What to look for in electronics troubleshooting

From MKRD.info Wiki
Jump to: navigation, search

Troubleshooting of real electronics assemblies and devices requires a systems approach and understanding of basics of reliability analysis.

Electronics assemblies fail due to one of the following:

  • A process problem
  • A person problem
  • A part problem (unlikely)

A process problem is one found in the board manufacturing, population, and assembly stages. Problems at this stage are usually parts which are installed incorrectly, solderability problems (shorts, lifted pins, cold joints, joint fractures, damaged traces, etc). This failure possibility is covered by a thorough visual inspection of the board, preferably at 4x to 10x magnification.

A person problem is one of a manual process. There are more chances for an incorrect implementation of a specified manual rework. In general, work done by a manual operator should be inspected by a qualified electronics inspector or electronics technician. I have seen the following done by a human rework operator: work on a wrong part, damage to the board due to rework in the area (cut traces with a knife, overheated components, damage to solder, damage to parts), rework not actually implemented, part backwards / orientation, part not soldered down, bad discrete part.

A part problem, usually being suspect by a novice troubleshooter, is actually the least likely occurrence. When attached to the board properly (by the previous two steps discussed above), and when driven properly (i.e. no overload, overvoltage, shorts, etc are present), modern semiconductor components are very reliable. There are a few exceptions:

Power MOSFET drivers fail frequently. They often fail after the MOSFET itself fails. When troubleshooting one or the other, always check its corresponding part before powering up the unit for verification. An electronics troubleshooter should be familiar with the many harsh signal conditions with which a MOSFET driver must operate. These include, but are not limited to: excessive drive current required (low or no gate resistor), current being forced into the output pin (when MOSFET drain swings in voltage, current passes thru the gate-to-channel capacitor), transients and ringing (when attempting to drive a MOSFET quickly), etc. A technician should look at datasheets of a few MOSFET driver ICs to get familiar with conditions such as these.

Digital CMOS gates. These fail from overvoltage (some parts cannot tolerate more than 5.5VDD), excessive fan-out (overloading), logic latchup (input pins left floating), race condition (two output pis connected together, one pulls high while another pulls low).

ICs can also be destroyed by: overheating, overload, floating pin (CMOS), transient (FET driver, inductive kickback, ESD), bad or unconnected external component, misbehaving power rail.