smt reliability vs thru hole

Dave: I haven't located something to support your contention, yet, and I'm not going to comment on the factors that drive reliability beyond component selection, because I know that you've got some sweaty-palmed sales type wringing his hands waiting for THE ANSWER. Anyhow, this isn't IT, but it is intriguing:

In 1991, Janet Sterritt, of Hollis Automation, compared the defect ratio between SMT and through hole connections when wave soldering.

BOARDS USED IN THE TEST

Board type 1: Mixed technology with top side surface mount, bottom side wave soldered surface mount, and leaded through hole components. It measured 2"X8"X0.128" and contained 347 bottom side solder joints. The bottom side SMT components were leadless capacitors, SOTs, and SOICs. Of the 347 bottom side connections, 201 were surface mount and the remaining 146 were leaded through hole. Not a simple board. Board type 2: Primarily leaded through hole comprised of DIPs and a few bottom side chips capacitors. It measured 8"X10"X0.128" with a total of 2002 solder joints, comprised of 1980 through hole and 22 surface mount connections.

SOLDER MACHINES

Standard machine: Finger conveyor, 4 ft/min, foam fluxer, SMC OA flux, 4 ft IR bottom side preheater, dual wave, 500�F, SMC grade solder, air knife, finger cleaner. Optimized machine: Finger conveyor, 4 ft/min, foam fluxer, SMC OA flux, 5 ft IR bottom side preheater, 2 ft top side preheater at wave end, dual wave, 470�F, SMC grade solder, air knife, finger cleaner.

On the optimized machine, changes to the preheater were made to maintain similar 3.6�F/sec ramp rates and the 212�F delta T rule. Air knives were set differently on both machines. Board type 2 required a 1/8" greater immersion than Board type 1.

METHOD

Machine profiles were set-up using the thermocouple, place on the bottom side of the board and profiler. The standard machine was intentionally set to the best profile it could achieve on the boards. The profile on the optimized machine was intentionally set to "match" the profile of the standard machine. 100 of each board type were run on each machine. The boards were assembled, waved, and inspected at the same time. The same team of inspectors visually inspected both boards. Inspectors had standard defect definitions and didn�t know which machine the boards were run on.

RESULTS

Board type|Standard machine|Optimized machine|Defect ratio 1|1|0.37|3X 2|1|0.07|14X

Board type 1 showed an improvement of 3 times fewer defects on the optimal machine versus the standard machine. Board type 2, primarily through hole, showed 14 times fewer defects than the non-optimal machine. This show the enormous effect that optimizing process profile can have on reducing solder defects.

Good luck

Dave F

reply »

Dave: Try: http://rac.iitri.org/ they might be able to help. Good luck. Dave F

reply »

Dave: Find:

"Comprehensive Surface Mount Reliability Model Covering Several Generations of Packaging and Assembly Technology", by J. Clech et.al. IEEE Trans on Comp. Hyb. and Manuf. Tech. Vol. 16, No. 8 Dec. 1993, p.949. It listed cycle to failure and damage assessment based on stress test, storage, transport, operation, worst days and office enviornment work conditions.

Dave F

reply »