What is System Integrity?

If you manufacture, operate or maintain critical electrical and electronic systems - whether as part of an aircraft, or a satellite ground station - then you need the confidence and assurance that these systems will continue to carry out their intended function correctly and safely throughout the entire duration of their intended usage and life-cycle. This is the concept of System Integrity. A more formal definition is that System Integrity is; The ability of a system to function correctly without being degraded or impaired by changes in its internal or external environments. A major indication that System Integrity has been degraded or compromised is manifested through intermittent faults and ‘No Fault Found’ occurrences What is meant by No Fault Found (NFF)?

 

A NFF is a reported fault for which the root cause cannot be found.

The NFF phenomenon is usually encountered as a symptom that has been reported by the operator of a platform, system or piece of equipment and can not be reproduced or replicated. It is a chain of events that starts, chronologically, with the end user experiencing a fault symptom. The chain of events might progress from a sortie abort to reporting of the fault to the relevant technicians, through to diagnosis and rectification activity. If the diagnosis is a success then the genuine root cause of the fault is isolated and rectified and the aircraft is signed off serviceable having been made fit-for-purpose and airworthy once more. However, if the root cause cannot be found – in other words, the activity has resulted in diagnostic failure– you are dealing with a NFF situation.

 

What is the impact of NFF?

NFFs become a distraction to all involved in the operating of the platform, in terms of the ability to perform and provide the desired output, the frustration to repair the system to be available, the expense of providing a supply chain to support shot-gun maintenance activities, and of course the end customer and their annoyance at the level of service provided by a unpredictable system. How many problems does NFF cause on aircraft? The precise extent of the problem is difficult to quantify because published aerospace statistics are far from definitive, but they do illustrate the approximate size of the problem: avionics generally constitute 75% of aircraft NFF occurrences and, in turn, avionics NFF rates are (conservatively) in the region of 30% or higher. What are the main causes of NFF? The main causes are: Intermittency Maintenance Knowledge/Data Test Coverage Maintenance Human Factors

 

What is Intermittency?

Intermittency can be described as a temporary deviation from normal operating conditions of a circuit or device. It manifests itself as intermittent faults as the system or platform degrades through 100% availability to a ‘hard fault’ (ie permanent) state. What causes Intermittency? Intermittency is caused by minute changes in a circuit’s characteristics such that they have repercussions on the operating parameters of the system in which they occur. Digital systems and, in particular, databus circuits are more susceptible to intermittent faults, with micro-changes causing drops in voltage which then lead to data bits being dropped from the data-stream and thus causing parity errors. These errors result in either the system locking out/shutting down/restarting, or data being transferred again and again until a hand-shake is satisfactory, with delays in data transfer. What is the different between the definitions of Serviceability and Availability? A system or platform can be considered ‘serviceable’ if it can be demonstrated successfully, at a specific point in time, that it will successfully fulfil its intended function. Availability, on the other hand, is the measure of the system or platform’s ability to remain serviceable and operational for the entire duration of its intended duty cycle. Therefore, Availability can be described as a system with full functionality and integrity, such that it remains fit-for-purpose over the intended duty cycle. Is NFF known as anything else? Yes, it’s also called: Cannot Duplicate (CND) Re-Test OK (RTOK) Unable to Reproduce Fault Unable to Reproduce Symptom

 

Why should I bother even looking for intermittent faults?

Generally speaking, intermittent faults are the basis of virtually all NFF occurrences, and are responsible for over 40% of all reported No Fault Founds faults in electrical and electronic circuitry. Other main causes are: Maintenance Knowledge/Data Test Coverage Maintenance Human Factors.

 

What is the impact of NFF?

NFFs become a distraction to all involved in the operating of the platform, in terms of the ability to perform and provide the desired output, the frustration to repair the system to be available, the expense of providing a supply chain to support shot-gun maintenance activities, and of course the end customer and their annoyance at the level of service provided by a unpredictable system. How many problems does NFF cause on aircraft? The precise extent of the problem is difficult to quantify because published aerospace statistics are far from definitive, but they do illustrate the approximate size of the problem: avionics generally constitute 75% of aircraft NFF occurrences and, in turn, avionics NFF rates are (conservatively) in the region of 30% or higher. What are the main causes of NFF? The main causes are: Intermittency Maintenance Knowledge/Data Test Coverage Maintenance Human Factors

 

What is the difference in testing capability between an IFD and Ncompass?

