Most of FCL's support to clients has included a significant proportion of hardware 'hands-on' work. Laboratory based work has been necessary to verify performance to specification, to solve issues and to trial various improvement options. The following paragraphs describe some examples..
FCL has significant experience at RF and microwave frequencies with microstrip, stripline, co-planar waveguide (CPW) and combinations of these. For frequencies even into the gigahertz region, the PCB dielectric substrate is often a type based on the industry standard FR4, of which there are now many 'refinements' like finer and more even fibre weave and better (lower loss) substrate materials. Many of these are multi-layer architectures, sharing RF/microwave, DC and high speed digital circuits with suitable provision to minimise interference between them. In fact, it has been surprising how adequate basic, low cost FR4 has been for many clients' hardware solutions. Being much cheaper to produce than the dedicated low-loss dielectric types, such as Rogers®, sometimes the necessary small compromise in RF performance is tolerable. Attention to detail and strict design of distributed features including controlled impedance lines, through hole technology, grounding and layout yields a product which works well. Where necessary, double sided PCBs have been used, with intelligent distribution of circuit function areas to each side to optimise performance including isolation, EMC, power handling, ease of assembly and test.
At the higher microwave frequencies, FCL has also worked with flexible, Rogers® type low loss dielectric materials and rigid (and brittle) alumina types for space hardware applications. On circuits like these FCL has successfully solved many RF isolation, screening and EMC issues with techniques such as compartmentalisation, frequency-appropriate attention to layout, selection of suitable high Q factor components and very good (low resistance and low inductance) grounding provision.
FCL is often engaged to undertake hands-on debugging work. The problems solved have ranged from initial power-up issues: circuit breaks, excessive loss or reflections through to eliminating some unusual spurious products caused by a strange mixture of harmonics and intermodulation products. Sometimes these issues only manifest at temperature extremes, so hardware often needs to be set up in a thermal chamber and all of the challenges of test equipment interfacing and de-embedding of the cable characteristics satisfactorily addressed.
Initial power-ups are of course done very cautiously. After a very detailed visual examination and verification of the connections and checks for short circuits, the voltage would normally be slowly increased whilst monitoring the current. Some circuits have non-linear DC current voltage relationships which must be allowed for. Also there may be particular current demands for initial power-up from cold and boot-up. For the biggest challenges, FCL has often verified proposed solutions by building, testing and documenting temporary modifications to existing hardware and then following these through to the formal engineering changes and product improvements. The debugging process is often performed at the same time as other circuit changes, enhancements or cost saving measures.
FCL is often required to support clients during the development and building of prototype hardware to tight timescales. Often this results from a commitment made to the customer by the client. The experience that FCL has acquired in similar industries enables it to rapidly provide this support. FCL is familiar with many types of RF and microwave test equipment such as swept CW and pulsed vector network analyzers, spectrum analyzers, noise figure meters, power meters, high speed oscilloscopes and dynamic signal analyzers. This extends to how to fully exploit their capability to produce reliable and accurate measurements including instrument control and data acquisition over the general purpose interface bus (GP-IB/HP-IB/IEEE488) and TCP/IP (Ethernet® and Gigabit® LAN). FCL has a good grasp of what can be done realistically and at what frequencies. Evaluation assignments like these often include writing or updating the associated design or test specifications. At the same time FCL can sometimes find ways to make the test engineer's job easier and more productive.
Several clients have engaged FCL to build and verify demonstration equipment whilst they concentrate on other priorities. FCL may well have enough previous experience with similar equipment to quickly understand what the demonstrator is intended to do and to quickly familiarise with the objectives. FCL can treat this process as a stand-alone project: assess the priorities such as long lead items and the fulfilment of commitments that the client has made, and get to work on it. Clients have then required FCL to play a central role in describing, documenting and demonstrating the equipment to review teams and potential new customers.
As well as the electrical functionality of a product, FCL also understands many of the non-electrical properties that are part of the product specification and have to be addressed at the same time. For example, calculations for power dissipations and heat transfer mechanisms by convection, conduction, radiation, and combinations of these. FCL is also conversant with the typical product physical properties such as mass, weight, moment of inertia, centre of gravity and centre of mass.
FCL has written source code in Excel Visual Basic for Applications®, Microsoft Visual Basic 2015 Express®, Visual C/C++ 2015 Express® and Matlab 2020b® using the instrument server application. These programs and scripts have been used for controlling test equipment, extracting the data, processing it as necessary and its subsequent reporting and presentation. Physical interfaces used have been GPIB, universal serial bus (USB) and TCP/IP (Ethernet/Gigabit®). FCL has also achieved highly effective control with various ad-hoc parallel and serial interfaces.
Sometimes equipment fails to meet specification due to some particularly difficult parameter to achieve or perhaps something unexpected happens at an extreme end of the operating temperature range. FCL has experience of successfully finding and clearing many challenging faults such as these. A very good understanding of the product is required: how it operates and was designed together with fully utilising the test equipment available, sometimes using unusual techniques.
In one example a frequency synthesizer was loosing lock randomly at low temperatures. Some painstaking investigations showed that the associated voltage controlled oscillator (VCO) was coming to the end of its range prematurely. An examination of the design documents and bill of materials yielded details of the components used. Some further research into their characteristics revealed that some had inadequate high frequency performance. These were replaced with correctly specified ones which enabled the VCO to work within specification. The problem centered on the core materials used for some inductors which were magnetic (μr>1;) and not air-cored (μr=1;) types. The magnetic properties of many materials vary significantly over temperature and this was no exception.
One product that FCL investigated, designed for operation in an external environment in the UK, was failing due to overheating at several of the customer's sites on summer days. Internal power dissipation was quite modest and convective cooling was thought to be adequate for external use in the UK based on yearly air temperature statistics. However further investigations revealed that insufficient margin had been allowed specifically for solar (thermal or infrared) radiation, previously not thought to be of particular concern in this relatively cool climate. FCL quantified the level of solar heating expected by application of the Stefan Boltzmann 'fourth power' radiation law after making allowances for the surface emissivity, infrared spectral density, atmospheric absorption and integrated daily solar power flux. A representative unit was tested in the laboratory using solar lamps suitably orientated and calibrated for infrared content. The effects of solar radiation were found to amount to several hundreds of watts equivalent heating power thus confirming the cause of the overheating. Several recommendations were made for ways of mitigating the effects.