Dense packing of electronic components on both sides of circuit boards makes jobsite troubleshooting and repair difficult.

In normal situations, when an equipment shutdown is traced to the failure of a printed circuit board, the board is replaced, and the old one is discarded or returned to the manufacturer for repair. You don't have this luxury in an emergency situation, especially when a spare board isn't available. In these instances, you have to resort to some ingenuity and make the necessary repairs yourself so that your production equipment can be up and running ASAP. Let's discuss an actual emergency situation and see how the onsite repairs were made on a printed circuit board having surface-mount electronic components.

A maintenance call

Modern printing presses are an assembly of individual equipment, all of which must function together to produce the finished product. Very extensive controls are required for each piece of equipment, as well as additional controls to make individual pieces function together as a complete line. Precise speed control is an essential element because of the high speeds involved; DC drives are often used for this purpose.

Our example centers on a problem that developed with a water roller, one of the components of an offset printing press. Its motor would run for about 10 min and then stop. If the operator turned the power off for several min, then restarted, the water roller would operate and then stop again after about 2 min.

A call was made to the maintenance shop for an electrician, who made a visual inspection to make sure that there wasn't a mechanical problem and that the water roller was properly set up. The DC motor was checked and seemed to work properly.

The printing press was started once again, and the result was the same: The water roller operated for about 2 min and then stopped.

A quick check with a VOM showed that there was field voltage; however, the armature voltage measured only 0.75 VDC, rather than the expected 90VDC. The problem had to be with the DC drive's circuit board. And there was no spare board on hand.

A replacement board was ordered, but it would be two days before it arrived. This problem needed immediate attention, since this particular press was scheduled for a rush printing job.


The suspect circuit board was removed from the drive and its schematic diagram retrieved from the machine files in the maintenance shop. Its circuit was complex, with various transistor and operational amplifier arrangements going to the transformers that gate the SCRs (semiconductor controlled rectifiers).

As shown in Photos 1 and 2, the circuit board basically had large devices mounted on one side and miniature surface-mounted devices on the reverse side. Because of their small size, these active devices sometimes don't have identifying numbers printed on them. Thus, a maintenance electrician must refer to the schematic diagram of the circuit board to find out what type of device the transistor is and to determine its identification number.

This circuit board's schematic diagram had a note stating: "Engineering changes may have been made after publication date. Any departure from this drawing should be checked with the factory." Clearly, the first step was to trace the circuitry with a VOM to identify the items, check the foil continuity, and confirm that the diagram and board circuitry matched.

A VOM set at the "R x 100" ohmmeter range was used to check these items, with the larger components checked first. [ILLUSTRATION FOR PHOTO 3 OMITTED]. Continuity readings of transformers indicated they were OK. The large SCRs were removed from the board, as shown in Photo 4, and tested with an in-house developed SCR/transistor tester. On the board's other side, the transistors were checked for emitter-to-base shorts. Also, the diodes and capacitors were checked for shorts. Everything checked out OK. The trim potentiometers were cleaned with contact cleaner, as shown in Photo 5, and tested; they seemed to be in working order. Thus, the problem was narrowed down to the surface-mounted integrated circuits (ICs).

Replacing the ICs

Three ICs were identified while tracing the circuitry. Ohmmeter readings of the pins were compared; considerable differences were found between them, but it was difficult to tell whether they represented differences in the ICs or circuit configurations. The best bet was to replace all three ICs.

Surface-mount, large-scale ICs can have up to 80 pins, 20 on each side of a square package. These devices require specially designed desoldering equipment for removal. Fortunately, in this instance, the ICs had only 14 pins, as shown in the diagram on page 24. While the maintenance shop didn't have a vacuum soldering and desoldering station required to remove surface-mount ICs, it was possible to remove these ICs with a low wattage soldering iron and solder wick.

However, this was not the total answer. The pins have to be desoldered at the same time so that the package can be picked off the foil.

To remove the ICs in the manner required, a tool "made for the purpose," which is shown in Photo 6, had to be fabricated in the shop. A piece of heavy-gauge brass was bent into a "U" shape and an allen-head screw attached to act as a handle. A 6-32 set screw with nuts was then inserted through holes drilled into the flanges of the "U." This set screw was used for adjustment so that the brass "U" fit over the surface-mount chip and touched all pins. A sketch was made of the foil around the IC, just in case the foil was damaged during the removal.

The brass "U" was heated with a propane torch and, using needle-nose pliers as shown in Photo 7, was placed over the first IC. The chip lifted off the circuit board with no damage to the foil.

Replacement devices were purchased from a local electronic distributor. Using tweezers, the small ICs were held in place on the circuit board. Two corner pins were soldered, as shown in Photo 8, and an inspection was made with a magnifying glass to make sure the other pins lined up with the foil traces. These pins were then individually soldered to the foil using .020-in.-dia electronic resin-core solder.

Next, the magnifying glass was used to inspect the soldering job, as shown in Photo 9; two pins were found to be bridged with solder. To remove the bridge, a soldering iron and a solder sucker were used, as shown in Photo 10. The IC's pins were inspected again to verify the removal of the bridge.

Once the three ICs were replaced, the circuit board was placed back into the drive and the offset press turned on. Voila! The water roller operated properly.

The whole repair procedure took about five hours, which is considerably shorter than the two-day delivery time for the new circuit board. (With proper desoldering equipment, the ICs could have been replaced in only minutes.)

When the replacement circuit board arrived, the repaired circuit board was removed from the drive and stored in the shop as a spare board.