LVDTs
In an earlier article in this series, we looked at proximity probes and discussed the challenges of measuring distance and displacement electronically. Now it’s time to look at one of the other key transducers used to tackle this problem.
Very often we are presented with a case where one object moves in relation to another. This can be a hydraulic cylinder or other kind of piston. It could be the position of two parts of an arm for a robot or earth mover. Any situation in which you need to know the distance between two objects and can attach an item between those two points, in theory, you can use an LVDT. There are some practical limitations; most LVDTs, even those described as long stroke, usually have ranges of less than 0.5m, so it is important that the range of distance between two objects be well-defined so that we know the distance won’t exceed the maximum stroke of the LVDT.
Now, with these limitations, you might think that the LVDT is somewhat of a niche device, but actually, even with its limitations, it’s an option in a wide array of instrumentation challenges that we encounter. Machines where we need to know the distance between two points where they move more than 0.5m may seem common, but as those components are typically connected to the machine elsewhere, we can usually measure displacements elsewhere on the device and then use geometry to calculate the desired metric.
But what actually is an LVDT? Whilst they are nearly universally called LVDTs, what that stands for is not actually agreed upon. This is why I bucked convention and didn’t define it until now. From reputable sources, you’ll find LVDT defined as Linear Variable Displacement Transducer, Linear Variable Differential Transformer, displacement transformer, differential transducer… and so on.
I prefer linear variable differential transformer because it actually tells you exactly how they work. Displacement transducer, on the other hand, explains what it does, which is also very useful depending on the context. So why does Linear Variable Differential Transformer explain how it works?
Let’s have a brief look at transformers. You may remember from physics lessons that a transformer is a device that we often use to step up or step down the voltage of AC current. It consists of an A coil with a certain number of windings, typically around an iron core, and a B coil also wrapped around the same iron core. An alternating current in coil A induces an alternating magnetic field in the core, which in turn induces an alternating current in the B coil. Simple, right?
Now, what if that iron core could move? If the position is variable. If we move it out of the B coil, the coupling between the A coil and the B coil will be weaker; if we move it into it, it grows stronger, providing the core doesn’t also exit the A coil. And there we have a rudimentary LVDT! Now the slightly less intuitive part is that the vast majority of LVDTs have two B coils (I would say all, but that’s just an invitation to be corrected by some obscure example). These two B coils are arranged in series with each other, but one is inverted compared to the other such that it generates an inverted AC current. These are arranged on either side of a central A coil. When the core is central, the inverted current in one side of the B coil is balanced by the current induced in the other side of the B coil, such that it reads zero.
As the core moves one way, the influence of the coil on that side becomes more dominant as the differential is increased. This is why they are called variable differential transformers. This increases the sensitivity of the arrangement to movements of the core, producing a much more effective instrument for comparatively little addition in physical complexity (even if conceptually it is a step up).
The real complexity with LVDTs, and the reason we bother to pay brilliant manufacturers such as Honeywell and Omega the big bucks, is producing the AC current and reading it. Like most transducers, LVDTs are typically powered by a voltage input or as part of a 4-20mA current loop. Read our article from a few weeks back to find out more. The LVDT not only needs to transform the DC power into an AC signal but do it with incredible precision. Small variations in the current energising the A coil will change the response in the B coil, which could read as movement when there is, in fact, none. This means that LVDTs include a high precision oscillator to produce such a signal.
On the B coil side, a demodulator converts the AC signal back into a DC signal, either milliamps or volts. AC signals are also easier to amplify than DC signals, at least without introducing drift or noise, so if this is required, this step is done first. If this is done correctly, LVDTs can produce readings that are linear with error measures below 0.1% of their range. For small ranges in the 10s of mm, error margins an order of magnitude lower are readily achievable, getting into micrometre resolution. This allows LVDTs to be used to measure tolerance and surface finishes for machined parts, a very common use for this type of sensor.
The conversion of the AC signal into DC is either built into the LVDT or built into a separate module. Flintmore has the capability to achieve either depending on need. Quite often, manufacturers will supply LVDTs and a combined oscillator demodulator module separately, which gives the instrumentation engineer more flexibility in terms of installation. It also gives the potential of using a single oscillator/demodulator module to drive several LVDTs and other sensors that require the same processing, which minimises duplication of parts and can bring down costs.
In my personal experience, LVDTs are another transducer that falls into the category of ‘they just work’, unlike other transducers that can be a bit more ‘sensitive’ (I’m looking at you strain gauges). LVDTs have several excellent qualities: they give absolute rather than relative position; the electromagnetic coupling for a certain displacement should always be the same, so they can be turned on and off and when re-initialised still read the same. They exhibit little to no drift or hysteresis. They also do well over time; whilst the core does move, it doesn’t have any contact with the coils, the coupling being purely electromagnetic. Any contact is purely through mounting bearings that hold the core steady as it moves. Whilst, like any moving part, the bearings will eventually fail, they have an extremely long life, meaning that LVDTs are very reliable with a low failure rate over time.
Overall, LVDTs, alongside proximity probes and linear encoders, are a brilliant tool in the instrumentation and test engineer’s arsenal for characterising linear displacement. Do you have a complex task involving the measurement and characterisation of displacement? Flintmore can provide the expertise, equipment, and hands-on engineers on the ground to get your job done. Get in contact today to talk through your challenge.
Further Reading
LVDT’s are widely used and as such have a wide array of excellent articles.
Honeywell’s article is very good with additional information on oscillators and demodulators that do the hard work of turning the lvdt from a concept to a functional and rugged sensor. They use Linear Variable Displacement Transducer:
One of my go to sources for transducer information when my memory fails is Omega, they use the name Linear Variable Differential Transformer:
https://www.omega.co.uk/technical-learning/linear-variable-displacement-transducers.html
And, as is often the case, the wikipedia article on LVDT’s is factually accurate (at the time of writing) with good diagrams and explanations, wikipedia lists all the definitions:
https://en.wikipedia.org/wiki/Linear_variable_differential_transformer
