Custom –v- standard sensors
18 January 2011
Why do some engineers choose custom sensors and others standard ‘off-the-shelf’ versions? Often the right choice is not straightforward. Mark Howard from Zettlex examines the pros & cons of custom versus standard position sensors and explains why new technology is changing the rules.
Why would an engineer choose anything other than a standard ‘off-the shelf’ sensor? There is a huge selection; no shortage of manufacturers, and the Internet provides a fast and easy method of comparing price, performance and availability. Despite this, there is still demand for bespoke, customised sensors across all sectors.
The first reason to choose a custom sensor is that there is not a standard ‘off-the-shelf’ sensor that meets all, or enough, of your requirements. The second, less obvious, reason is that standard off-the-shelf sensors could actually be more expensive than a bespoke or customised sensor.
We will use the term ‘bespoke’ to refer to a sensor that has been designed and engineered to meet a specific customer’s requirements. The term ‘customised’ refers to a sensor, which is based on a standard product but whose software or hardware has been modified so that it meets a specific customer’s requirements. For example, a customised sensor might be a standard unit but fitted with a mil-spec connector, high temperature electronics or a specific communications protocol. A standard or commercial off-the-shelf sensor (COTS) sensor refers to a sensor whose specification or build standard remains unchanged.
Most sensors – by both volume and value - are COTS sensors. Typically, sensor manufacturers group together a similar set of market requirements then develop and make a range of products to suit that group. If you are buying a small number of fairly run-of-the-mill, low to medium specification sensors then such standard ranges are ideal. Although you will pay a premium price for low volumes, the economics are still in your favour since any further effort (and hence cost) you might expend in searching and negotiating for the very best deal may well incur more cost than you save.
At the other end of the spectrum, we can consider higher volume, higher specification sensors – for example, those sensors found in robotics, CNC machinery, military or aerospace equipment. Even if you can find a standard COTS sensor that fits a demanding specification, it is likely that such a sensor will have a mechanical form, electrical interface, functionality etc. that has been designed for a group of applications rather than your specific one. As such, you will be paying for hardware and functionality that you may not need. Similarly, you may have to modify the design of your host equipment to accommodate a standard unit. Such changes might include a power supply specifically for the sensor, special cabling, cable routing, different connectors, couplings or mechanical mounts. Your overall unit costs will increase.
In higher volume, higher specification areas, bespoke or customised sensors are more likely to be a good choice. In these cases, the initial costs that the sensor manufacturer will charge to engineer a bespoke or customised solution can provide a rapid payback. You are likely to save money because you will get (and only pay for) exactly what you need, nothing more, and you will minimise the costs associated with the sensor’s mechanical and electrical interface because you can specify exactly what best suits the host equipment.
The decision in favour of a bespoke or customised sensor therefore depends on the non-recurring engineering/tooling cost versus the unit cost savings. For traditional sensor technologies, such costs may include significant mechanical design, tooling and re-engineering costs. These high costs for engineering and tooling can be prohibitive for many, mainstream applications and this means that, in many instances, a standard sensor has to be ‘shoe-horned ‘ in to the host equipment and made to work as best it can.
New technology is changing the decision point between standard and custom sensors by reducing or eradicating much of the engineering and tooling costs needed for a customised or bespoke sensor. New generation inductive devices are based on PCBs and so a custom or tailored device can be engineered by simply relaying out a new PCB to match the specific mechanical and electrical requirements. The fundamental physics behind these new generation inductive sensors is similar to those of resolvers and linear transformers and it is this fundamental physics that enables measurement stability even in harsh environments.
Similarly, these new generation inductive sensors do not need precision alignment or installation for precision measurements. This means that the mechanical components required to seal, protect and orient the sensor components are simply no longer needed. Instead, the main sensor parts can be mounted directly to the host machinery.
The net effect is that the costs required to engineer a customised or bespoke sensor solution are massively reduced. The sensor unit costs are also reduced since there is no need for a sensor housing, seals, bearings or couplings. The sensor is simply a PCB assembly mounted and housed within the host equipment.
The sensor’s circuit boards can be encapsulated or conformally coated to provide protection against even the harshest of environments; sensors can be powered from 3.3VDC - 240VAC; any connector can be used and the mechanical mounting points chosen to suit the host’s own mechanical parts. Shapes include rotary, linear, curvi-linear, 2D & 3D and measurement ranges span from 0.1mm to 10m.
As an example, UK-based Flow-Mon makes flowmeters. Christian Freeman, Flow-Mon’s commercial said: “Originally, we used a standard potentiometer to produce a 4…20mA signal of flow. The solution worked well but needed a lot of labour to set up. Then we found that we could swap to a next-generation inductive sensor that gave us better accuracy and radically simplified installation. The move to a non-contact sensor was also perceived by our customers as a more reliable, long-life solution. We specified the new sensor so that fitting just needed two screws; we could program the sensor with a PC and the moving part of the sensor also acted as a pointer to give visual indication of flow against a scale. The unit cost of the new generation inductive sensor was about the same as the potentiometer solution but the big savings on labour and simplification of the production process meant a rapid payback on the engineering cost. Our customers love the solution and now we also use an ATEX variant of the sensor.”
Most Viewed Articles...