BE Radio March/April 1996 RF Engineering Solid-state RF By William Fawcett ------------------------------------------------------------------------ William Fawcett is president of Mountain Valley Broadcast Service, Inc., a broadcast engineering firm in Harrisonburg, VA. ------------------------------------------------------------------------ If you have been servicing broadcast transmitters for more than a few years, chances are you are already "up-to-speed" on tube-type units. Being knowledgeable about tube transmitters will probably remain a uniquely marketable skill for another decade or two. Today, of course, the proliferation of solid-state devices in modern RF power amplifiers requires familiarity with another set of troubleshooting techniques. The first and second generation of solid-state RF amplifiers were (and still are) problematic. Devices had to be closely matched, and they often would self-destruct while repairs were being made. If you have such a unit, check with the manufacturer to see if there have been any upgrades. Another problem with the older units is the current availability (and cost) of replacement parts. One major manufacturer now suggests the wholesale replacement of their failure-prone IPA shelf with an outboard power amp. It's a worthy idea. Given the variety of devices available and the ever-advancing state-of-the-art, almost every transmitter design is unique. Fortunately, a deep knowledge of circuit theory is not needed to effectively troubleshoot most RF power amplifiers. Even with rapidly changing technology, the fundamentals of logical troubleshooting remain constant. Baseline measurements The first essential is to establish a baseline. While the transmitter is new or properly operating, record every reading available. It is best to have the readings made on your specific unit, at its actual nominal operating power and line voltage. In a pinch, the manufacturer's test or generic data is better than nothing. Aside from the usual voltage, current and temperature readings, inter-stage forward and reflected power measurements are very useful. When trouble occurs, these baseline readings can often be used to locate the problem down to the module, or perhaps to the actual component. Reducing the power level or drive may enable you to take readings on a transmitter that shuts down at full power. Using a dummy load and a wattmeter will allow you to eliminate the possibility of an external problem. Next, do a complete check of fuses throughout the unit. Some transmitters incorporate "fuse-finder" readings or have front-panel test jacks. Follow the instructions in the manual. If you're working on an AM transmitter, you may want to take some preliminary pulse-width measurements with an oscilloscope. Without these measurements it is impossible to separate a modulator failure from an RF power-amp problem. Armed with this information, you are now ready to find a quiet room and review the block-diagram and schematics of the unit. More problems can be solved with a cup of coffee and a sandwich than with a room full of fancy test gear. If the problem is not obvious, swallow your pride and call the factory. You could save hours, or even days of hair-pulling, because they are already aware of the most common failure modes and peculiarities of your unit. Having the baseline and present readings available will give the factory technician a good starting point. Three techniques Once you have isolated the problem to a certain section, three basic troubleshooting techniques will be helpful. Substitution of a module will often verify the location of the problem. If you do not have a replacement module, and more than one module is used in the transmitter, swap it with another module and see if the problem moves to that section. Be aware that this can sometimes cause damage to other components, so use this technique with caution. Make a list of test measurements with each swap, and if possible swap only one module at a time. (In the event of a catastrophic failure with multiple blown-modules, module-swapping is guaranteed to drive you buggy.) A less dynamic approach is to take resistance readings on the suspect module and on another board that is known to be good. A DVM with a diode position is often helpful here. Removing the solid-state device and comparing resistance readings with a known-good device may allow you to locate the defective part. Triggering a MOSFET for dual-state testing may require a high-output-voltage ohmmeter like the venerable Simpson 260. Testing a blown device may allow a determination of the normal failure mode, which will assist in identifying other bad parts. Measure from all three leads, in both polarities (six measurements), looking especially for blown junctions or high leakage. A third technique is the "shotgun" approach, which on the surface seems extremely wasteful. In this case, replace all suspect components -- perhaps all semiconductors on the defective module. When you factor in the value of your servicing time and the expense of transmitter downtime, the shotgun approach may actually be the most cost-effective. In cases where the component "ohms out" OK but breaks down at full power, the shotgun method beats all others. The obvious downside of this method is its need for ample stocks of replacement devices. Replacing components Although AM transmitters may use socketed TO-3 type devices, the higher frequency FM transmitters usually employ a tab-type device soldered to the printed circuit board. For these types, in-circuit testing is preferable. If you must remove the device, consider nipping the four tabs off prior to desoldering. Such removal requires less heat, and preserves the printed circuit pattern, which may otherwise suffer damage after a few replacements. Of course, this is not practical if you are removing the device for an out-of-circuit test and you don't have many replacements on hand. You may be reluctant to perform destructive removal on a hundred-dollar device, but replacement boards are expensive and hard (slow) to obtain. Buying generic replacements in quantity will often give you three components for the price of one from an OEM. Having such a stock on hand will allow you to effectively utilize the shotgun method and destructive removal when necessary. If you use a generic replacement, be sure it is appropriate. It is not uncommon for RF power amps to incorporate hand-selected components, or to use a particular manufacturer's device that may have a higher rating. If the components are utilized in pairs, or quads, or complements, replace all devices in that set with one made by the same manufacturer. Otherwise, one device may take on more than its share of the load and subsequently fail. Pay close attention to the means of mounting the solid-state device to its heat sink. It is imperative that you properly re-assemble the unit in the same manner. The device may or may not be insulated. Utilize the same type heat sink compound, and be sure to use proper handling procedures for static-sensitive devices like MOSFETs. Stable operation Vacuum tubes age and solid-state devices don't. With most modern circuits, touch-up tuning is not required. For broadband circuits, tuning is not even an option. It's not uncommon for many years to elapse between failures of modern solid-state power amplifiers. Ironically, this reliability makes engineers less familiar with the inner workings of such complex devices, which can turn a major problem into a most befuddling experience. The methodical approach of logical troubleshooting, coupled with adequate stocks of replacement parts, will provide you with a a problem-solving technique for solid-state transmitters that really works. Copyright 1996 Intertec Publishing Corporation