results. Either too much heat is generated by epoxies, or they take an extremely long time to cure. SamplKwick acrylic resin has been developed by Buehler specifically for this application. It has a quick cure time of less than 10 minutes and brilliant wetting characteristics. It is recommended that you use a mixture of one part powder to one part liquid.
Prior to mounting, each of the coupons is dipped into the liquid part in order to improve the surface coverage of acrylic material inside the holes. This is especially important when using smaller hole sizes.
You should fill each cavity to the top edge of the coupons with SamplKwick (Figure 6b). Each cavity should be filled to an equal height without overfilling, as surplus mounting media may increase grinding times without producing any benefit.
Mounting media should be used to fill any empty cavities in order to ensure an even level of support around the Sample Holder. The molding plate can be removed once cured, and the sample holder is then separated ready for both grinding and polishing (Figure 6c).
Figure 6. (a) Coupons mounted into holder prior to pouring the resin (b) cured mounts in the mounting assembly (c) the grinding fixture with the molding plate removed
Step 1: Coarse Grinding
Using silicon carbide paper, it is possible to quickly grind through polymer based PCB coupons. The opening step begins with 240-320 grit SiC paper and running water until you reach the Long Diamond Stop. The SiC paper should be changed every two minutes and grinding continued until all of the Diamond Stops contact the surface.
A useful way of confirming contact with the grinding surface is to mark the end of each stop with permanent marker.
In order to prevent cross-contamination, thoroughly rinse the specimens and holder with water after each step. Back out all three of the Long Stops by half a turn.
Step 2: Fine Grinding
Using 600 grit SiC paper and water, grind until all Short Diamond Stops have been reached. The specimens and holder should then be rinsed thoroughly with water.
Back out all three of the Short Diamond Stops by half a turn. There should be a reading of 7 mil or less on the Dial Gage after step 1, and 3 mil or less following step 2.
Step 3: Coarse Polishing
Use 3 μm Diamond Paste with MetaDi Fluid. Evenly apply the paste to a TexMet C cloth and spray enough MetaDi Fluid for lubrication (paste can be re-applied every 5-10 uses for repeated use). Polish for three minutes, after which you should clean the sample and holder with water. Spray with Ethanol and then dry with warm air.
In order to confirm that SiC scratches are completely removed during this process, inspect the prepared surface.
Step 4: Fine Polishing
In the final polishing step, 0.05 μm alumina suspension (MasterPrep) is applied to a medium napped polishing cloth (MicroCloth). It is then polished for 90 seconds. Rinse with running water during the final 15 seconds of the polishing cycle, in order to clean the polished surface and the polishing cloth. Do not over-polish during this step.
If you run for extra time, rounding will occur at the edges of the sample – particularly on soft material such as tin-lead coating. This can interfere with the taking of accurate measurements.
Depending on the number, thickness, and type of coupon material used in both accessory kits, there may be a requirement to vary the cross-section preparation procedure.
Figure 3. Dimensions of test coupon for use with PC-Met and PWB-Met
Table 2. Modern Preparation Method
|Head Speed (rpm)
||CarbiMet SiC Paper
||until Long Diamond stops are reached*
||CarbiMet SiC Paper
||600 grit (P1200)
||until Short Diamond stops are reached
||0.05 µm MasterPrep (Alumina)
|Typical grinding/polishing time per set of samples
|Processing time including cleaning between stages per set of samples
(1 min for holder each step, 1 min for cloth)
|Processing time including cleaning between stages PER SAMPLE (PC-MET)
|Processing time including cleaning between stages PER SAMPLE (PWB-MET)
*Change the SiC paper after two minutes if all Long Diamond Stop are not reached.
Discussion and Conclusions
The per-sample processing time can clearly be greatly reduced by using semi-automatic preparation and multi-sample target grinding equipment. The time to grind and polish the specimens in this example was reduced from roughly 20 minutes to shorter than one minute. This clearly represents a significant benefit to any high-volume laboratory which processes multiple samples.
Additionally, the improvements in reducing the level of reworking, increasing the consistency in finish, improving repeatability between operators and over time, the capability to process numerous sample types with fewer consumables, and enhancements in the reproducible targeting of features, will improve the performance of laboratories producing fewer samples as well.
This information has been sourced, reviewed and adapted from materials provided by Buehler.
For more information on this source, please visit Buehler.