Key points
When I’m not delivering Certified PROFIBUS and PROFINET Training courses, I spend my time in a wide range of industries auditing and fault-finding on these industrial automation networks.
As you might expect, I see a lot of Variable Speed Drives on these networks and over the years I have seen many occasions where they, or more correctly the way they have been installed, have been the cause of intermittent outages on the networks, or at least put the network at risk of succumbing to them.
So, what is special about Variable Speed Drives that would cause network outages?
There are several reasons, but one of the most significant is the switched-based nature of the output section of the drive, i.e. that connected to the motor.
Whilst all drives are supplied from a 50 Hz ac supply, internally the voltage is rectified and then applied to the motor using a power-switching technique called Pulse Width Modulation (PWM).
If you were to use a high voltage differential probe between the phase conductors of the motor and connect the probe to an oscilloscope what you would see would be far from a sine wave and instead is just a series of very fast on/off pulses, with the amplitude of the pulses being around 600 V.
When this is applied to an induction motor, the inductance of its windings ‘integrates’ the pulses, resulting in the current drawn approximating to a sine wave whose frequency is changed by varying the pulse waveform of the applied voltage.
It is the switching, and specifically the very fast voltage transitions from off to on and on to off, that can lead to emissions of several MHz. If the installation does not take this into account, these emissions can travel around the installation and cause the kind of problems I see quite often.
When it's too late to fix a Variable Speed Drive installation
The problem is that, when I get involved, it is a little late in the day to suggest changes as these can then be a very expensive exercise to fix. This will not be new to some of you but it is clearly something that many designers, installers and end users, are not aware of, so what is going wrong?
Several years ago, I decided to look into this from a standpoint of an industrial electrician, what do the IET Regulations, BS7671:2018, have to say on this?
Section 444 of BS7671:2018 is entitled “Measures Against Electromagnetic Disturbances” with section 444.4.1 identifying typical sources of disturbances which includes Variable Speed Drives, so there’s a good start. This document is primarily concerned with bonding and cable separation.
Most people that I have spoken to look at the bonding from a safety-point of view, not understanding its importance in mitigating against potential EMC problems.
A bonding infrastructure that includes the use of green/yellow earth wire will be fine from a safety point of view but poor from an EMC point of view because of the inductance of a wire and the skin effect, both of which will impede current flow as the frequency increases.
Most EMC issues I have come across would have been prevented if the installers had at least followed these two guidelines. Section 444 takes up only 8 sides of A4 paper and can only be considered as a starting point in mitigating against EMC, this is why it makes the following statement: –
“The requirements of the following standards shall be applied where appropriate” and then goes on to list the following: –
BS 6701: Telecommunications equipment and telecommunications cabling. Specification for installation, operation and maintenance.
BS 50310: Application of equipotential bonding and earthing in buildings with information technology equipment.
BS EN 50174 Series – Information Technology. Cabling Installation.
BS IEC 61000-5-2: Electromagnetic Compatibility (EMC). Installation and mitigation guidelines. Earthing and cabling.
In other words, if your installation includes one or more of the listed sources of emissions then the designers, installers and end-users need to get hold of these standards, read them and, where appropriate, implement them.
It’s interesting to note that BS IEC 61000-5-2 has been in place since 1997, that’s over 20 years ago and yet very few installations I have seen follow it and equally more disturbing is that very few people have heard of it.
Reading the standards is one thing, understanding them and how to implement them is another thing all together. The problem seems to be that many installers and designers still have their 50/60Hz hats on and whilst this remains important they need to consider the realities that much of the equipment the use / install will have the potential of releasing MHz emissions into the control system, and beyond, that will have the potential of causing these annoying outages that no one seems to be able to get to the bottom of. The way forward has to be better education and awareness.
The main concepts that Variable Speed Drive designers and installers need to understand include:
- The real difference between resistance and impedance. The meshed bonded network needs to present a low impedance to currents at the “frequency of concern”. If we know the frequency of likely emissions from the equipment being used then we can work out the wavelength, in metres, and then relate this to the mesh dimensions.
- The concept of differential and common mode currents. All electrically-biased people inherently understand differential mode currents, i.e. though one wire and back through another. As long as the two conductors are close together then the magnetic fields due to current in each wire will be effectively cancelled.
Common mode is very different and is where current flows in both conductors in the same direction and hence there is potentially no cancellation of magnetic fields. A poorly installed Variable Speed Drive will result in uncontrolled flow of these common mode currents and the EMC-issues they cause.
The drive manufacturers are well aware of this and go to great lengths in their installation guidelines on how to avoid them. The problem is that not everybody reads them, or if they do, they don’t act upon them. - Cable Shielding – Depending on what the cable is carrying, the shield will either protect the internal signals from outside influences, or protect the outside world from cable emissions. PROFIBUS and PROFINET come in to the former, Variable Speed Drive motor cables come into the latter.
People often talk about “Earthing” a shield and then go on to discuss the issues of ground loops. It has long been the case that the only way a cable shield can protect against both electrostatic and electromagnetic coupling is for the shield to be bonded at both ends to a common reference.
The fact that this common reference is “earthed” is for safety reasons and nothing whatsoever to do with EMC mitigation. The standards around what to do with the shields in a hazardous area always create discussion as they stipulate single-ended termination only. In this case the designer has to follow the latter for safety reasons but needs to be aware that the consequence of following them will be that the shield can no longer protect against any magnetic field coupling. - Parallel Earth Conductors (PECs) – These are covered extensively in BS IEC 61000-5-2 and control the flow of common mode currents to mitigate against potential EMC issues. Most people think that a cable tray is for mechanical support whereas designers are encouraged to consider using these as PECs as well.
- The real effects of Pulse Width Modulation – Variable Speed Drives have used this technique for many years. From an EMC point of view, the problems occur on the rising and falling edges of the pulses. As most of the heat loss in a drive occurs at these transitions, there would be a real operational / economic benefit in reducing the duration of the transitions, i.e. increasing the rate of change.
This has been continually happening as new switching technologies became available, but an unfortunate consequence of this is that the MHz emissions have been increasing.
There is no end in sight to this ‘progress’, with VSDs based on Silicon Carbide (SiC) and Gallium Nitride (GaN) technologies being about to take over from Silicon PowerFETs and IGBTs Insulated Gate Bipolar Transistor), achieving at least a ten-times increase in switching speeds, with a corresponding ten-fold increase in both amplitude and frequency of potentially interfering MHz emissions.
A real understanding of all of the above, as well as reading, and both fully and correctly implementing, the installation guidelines of the Drive manufacturer will go a long way in avoiding these often costly and disruptive mistakes.
Having incorporated the techniques into the design and installation, you should always remember to perform an installation and operational qualification on your PROFIBUS and PROFINET networks as doing so can confirm that there are no underlying issues that may come back to haunt you in the future.