Couldn’t I use a highly accurate Digital Multi Meter to find the intermittent faults? Intermittent faults can be found using traditional test equipment to a limited extent, but this is only possible with conventional test equipment when the intermittent faults have degraded to the extent that they are closer to becoming hard faults. Additionally, conventional test equipment only tends to find intermittent faults on system circuits that are well understood and where faults have been found before after a considerable amount of time has been expended on finding them. More significantly, traditional test equipment is only used for point-to-point, single point-in-time testing, so their restricted ability to detect the micro-changes at the root cause of intermittency, and NFFs, is very limited. As test equipment is developed to become more accurate...

 

...does this increase the chance of detecting intermittent faults?

Intermittent faults occur randomly. Whilst it could be argued that there is no such thing as random, the context of ‘random’ in this instance is centred around the fact that intermittent faults occur at unpredictable points in time depending on mechanical, environmental and of course electrical stresses/conditions on the test point. Therefore, as with anything that is random, detection must be more aligned to probability than accuracy; the experience of taking a perfect photograph of a lightning strike with a highly accurate shutter speed setting would be deemed by chance not by accuracy! Accordingly, the art of finding intermittent faults is all about increasing the probability of detecting a fault and isolating its existence. By testing more than one point-to-point - ie by simultaneously testing multiple test points - and continuously over a period of time using a the IFD/Ncompass means that there is a 1x10^6 increased chance of detecting an intermittent fault. Taking the analogy one step further, improving the chance of taking a perfect photograph of a lightning strike would require a long shutter period and in multiple directions, ie a shift from highly accurate to one of continuous and simultaneously

 

What is a Neural-Network?

Traditionally, the term neural network had been used to refer to a network or circuit of biological neurons. The use of the term in relation to the IFD/Ncompass refers to artificial neural networks, which are composed of artificial neurons or nodes. Thus for the IFD/Ncompass, the Neural-Network allows multiple test point signals to be combined simultaneously, through a patented technology, and thus allows the rising signals to be detected and referenced; this unique capability allows the IFD/Ncompass to simultaneously monitor multiple test points at the same time, and as explained in another FAQ, thus increase the probability of detecting an intermittent fault.

 

Why use an analogue over a digital test signal??

As explained in another FAQ, detecting an intermittent fault is about increasing the probability of detection rather than the accuracy of detection, therefore by increasing the time domain (ie an analogue signal) of the test signal and monitoring the returns is more appropriate than a digital signal. Conversely, digital signals are synonymous with sampling and digital averaging, and both of these side effects decrease significantly the probability of detection. While the testing world shifts to accuracy, the IFD/Ncompass exploits the benefits of an analogue signal, combined with the patented technology of Neutral-Networks.

 

What is the difference in testing capability between an IFD and Ncompass?

Generally speaking, intermittent faults are the basis of virtually all NFF occurrences, and are responsible for over 40% of all reported No Fault Founds faults in electrical and electronic circuitry. Other main causes are: Maintenance Knowledge/Data Test Coverage Maintenance Human Factors.

 

What is the impact of NFF?

An IFD is a dedicated Intermittent Fault Detection tester, whereas the NCompass has more testing capabilities such as Continuity, Short, Trace and other point to point testing modes, plus Analyze and Z-Sweep.

 

As test equipment is developed to become more accurate...

...does this increase the chance of detecting intermittent faults? Intermittent faults occur randomly. Whilst it could be argued that there is no such thing as random, the context of ‘random’ in this instance is centred around the fact that intermittent faults occur at unpredictable points in time depending on mechanical, environmental and of course electrical stresses/conditions on the test point. Therefore, as with anything that is random, detection must be more aligned to probability than accuracy; the experience of taking a perfect photograph of a lightning strike with a highly accurate shutter speed setting would be deemed by chance not by accuracy! Accordingly, the art of finding intermittent faults is all about increasing the probability of detecting a fault and isolating its existence. By testing more than one point-to-point - ie by simultaneously testing multiple test points - and continuously over a period of time using a the IFD/Ncompass means that there is a 1x10^6 increased chance of detecting an intermittent fault. Taking the analogy one step further, improving the chance of taking a perfect photograph of a lightning strike would require a long shutter period and in multiple directions, ie a shift from highly accurate to one of continuous and simultaneously

 

What is a Neural-Network?

