Learn how cables affect drive performance and get tips to help ensure your variable frequency drive (VFD) system operates as expected …And not interfere with hearing aids, cochlear implants, and audio frequency induction (“hearing”) loop systems.
By General Cable, with “Editor’s Notes” pertaining to hearing technolgy by Dan Schwartz
▬▬► Editor’s Note: This article explores a major source of radio frequency and electromagnetic interference (RFI/EMI) that affects hearing aid T-coils when used with baseband induction “hearing” loops; and also hearing aids and CI’s that use weak 10.6 mHz signals for inter-ear communication & coordination, and also 10.7 mHz for implant-to processor telemetry — More on this in the Bootnotes at the end of this article… ~DLS
Variable-frequency drive (VFD) cables emerged in the 1990s to help minimize operational issues with VFD systems. With the incorporation of Insulated Gate Bipolar Transistors (IGBTs) into inverters, switching speeds have nearly doubled from earlier Silicon Controlled Rectifier (SCR) and Gate Turn-off Thyristor (GTO) technology drives. This causes higher-frequency output signals from the inverters, which traditional power cables can’t always adequately handle. This can lead to system problems.
In some installations, VFD cables help solve system problems that range from motor issues and premature cable failure to control issues. In other installations, standard power cables work just fine. Why the difference?
VFD systems are complex, involving many components and variables such as drive types and specifications, filter or reactor installation, cable lengths, and proximity to other electronics. Additional variables include the type of cable glands, terminations, high-frequency bonding straps and installation variables.
Let’s take a closer look at how VFD cables can help minimize issues in these complex systems.
Issues Related to Other Electronic Systems
Noise radiated from VFD systems can negatively affect instrumentation, radios, alarms and other equipment. This effect is a function of the equipment proximity to the drive system, and the switching speed of the drive, cable layout, cable length, cable terminations and the nature of the VFD itself.
Installation practices often require inverter-to-motor power cables to be situated close to lower-voltage data, communication and control cables. With more control and automation systems installed in facilities today, these two types of cables can come in close proximity to each other at multiple points along the cable path.
Prior to IGBT technology, this wasn’t a problem. However, upgrading an older VFD installation to a newer IGBT technology requires evaluating the existing cable system, even if the older VFD system operated as expected.
Older VFDs using SCR or GTO technology have rise times that measure in microseconds. Newer VFDs using IGBT technology have rise times that can be in the tens of nanoseconds. While faster switching speeds reduce power losses in the drive, they also produce stronger electromagnetic (EM) fields around the cable. These stronger fields cause induced voltages/currents in nearby cables and other electrical systems.
In cables, this can result in dangerous currents and voltages and adversely affect data and control signals. In electrical systems, problems can range from intermittent operational issues to component failure and damage. Some of these electrical systems could even include a nearby inverter in another VFD.
To better control the electromagnetic interference (EMI) from the cable, VFD cables are designed with a continuous overall shield that provides low transfer impedance at high frequency. The cable terminations also should be designed and installed to provide a low to ground, low impedance at high frequency and a 360-degree connection to the cable’s overall shield.
System Issues Related to Cables
Newer VFD systems can put more stress on a cable’s insulation and the motor than was seen in traditional 60-Hz systems. This can lead to premature cable failure. Let’s compare some voltages.
A traditional 60-Hz sine wave has a zero-to-peak voltage V0-P of
V0-P = √2 * VRMS
…Where VRMS is the root mean square (RMS) voltage of the system.
In a traditional 3φ 480 VRMS 60 Hz system, the peak voltage sent to the motor is:
480 VRMS * √2 = 679 V0-P
In contrast, VFD inverters do not produce sinusoidal output voltage but instead generate a series of pulses. The peak voltage of these pulses at the VFD inverter is equal to the VFD’s DC bus voltage (Vdc), which is produced by a three-phase rectifier and expressed as:
…Where VRMS is the RMS voltage of the supply to the VFD. This means that in a 480 VRMS drive system, the peak voltage output from the drive is the same as the DC voltage, which is:
or 648 V0-P, which is 31 V less than the traditional 60 Hz system.
