What if a variable-frequency drive (VFD) on a critical system has a fault indication, the “error” LED on a DC power supply has turned on, or a conveyor belt has started behaving erratically — most notably when the conveyor speeds up? You need answers fast, but it’s unclear where the problem lies.
The first reaction for most electricians would be to reach for their digital multimeter (DMM) to start the troubleshooting process. However, that’s not always the right move. As this article will demonstrate, the use of a digital oscilloscope can help speed up the troubleshooting process for industrial control systems and be more effective in pinpointing the exact nature of the problem.\
Analyzing an undervoltage condition with a VFD
The smooth operation of many industrial processes requires that VFDs be kept in good working order. When a fault indicator comes on, it’s important to diagnose and fix the problem quickly.
For example, let’s say a VFD on a critical system has a fault indication. The error code on the device is “F4,” which means the drive has detected an “undervoltage” condition. At this point, the drive has shut down. But is the failure in the drive, the motor, or a distorted supply voltage? You can check the line input voltage with a DMM, but depending on the type of distortion, a voltage reading alone may not reveal whether a problem exists.
Your next step should be to check the VFD DC bus voltage. Is it directly proportional to the peak of the input line voltage? Check for any distortion or error in peak amplitude of the line voltage, which can cause an over- or undervoltage error.
Let’s say when you check the drive’s DC bus voltage, you find that it’s about 20% under the nominal value. That’s a problem, but do you replace the drive controller, the motor, or both? You really need more information before you can proceed.
An oscilloscope can give you a clearer picture. At the power input of the drive, connect the scope to phases L1, L2, or L3 and the ground lead to neutral, and take a look at the waveform. Let’s say you do this, and it reveals an almost flat-topped sine wave. What does this mean?
Flat-top waveforms are often caused by nonlinear loads attached to the same feeder circuit. This confirms the problem doesn’t lie with the VFD or the motor. The oscilloscope graphically shows not just amplitude, but also any distortion, disturbance, or noise that may be affecting the waveform, giving you the information you need to accurately diagnose and solve problems.
Analyzing Transient Reflections at the output of a VFD
VFDs control the speed of machinery and processes by varying the width of (i.e., “modulating”) the voltage pulses they send to motors. These voltage pulses, which are essentially square waves, have fast-rising edges that can cause transient voltage reflections in the cable between the VFD and the motor. Many factors affect the size of the reflected waves, including cable length, motor load, the surge impedance of the cable and motor, and the rise time, spacing, and magnitude of the drive pulses. Over time, these reflected wave voltages can place too much stress on cable and motor insulation, which can cause either to fail.
Making voltage measurements with a DMM can reveal excessive voltage at the motor terminals when transients are present. To check the voltage, connect the multimeter at the motor terminals between phases L1, L2, or L3. Although you can use a DMM to measure the peak voltage of a pulse width modulated (PWM) signal, the limited bandwidth of the typical DMM makes measuring pulses with short rise times problematic. Therefore, the readings may not provide the complete picture of the presence — or full amplitude — of any reflections or transients. Unfortunately, even the best DMM can’t show you what the waveforms actually look like.
Oscilloscopes display a signal as a line that moves across the screen from left to right. The higher the voltage, the higher the signal appears on the oscilloscope screen. Measurements made with an oscilloscope reveal not only the presence of excessive voltage at the motor terminals, but also give you a clear picture of size, location, and duration of any voltage transients.
To take a reading with an oscilloscope, connect the scope to the motor terminals between phases T1, T2, or T3, just as you did with the multimeter. An ideal PWM signal with normal peak voltage will look like the waveform in Fig. 1. If high-voltage transients are present from voltage reflections in the cable, then the waveform will look like it does in Fig. 2. These voltage reflections are often caused by improper cabling.
Analyzing a Rotary Encoder that has been Operating Erratically
A rotary encoder converts the position of a rotating shaft into a series of digital pulses. In this example, the pulses are monitored by a controller that uses them to control the speed and position of the conveyor system to which the encoder is attached (Fig. 3). The conveyor belt has started behaving erratically — most notably when the conveyor speeds up — but it’s not clear if the problem lies with the rotary encoder or the controller that operates the belt.
Measurements with a DMM can reveal the voltage, frequency, and duty cycle of the signal from the rotary encoder. So you break out your DMM and make these measurements. When you’re finished, you realize that all of the values appear to be normal. Because the controller’s programming hasn’t been modified and has worked well for years, chances are that’s not the problem. Looks like it’s time to break out the oscilloscope again.
An oscilloscope connected to the output of the rotary encoder can display the signal’s average voltage, peak voltage, and frequency, just like a DMM. However, as noted in the previous section, an oscilloscope can also display the voltage change over time. A good digital output from the encoder will look like the waveform shown in Fig. 4.
But what if the signal waveform is jagged and electrically “noisy” like the one shown in Fig. 5? This is a clear sign of trouble that can arise when there is inadequate shielding on a signal line. Upon inspection of the cabling, we find the cable shield of the controller wasn’t fastened down properly. This ends up being a simple problem to fix.
Continue to Page 2
Analyzing the Erratic Operation of a Mechanical Proximity Switch
Mechanical switch contacts deteriorate over time, because they are subject to mechanical wear. In addition, arcing can cause pitting of the contact surfaces. Usually, these problems arise slowly and gradually worsen before the switch completely fails.
Troubleshooting a mechanical switch that has failed completely (i.e., a “hard fault”) can often be done with just a visual inspection or with simple measurements made with a DMM. It’s the gradual or intermittent switch failures that usually cause real headaches.
In this example, a failing mechanical proximity switch is causing erratic operation in a conveyor system. After connecting a DMM to the output of the conveyor belt controller (which receives its input from the proximity switch), we see the “voltage peak max” and “voltage peak min” readings show the switch is turning on and off as expected. If you have a DMM with frequency measurement capability, you could also see the rate at which the switch is changing state. However, none of this information helps diagnose the real problem.
A modern digital oscilloscope can give you a lot of the same numerical information as a DMM, including the pulse voltage and frequency information noted above. It also allows you to see switch on/off state timing and time relationships between a source and output signal (i.e., the switch and the controller output).
After connecting the oscilloscope to the output of the controller, it reveals nothing out of the ordinary. The pulse waveforms are well-formed and free from electrical “noise” (Fig. 6). We then connect the oscilloscope so it captures the signals from the proximity switch on the first channel and the output pulse from the controller on the second channel (Fig. 7). When we examine the result, it’s immediately clear that something is wrong. The bottom trace (the output from the proximity switch) is not stable in relation to the top trace (the controller output).
“Zooming in” on the signal (Fig. 8) reveals that the output from the proximity switch (lower trace) isn’t changing from OFF to ON in a single transition (red circle). Instead, faulty switch contacts are bouncing on and off for about 5 msec before the output signal stabilizes. The controller is unable to read this jagged voltage correctly, so its output varies broadly (over the time range between the red bars in the top trace). That is what’s causing the erratic behavior.
Analyzing a Power Supply with an Intermittent Failure
The DC power supply unit (PSU) is one of the most critical components in any automation or process system. If a PSU suffers a “hard failure,” you simply replace it and move on. But what if the problem is intermittent, or the problem returns a short time after you replace the failed PSU with a new one?
In this example, let’s say the “error” LED on a DC power supply has turned on. Your troubleshooting task is to determine if the problem is the power supply, the input supply voltage, or a load change on the demand side of the unit.
Using a DMM, you measure the line input voltage, and see that it looks fine. Next, you check the DC output voltage. Again, everything looks good. At this point, you decide to swap out the PSU with a known good replacement unit and hope for the best. However, when you come back 2 hr later, you see that the fault indicator LED is lit again. What should you do now?
It’s time to check the PSU’s input and output waveforms with an oscilloscope. First, connect the oscilloscope to the power supply’s AC line input terminals and check the input waveform visually for any fluctuation, distortion, or dropouts. Let’s assume you see the AC line voltage is a perfect sine wave. Having established that the AC supply voltage is good, your next move would be to check the DC output voltage. It looks good as well. Note: Some oscilloscopes require a separation transformer or a differential probe to measure the line input voltage and the DC output voltage simultaneously.
Because no problems are immediately obvious, your next move is to check the PSU’s input and output waveforms over time. Let’s say you determine the line input voltage dropped to 71V after 14 hr, 23 min., and 15 sec, causing the “error” LED to turn on. This would tell you the fault is in the AC supply, not the PSU. But what if the AC line voltage input is OK over a significant period of time? Then you need to check the PSU’s DC output voltage. To measure the PSU DC output current and voltage over time:
- Place a current clamp around one of the DC supply conductors and connect the clamp to channel A of the oscilloscope.
- Then, connect channel B to the power supply’s DC output voltage.
Now you can plot both the PSU’s output current and voltage over time.
Let’s assume the load current exceeds the maximum rating of the power supply at 16 hr, 33 min., and 59 sec, which causes it to drop out (Fig. 9). Now it’s either time to look “downstream” of the supply to see what’s drawing too much current, or it’s time for you to put in a bigger power supply.
Hammond is a product manager with Fluke Corp. in Everett, Wash. He can be reached at Hilton.firstname.lastname@example.org.