Troubleshooting Variable Speed AC Motor Drives

Solid state electronic AC motor drives are becoming increasingly more commonplace within industrial and commercial facilities. They control air handlers, chillers, pumps, conveyors, machine tools, mixers, draw lines, and numerous other applications once considered either strictly DC applications, or constant speed.

The increased usage of variable-frequency drives (VFDs) presents the maintenance and service community with both the opportunity and challenge of learning to maintain, troubleshoot, and operate this equipment.

First, modern VFDs are reliable, in fact, they are extremely reliable. Keep this fact in mind if a continued or repetitive failure is observed on a particular application. It usually isn't the VFD, regardless of the number of fingers pointing toward it. Unfortunately, this hasn't always been the case.

Back in the late 1970's when I first began doing field service on AC drives, the typical service call involved pulling the charred remains of drive components out of a cabinet, and listening to the plant personnel remembering exactly what they were doing when the VFD exploded. Just in case you're wondering, explode is not an exaggeration. I've seen cabinet doors blown off the their hinges, foil traces hanging from the circuit boards, and shrapnel all over the motor control room.

I arrived at one site in Texas to repair a 200 HP VFD, where in maintenance supervisor, prior to my arrival, had picked the unit up with his fork lift, and thrown it into the dumpster. He threatened to kill the next OEM that installed one in his plant. Reliability, top say the least, wasn't what earlier versions of VFDs were noted for, Fortunately, that has changed.

A block diagram of a typical Pulse With Modulated (PWM) VFD is shown below, PWM drives are the most prevalent type of AC drives being sold today. You'll note that the AC line is converted to DC (within the converter Section of the drive) and the synthesized back into a variable frequency, variable voltage output. The theory behind this is, motor speed is varied by changing the frequency, and motor torque is maintained by keeping the volts to frequency ratio constant (for most applications).

How do you troubleshoot it? Not to disappoint the electronic techs out there, but you are going to do very little electronic circuit board repair on a modern, microprocessor based motor drive. Fortunately, most failures are not on the circuit boards, but rather within the power sections, and that isn't really difficult to troubleshoot. Companies like ours can repair these boards, but the typical plant maintenance person simply isn't going to see enough of the failures to develop any expertise in repairing them.

Effective troubleshooting, whether on a VFD or an old automobile requires a methodical approach. I have generally observed two techniques, one is the classic divide and conquer, and the other which I'll term stochastic. The classic method is taught by most technical schools, and is particularly effective when knowledge of the equipment is only rudimentary. The classic method is best illustrated in figure two. Assume a signal is present at A, but is not at B. What are the most number of steps a good troubleshooter would use to isolate the box that isn't passing the signal?

The answer is three! A classic troubleshooter would divide the circuit in half, by checking between boxes 4 and 5. If the signal were absent, then the next check would be between blocks 2 and 3. If the signal were present here, then a final check between blocks 3 and 4 would isolate the problem to either block 3 or block 4. Remember, divide and conquer!

Stochastic troubleshooting takes advantage or more in-depth knowledge and experience, and attempts shortcuts. The term stochastic refers to an educated guess, based upon random tests and observations. Basically, the more familiar someone is with any particular piece of equipment, the better they can deduce the problem from random tests, without a time consuming disassembly in order to make measurements, as would be required with the classic method of troubleshooting. This is why I often stress to young drive technicians the necessity of understanding drive theory. That theory, however, is beyond scope of this article.

So how can you quickly troubleshoot a dead VFD?

First be careful. The capacitors within the power section can maintain a lethal charge even after the power is removed. Don't put your hands into the power section before determining that the capacitor voltage has been discharged.

With the power off, check the power sections. Here's how. Place your digital VOM in the Diode check mode. Find the positive DC Bus (Sometimes this brought out to terminal, sometimes it isn't). Place your negative lead on the + Bus, and then check in turn with your positive VOM lead each incoming phase. You should read a diode drop from each phase. If it reads open, then the charge resistor (see figure 1) is open. This is a common failure.

Next place the positive VOM lead on the - Bus, and place the negative VOM lead on each incoming phase again. You should read a diode drop, not a short or an open. Place both VOM leads on the Bus, one on the + and the other on the -. You should read the capacitors charging; you should not read a short.

To check the inverter section, place the positive VOM lead on the - Bus, and reach outputphase by placing the negative VOM lead on each one. You should not read a short, and in fact should read a diode (there are diodes connected across each output transistor). Check the remainder of the inverter section by placing the negative VOM lead on the + Bus, and checking each output phase again with the positive VOM lead. You should not read a short, but rather a diode drop again. If you read open from either of these checks, then the Bus fuse is open (see figure 1) ( The charge resistor and the Bus fuses may be in the + or - Bus, depending upon the manufacturer).

If no problems are present within the power section, and the unit will still not go, you have it either connected improperly, programmed wrong (the most common problem we find), or you have a bad circuit board.

If you do have a shorted transistor ..... here's a word of caution. If you have an older generation PWM drive which uses Darlington transistors in the Inverter section, be careful if you find one shorted. When they short the almost always fail the base driver circuit as well. This means that if you just charge the transistor again as soon as you attempt to start.

Later PWM drives use IGBTs in the output section, and they are much less likely to fail the driver sections.

Nervous about all this? Consider that the cost of VFDs has dropped dramatically within the past few years, along with an increase in features and reliability (just like computers). So, buy a spare and sent the broken one to us! We make a living repairing them for you! By: Eddie Mayfield