Traditionally, the term neural network had been used to refer to a network or circuit of biological neurons. The use of the term in relation to the IFD/Ncompass refers to artificial neural networks, which are composed of artificial neurons or nodes. Thus for the IFD/Ncompass, the Neural-Network allows multiple test point signals to be combined simultaneously, through a patented technology, and thus allows the rising signals to be detected and referenced; this unique capability allows the IFD/Ncompass to simultaneously monitor multiple test points at the same time, and as explained in another FAQ, thus increase the probability of detecting an intermittent fault.

 

Why use an analogue over a digital test signal?

As explained in another FAQ, detecting an intermittent fault is about increasing the probability of detection rather than the accuracy of detection, therefore by increasing the time domain (ie an analogue signal) of the test signal and monitoring the returns is more appropriate than a digital signal. Conversely, digital signals are synonymous with sampling and digital averaging, and both of these side effects decrease significantly the probability of detection. While the testing world shifts to accuracy, the IFD/Ncompass exploits the benefits of an analogue signal, combined with the patented technology of Neutral-Networks.

 

What is the difference between testing capabilities within the NCompass?

The revolutionary technology within the NCompass has been developed to detect and isolate intermittent faults in a circuit. However, there are other test modes that come with the NCompass that makes it more cost effective in allowing the user to perform other tests using the same interface cables and exploit the time needed to map the test points for a particular Unit Under Test (UUT). The other testing modes are: Point-to-Point including: Continuity and comparing theoretical or previously measured test results against each test point. Short testing against a predefined map of the test points. Tracing of characteristics between two test points. Measurement of a particular characteristic (not insulation testing) using an external tester on a BNC connector on the NCompass. Analyze allows all the test points to be characterised by using a impedance test and then allow a comparison of the test results with previous test results taken for the same UUT. Z-Sweep is similar to Analyze but is more dynamic in the testing regime and compares test results on in real time. Please get in contact with the Copernicus Team for more details on testing capabilities and how to exploit them.

 

If the IFD/Ncompass is so capable and sensitive...

...will it not highlight too many problems which will be too expensive or time-consuming to put right? In essence, a detected intermittent fault means that there is a fault, and therefore, this fault will be causing or may start to cause problems; so the short answer is ‘Yes’. However, in the short-term there are other options aside from repairing the fault immediately. Once these faults are detected and known, it is now all about managing what you do know and not about not knowing the extent of the don’t knows! With the IFD/Ncompass you can establish the severity of the intermittent faults and where the problems within the system are. Just because an intermittent fault is found, it does not mean it has to be repaired; it can now be managed. Knowing a system has intermittent faults, and in relation to the system’s criticality, other solutions or courses of action can be perused; this may include: Informing the operator that the intermittency actually exists and the symptoms that it causes so it does not become a surprise and thus increase stress levels and thus lead to compromising Flight Safety. Enhanced training for the operator so that they are familiar with the actions needed to manage the symptoms caused by intermittency. Manage the reporting of the symptoms and track the level of intermittency and thus No Fault Founds by system. This helps to alleviate maintenance time and resources by not unnecessarily searching for intermittent faults that have already been established as being there! This leads to reducing: Supply chain assets which normally increase to back fill spares used on specification to try and repair the fault. The stress of the maintenance technicians and the supporting staff in trying to reproduce an intermittent fault. Assets can be selected for specific operational tasked based on their Intermittency Log as you would currently for Deferred Defects and Limitations Log.

 

How will this provide a Return On Investment (ROI)?

