| Operation Peek Inside |
Print |
Backplane X-ray Analysis
This is one of those things that drive people in this business mad. A very expensive and difficult board, which once on the job has to endure in a harsh environment, was manufactured successfully. All the electronic and environmental testing was completed and the board went into service. Then, to everyone’s shock, months later the board began to perform erratically in the field.
 |
| Figure 1: Open solder connections at component pins. |
|
Of course the failure was “high visibility” and everyone is under the gun to find out how and why this is happening. And fix it! The customer urgently directed all possible resources to bear and after a week or so of testing found the source of this very sneaky problem. They discovered a dual challenge: a very thick board and leads that were (on occasion) contaminated – most likely oxidized.
There was only one component type involved and this component was placed at about a dozen locations per board and not every component demonstrated the problem. What happened was, at the affected holes on the through-hole component, barrels incompletely filled with solder, which caused the solder connection with the component lead to occasionally fail after the board had been in the field working for some time. Did I mention these boards were performing a mission critical function? Sorry if I missed that.
When the boards arrived at Circuit Technology Center our challenge was to first identify the holes that were inadequately filled and fill them. At the same time the customer wanted to ensure that all pins on the problem part were reflowed to ensure they would forevermore demonstrate proper wetting. Once this was complete, there had to be proof that the holes were all filled to everyone’s satisfaction. To complicate matters these boards were covered with a thick conformal coat.
Here’s a look at some of the particulars.
Tools and equipment:
Air Vac BGA Rework Machine with 280DVG Nozzle
STTC 126 soldering tip with Metcal soldering station
Water Soluble Flux (Kester 450)
X-ray system
Procedure – X-ray:
To facilitate location of under-filled barrels and organization of this lengthy process - from x-ray through various rework steps and back through x-ray again - each location to be reworked had to be accompanied by a lead (as in the metal lead) numeral placed on the solder side of the board at the rework location. This lead numeral will be seen in the x-ray and would uniformly identify the specified locations.
The rather large board had to be placed – solder side up - at about a 30 degree angle to the x-ray emitter head in order to permit full view of the barrel length. By the way, due to various board and component constructions, not all through hole components can be viewed in this fashion – we were a little lucky in this case.
Once the board was under view, each location was photographed and a marker placed on the photo noting the holes that were clearly under-filled. Figure 1 is an example of an x-ray snapshot of voided plated through holes.
Procedure – Reflow:
Now came the hard part. It’s easy enough to talk about reflowing ground planed pins on a .150-inch thick conformal-coated board, but to actually do it???
In order accomplish this task without damaging non-affected areas the board had to be prepared by placing thermal resistant tapes and heat deflecting material on the surrounding surface areas and components to protect the unaffected components and the solder side conformal coating wilting.
A vacuum-desoldering tool was used to remove the conformal coating from the solder-side leads to permit flow and fill at those locations. The interesting thing is that this rework was going to be performed on a BGA rework machine in order to maintain the board at a steady state high temperature using the BGA rework machine’s bottom heater plates to warm the whole substrate and the topside air nozzle to drive even more heat into the specific rework location.
In order to maintain control of this amount of heat, thermocouples were placed under the rework locations. Thermocouples were also placed to monitor the heating of sensitive devices near the rework area.
 |
| Figure 2: Correctly filled solder joints. |
|
The board was slowly heated until the proper base temperature was reached then a vacuum de-soldering iron was used to suction out the existing solder ridding the barrel of any suspect material. It’s useful to note that the technician operating the procedure had to be fully protected from the general heat of the machines and the board – not an easy task. It was very warm at that station.
The final step of the soldering phase was to add solder at the target barrel(s) using the soldering iron and wire solder. This process was repeated until all affected plated through holes were properly filled.
X-ray Inspection:
The board was then cleaned in a de-ionized water washer. X-Ray Inspection: Once again the board was x-rayed at every location. The numbers were put back in place and the formally noted failures were noted as filled. Figure 2 is an example what that looks like.
That’s the quick and dirty of it. In reality it wasn’t quick, but it was hot, down and a little dirty.
The x-ray inspection on either end was very tedious, while trying to maintain organization during the normal distractions of a day and reading x-ray images (fuzzy white to grey to black) can be blinding. Then, of course, the rework itself was hot and demanded great patience and skill to avoid over and under-filling the holes. It ain’t always easy, but somebody’s got to do it.
Several members of the Circuit Technology Center team contributed to this feature story.
|
|
CircuitMedic