IPC-4552.pdf - 第15页

IPC-4552 Proposal June 2001 0 1 2 3 4 5 6 SS XRF std XRF XRF type Means and Std Deviations Level Number Mean Std Dev Std Err Mean SS XRF 240 2.66542 1.24144 0.08013 std XRF 240 3.08750 1.47084 0.09494 Fig 3: compar ison …

IPC-4552 Proposal June 2001

Nickel Thickness:

Determining the required thickness for the nickel barrier layer was arrived at easily through consensus. From the

input on “normal” plating times being used in the industry a specification range of 120 microinches to 240

microinches was agreed upon. At the 120 lower level it was felt this provided more than enough basis metal to act as

a diffusion barrier to copper migration as well as providing a suitable stable metal to solder to. The concern with

excessively thick nickel was not from a soldering viewpoint rather from the use of the finish for compliant pin and

the increase in insertion force required to overcome the reduction in plated through hole diameter.

Gold Thickness:

The immersion gold thickness was not that straight forward. A series of questions were raised:

• How much gold makes a good nickel solderability preservative?

• How much gold is needed for a minimum shelf life of I year?

• What are the processes out there capable of?

• Is better? Is more gold possible?

• How much gold is needed for “Contact Resistance”?

In an effort to answer these questions the committee designed and executed a series of capability studies. The studies

included:

• A capability study to determine what thickness of gold is the immersion process capable of.

• Solderability analysis by Wetting Balance method to determine the minimum gold thickness that is solderable.

• Solderability after aging.

• Contact Resistance testing using carbon membrane switches.

Round Robin Testing:

The committee has five chemical suppliers of ENIG, for the list of participants. What became apparent early in the

specification writing meetings was the variation in chemical management from the five different suppliers.

A quick round robin test was set up to determine the range of nickel and gold deposits that were achievable by the

five suppliers. The panels for the deposition rate test would also be used for solderability evaluations. All the panels

to be coated with ENIG were supplied from one PWB supplier and would be acid copper plated with no solder

mask. Based on the variations in chemical bath control etc, it was decided to level the playing field by having all the

nickel samples plated in baths at 1 MTO and all the gold baths at 25% of their useable life.

The test would consist of plating a maximum of 10 panels all at the same time in the nickel bath; some vendors

could only do 8 panels due to basket size constraints. They would all be plated at the same time to ensure correct

bath loading so that a statistically significant mean and standard deviation value could be obtained. For the gold

bath, a panel would be removed from the plating bath every two minutes until all of the panels were removed. Since

the gold bath is an immersion bath, the loading factor is not as much of an issue compared to the electroless nickel

bath.

Measuring Gold Thickness:

The ability to accurately measure the ENIG deposit could potentially be a paper/test method document in its own

rights. The thickness of the gold deposit poses some fundamental problems. A number of the committee members

have experienced measurement capability issues with their customers. These have ranged from using the wrong

calibration standards, i.e. using electrolytic Nickel electrolytic Gold standards

Page 12 of 10

Download From http://bbs.infoeach.com

IPC-4552 Proposal June 2001

SS XRF std XRF

XRF type

Means and Std Deviations

Level Number Mean Std Dev Std Err Mean

SS XRF 240 2.66542 1.24144 0.08013

std XRF 240 3.08750 1.47084 0.09494

Fig 3: comparison of Gold thickness values by XRF machine type

Fig 4: Wetting Balance coupon

FR4 0.062” thick, acid copper plated

6 X 3 mm SMD pads/side

Page 13 of 10

Download From http://bbs.infoeach.com

IPC-4552 Proposal June 2001

vendor D vendor E

vendor

Means and Std Deviations

Level Number Mean Std Dev Std Err Mean

vendor D 300 4.07967 1.59078 0.09184

vendor E 300 2.99307 0.95200 0.05496

Fig 5: Comparison of gold plating thickness variation by vendor for similar bath life conditions

110

120

130

140

150

160

170

180

190

vendor D vendor E

vendor

Means and Std Deviations

Level Number Mean Std Dev Std Err Mean

vendor D 300 159.624 12.9179 0.74581

vendor E 300 133.816 7.6477 0.44154

Fig 6: Comparison of nickel plating thickness variation by vendor for similar bath life conditions

Page 14 of 10

Download From http://bbs.infoeach.com