Clearly all investment is based on the possible returns that can be achieved and thus a business case can be developed. Understanding the problem and the effects on operational capability, customer satisfaction, and maintenance and supply chain costs is the starting point. The cost of the Ncompass, deign and build of the interface plus the technical assistance (if required) becomes a small investment when looking at the returns that can be yielded by improving the availability of systems by detecting and isolating intermittent faults and solving these failures at their root cause. On one Line Replaceable Unit, its MTBF was increased by over 100% saving $1.1m/year in just maintenance costs alone and over $28m in asset costs.

 

How do I connect the IFD/Ncompass to my component or system for testing?

For each Unit Under Test (UUT) a testing interface will have to be made. Each test point on the UUT will be determined as part of the interface design process and then connected via a series of Sub-D 50-pin cable loom to the UUT. Where practical the Interface Adaptor (IAC) will include the UUT mating connectors to ensure where possible a fully representative connection to the UUT. Once the connections have been made physically to the IAC, the test points for the UUT will then be mapped using a User-friendly software tool which will create the map and graphic files needed for the IFD/Ncompass. Once the IAC has been designed and manufactured, it can also be used for a number of other traditional testing methods which can be performed by the Ncompass.

 

Will I have to make interface cables for every test point on my system?

No, the first step into eliminating intermittent faults and thus No Fault Founds, is to understand where the majority of the problems exist and isolate the problem down to a system or sub-system. Using, for instance, Pareto Analysis, you would then pick those systems causing the majority of the NFFs and concentrate on these. Once these problem areas have been tackled and the arising reduced, the emphasis can then shift on to other systems or sub-system which will now have proportional higher arising. This would process will then cease once a level of NFF is deemed acceptable in terms of risk to safety, cost and/or business output.

 

How much do the interfaces cost to produce?

Interfaces can be built by CTL and/or by the client. CTL provides a training course for design building interfaces and this can reduce the costs to the client. Depending on the application of the interface and, the cost of the matting connectors for the Unit Under Test, the main cost of the interface is at the build stage. For example, a 256 test point cable harness, this can take approx 2-man days plus approx £250 for the cables and Ncompass connectors.

 

How does the test system know what it is connected to and is it difficult to setup?

There are a number of ways that this can be done. Firstly, if the wiring diagram is available then the Ncompass MAP file can be constructed using the NODES software tool. Secondly, a Automap facility is available within the Ncompass software which will automatically construct the MAP file. This can be then imported into NODES and then, if required, it can be compared to the theoretical MAP file. Both these processes are easily carried out and seamless to the user.

 

Can I use some of the existing interfaces I already have made up?

Yes an interface could easily be made to allow the Ncompass to connect to other interfaces already made for other testers such as DIT-MCO.

 

There are some really good pieces of test equipment already on the market...

...why don’t I just use this? The Ncompass is the only piece of test equipment in the World that can detect and isolate intermittent fault and has not be designed to replace testers that test for isolation, accuracy in ohmic/capacitance etc, it merely complements these testers with its revolutionary technology to detect intermittency. In addition, to intermittency detection, the Ncompass can provide the user with continuity, shorts and scope testing; plus it has the benefits to connect to other passive testers such as TDR, Oscilloscope or high end DMM.

 

25. Is there any complex set-up...

...that would always require specialist advice to be called upon? No, the Ncompass is very simple to program and setup. A training course is available from Technician to Board room level and is tailored depending on the client’s abilities and current knowledge/experience. Training course range from data acquisition & analysis to design and development of interfaces plus the programming of the Ncompass, and then on to the reading/understanding the Ncompass results.

 

Is there a lot of programming required...

...to allow the Ncompass to be used effectively? No, there is absolutely no programming needed. The user needs to understand a few concepts that are new to the testing regime and then be able to enter test point data in the software in order for it to recognise what it is connected to; there are graphical representations available to help the user visually reference the test points on the unit under test.

 

Can you test databuses?

Yes. More importantly, the Ncompass is only tester that is able to find early stages of intermittency on a databus. There are methods of testing that are designed to find intermittency on a databus by logging failures to transmit/receive, handshake and parity errors. However, these are limited methods given they are point in time, and in some cases, sample rates are not high enough to detect small glitches in the databus line; this is especially so when parts of the databuses are tested in isolation and therefore when they are not under load. The Ncompass is designed to test all the test points continuously and simultaneously and as such has a much higher probability of chance of detecting intermittent fault.