▬▬► Editor’s Note: Those of you who have worked with three phase full-wave bridge rectifier arrays will recognize that the output has a 5% ripple, which corresponds to the (31 volt/648 volt) lower voltage when the DC bus uses choke (inductor) input filtering, as the inductor smooths out the voltage to the lower 648 VDC value. However, whenever capacitor input filtering is used, the output voltage increases to the zero-to-peak value of 480√2 = 679 VDC. More on this in the Bootnotes… ~DLS
How can lower voltage cause more stress? It doesn’t. The stress is caused by the higher frequencies creating reflected waves in the cable. Reflected waves occur because of a mismatch in the high surge impedance Zo of the motor and the low surge impedance Zo of the cable. The greater the mismatch, the closer the wave reflection is in amplitude to the original source waveform. The cable sees the sum of these two waveforms, which can approach twice the amplitude of the source wave. This is close to what we experience in cables connecting today’s inverters and motors.
Let’s go back to our 480 VRMS drive system and assume a worst-case scenario of the reflected wave being equal to the source wave. With no reflected waves, the cable will see a zero-to-peak voltage V0-P 648 volts. If a reflected wave is allowed to develop, that same cable now is seeing a zero-to-peak peak voltage of 1,296 volts, which is much higher than the peak voltage of 679 V seen in traditional 480 volt 60 Hz power cable systems. If the installed cable is rated for 600 volts (850 volts zero-to-peak with a 60 Hz waveform), the peak voltage of the reflected wave is 448 volts more than the cable was designed to handle. This higher voltage, along with the fast rise times of the VFD inverter pulses, adds significant stress on the cable.
▬▬► Editor’s Note: Notice that the original author has been expressing the peak voltages as zero-to-peak or ground-to-peak values; however in fact in conduit as well as in cables where the phase conductors are touching each other, the peak-to-peak voltage values — which are twice the ground-to-peak values — must be accounted for, as this will determine the capacitive shunt currents, which appear as a short circuit to the output terminals of the VFD, and monotonically increase with the frequency of each of the harmonics. More in the bootnotes… ~DLS
Fortunately, using shorter cable runs can prevent reflected waves from ever occurring. To estimate the maximum cable length in feet, multiply the VFD pulse rise time (in milliseconds) by 246. Not all installations can be designed with cable lengths short enough to eliminate the chance of reflected waves. That’s when a VFD cable designed to handle the additional voltage stress should be considered.
System Issues Related to Motors
Premature motor failure in modern VFD systems often is because of bearing fluting. One contributing factor to bearing fluting is high motor frame voltage with respect to the inverter frame. If the inductance between the motor and the VFD is large enough, the reactance at high frequencies can support voltage drops of more than 100 volts between the motor and inverter frames. This voltage will cause a current to flow.
If the motor shaft is connected via a metallic coupling to the gearbox or other machinery that is solidly grounded and near the same ground potential as the inverter frame, shaft currents likely will flow due to the shaft’s “better” path to ground. Sometimes referred to as “shaft-grounding current,” this current may flow through the motor or machine bearings, causing bearing fluting.
Why would there be high voltage on the motor frame? Again, this is because of the high-frequency components generated by the drive. In a slower analog 60 Hz system, the net current flow of the three conductors at any point in time is 0. This isn’t the case in a VFD system. Because the waveform of each phase is the approximation of a sine wave (made up of the aforementioned pulses), there’s a net current flow (i.e., common mode current). This current, which originates at the inverter and must return to the inverter, must be minimized to keep the voltage between the motor and inverter frames as low as possible.
▬▬► Editor’s Note: Elecromagnetic emissions from these common mode currents flowing in the ground conductor are what cause interference to hearing aid and “hearing loop” systems… ~DLS
Two methods of minimizing this [common mode] current include:
- Use of a shielded cable that has low transfer impedance at high frequency to reduce the impedance between the inverter frame and the motor frame.