 

Does it matter if the device is analogue or digital?

No. Clearly it is important to understand the circuit dynamics and where the test points need to be connected to the Ncompass to ensure maximum test coverage. However, from experience it is more likely that intermittency is due to the mechanical interfaces ie the matting parts of the Unit Under Test such as the cables, chassis and connectors. Therefore, test points should be chosen to ensure these areas are exposed to the Ncompass testing regime. CTL can advise clients on any aspect of this FAQ.

 

How do you test cable runs – do you have to connect to both ends?

Yes, the Ncompass needs to connect to both ends of the cable run. However, this can be achieved by having a fly-lead to connect to the other end of the cable run, or a shorting link at the other end of the cable harness could be used to provide a complete circuit. The latter solution is a better method for longer cable runs.

 

Does it test insulation?

No. The Ncompass is not designed to test for insulation. However, it can find intermittent faults between test points or ground which might be a result of insulation breakdown.

 

How much does it cost?

CTL provides both a service solution or a sales solution for their clients. The range of technical requirements coupled a need to understand the problems facing the client are key in the tailored solution(s) CTL would provide. If you require a consultation with one of our team and a quotation, then please contact us.

 

Does the Ncompass need calibration or self-test?

A limited self test is needed to ensure that all the test points are still functioning but this is a very quick test to perform and it does not need to be sent away for this. As for calibration, there is no need to calibrate the Ncompass for Intermittency Testing or number of the other tests it can perform. Calibration is only needed on an annual basis if the Analyse and/or Z-Sweep tests are used to compare ‘Gold’ items or compare a Unit Under Test against itself using a previous test results; this has to be carried out by CTL or USC.

 

Can it test transformers?

Yes. As long as all the windings of the transformer can be connected to the Ncompass via an array of test points, the transformer can be tested for intermittency, continuity, shorts and/or Log Scope

 

Will the length of the cable/harness run ...

....affect the performance of the Ncompass in its ability to detect intermittent faults? For pure cable runs, with no in-built components, such as ballast resistors or protection diodes (although in the majority of cases this does not pose a problem either), the cable run length which can be tested effectively can exceed over 100m (330ft); this is more than the length of A380-800 or Boeing 747-8 and includes the return path for the test point ie 200m test length. The Intermittent Fault detection system is not looking at measuring the circuit, but it monitors all the circuits continuously and simultaneously for changes. It is these changes, which are caused by intermittent faults, that are then detected and isolated by the Ncompass; it is from this data that the Ncompass then presents to the User the circuit on which the intermittent fault occurred.

 

Why shouldn't I use a SWR or TDR for long cable runs?

The short answer is you should, but only once you have detected and isolated the intermittent fault with the Ncompass. To find an intermittent fault using a SWR or TDR would take too long and is probably impossible to detect depending on the stage of intermittent fault. Clearly by placing a TDR/SWR on a single cable for say 1 hour there may be a chance to detect an intermittent fault if its duration is greater than a few milli-seconds; however, the probability of detection, even if the resolution of the TDR/SWR matches the fault, is highly unlikely even during an hour of testing. Moreover, if there is more than 1 line to test, this would take an hour for each line. Even a small cable harness of say 10 cables will take 10 hours to provide the same level of test coverage by using a TDR/SWR. By using the Ncompass, the time duration of intermittent faults that can now be detected falls into the nano-second range, and the probability of detection increases by 10^8 for 10 lines under test. Plus because the Ncompass continuously and simultaneously monitors all 10 lines at once with no sampling, digital averaging, the same test coverage for all 10 lines can be completed in 1 hour. Once the fault is detected and isolated on the faulty line, then depending on the data provided by the Ncompass on the band width of the fault, the TDR/SWR could then be applied on to the specific faulty line to establish how far down the the line the fault is. Therefore the Ncompass is an excellent tool for isolating the intermittent faults while the TDR/SWR are great tools for establishing further information on the fault's position. For this reason, the Ncompass 4500 has a built in SWR.

 

end faq