- Use of a symmetrically-designed cable with three grounds in the interstices of the power conductors to reduce the total induced current/voltage in cable grounds.
The best VFD cables are constructed with both of these methods to provide the best assistance in combating current flowing through bearings.
Tips for Improving Performance
The following three tips for cabling can help reduce or eliminate several performance issues in a VFD system:
- Shielded cables to address EMI issues in other electronic systems and reduce bearing currents.
- Cables with adequate insulation thickness to address standing waves.
- Cables with a symmetrical design to address induced currents related to bearing failure.
While it’s not easy to determine which installations require VFD cables, specifying cables that include all three of these cable attributes can help mitigate the issues discussed in this article and help ensure that your VFD system will operate as expected. These three VFD cable attributes are found in quality VFD cables offered by most major manufacturers.
This article is reprinted with permission of General Cable.
©2014 General Cable Technologies Corporation, all rights reserved.
Bootnotes from The Hearing Blog:
• As we stated in the lede, this article explores a major source of radio frequency and electromagnetic interference (RFI/EMI) that affects hearing aid T-coils when used with baseband induction “hearing” loops; and also hearing aids that use weak 10.6 mHz signals for inter-ear communication & coordination. This was originally conveyed to this author in mid-2011 by Widex, as their Clear (and subsequent) platforms have a feature that conveniently announces “Partner Check” into the ears of the wearer whenever inter-ear communications is lost. This is manifest when riding on inverter drive electric golf carts, as well as when the wearer is near inverter drive washing machines, and also neon signs. In addition to Widex, other manufacturers that use 10.6 mHz for inter-ear communication & coordination include Phonak (and Unitron), Siemens (and Rexton & Miracle Ear), and Oticon (and Bernafon). However, Starkey (and Audibel & NuEar) and GN ReSound (and Beltone) use UHF for inter-ear communications, which appears to be immune to this form of EMI/RFI.
• In addition, Advanced Bionics uses a weak 10.7 mHz reverse signal in their HiRes 90k system for implant-to-processor telemetry, which is also subject to this type of VFD and inverter-generated RFI/EMI — And may go a long way towards unraveling the cause of previously unexplained shutdowns due to loss-of-lock.
• When a DC motor drive is used, or if a VFD needs a lower DC bus voltage than the across-the-line diodes in a conventional 3 phase full wave rectifier will deliver, the diodes are replaced with SCR’s (a/k/a thyristors), which are fired at an arbitrary phase angle between 0° and 180°. Note the very sharp rise time in the tens of microseconds (which depends on the turn-on time ∂v/∂t of the SCR), which generate harmonics in the audio frequency range and beyond, which causes tremendous interference when hearing aid telephone coils are switched on for telephone or (especially) audio frequency induction (or “hearing”) loop reception:
• Here’s a neat trick we discovered in the All About Circuits article that is worth mentioning: It is indeed possible to obtain more pulses than twice the number of phases in a rectifier circuit in polyphase circuits. Through the creative use of transformer secondary windings, two sets of six full-wave rectifiers may be paralleled in such a way that more than six pulses of DC are produced for three phases of AC. This circuit leverages the 30° phase shift that is introduced from primary to secondary of a three-phase transformer when the winding configurations are not of the same type, i.e. a transformer connected either Y-Δ or Δ-Y will exhibit this 30° phase shift, while a transformer connected Y-Y or Δ-Δ will not. This phenomenon may be exploited by having one transformer secondary connected Y-Y (or Δ-Δ) feed a bridge rectifier array, and have another transformer secondary connected Y-Δ (or Δ-Y) feed a second bridge rectifier assembly, then parallel the DC outputs of both rectifiers to sum the outputs. Since the ripple voltage waveforms of the two rectifiers’ outputs are phase-shifted 30° from one another, their superposition results in less ripple than either rectifier output considered separately, and also yielding 12 pulses per 360° instead of just six: