How to test electronics out in the field without messing up the results.
Consumer electronic products are not just getting smaller, they are turning up in tinier nooks and crevices in systems that have to suffer dirt, dust and vibration. Electronic systems now routinely drive cars, trucks and aircraft. At the same time, the aerospace and defence industries are tightening environmental specifications. All of these trends require that instrumentation used in all phases of development - design, debug and validation - cope with increasingly severe environments.
These outdoor environments pose challenges for oscilloscopes and probes. Systems can experience drastic temperature swings, can be difficult to access, or can involve tiny devices that need to be tested, as well as other mechanical considerations. The good news is that, with some care, a simple understanding of what to expect in these 'extreme' environments, and some common sense, it is possible to make high-quality, reliable measurements using readily available probes and accessories.
'Extreme probing' can be any probing situation in which, without proper consideration, it is possible to cause damage to probes and accessories, or make an unreliable or poor measurement. Common hacks that engineers use to make these measurements often have dire consequences when it comes to signal integrity. Other extreme situations will lead to probe failure no matter what steps are taken; in these cases there are several strategies that can help extend probe life as long as possible.
With ever-shrinking components, engineers often find themselves probing leads that are difficult to see with the naked eye, and even harder to maintain good electrical contact with, particularly using hand-held browsing probes. Most probe manufacturers offer interchangeable probe heads. Systems like this are the best solution for electrical performance-critical applications, but they also offer added convenience. With these types of probes, engineers purchase a probe amplifier and any number of heads, each tailored to a particular probing situation.
With a solder-in probe head, the small inconvenience of carefully soldering the probe in place can save huge amounts of time when returning to the test point later: the connection only needs to be made once, which is very useful for hard-to-access connections. For designs with a considerable amount of test points, however, this can be costly, as probe heads are precision electrical components.
Variations of the traditional solder-in probe head offered by probe manufacturers solve this problem. Zero insertion-force (ZIF) probe heads and tips work with differential probe amplifiers and can reduce probe costs significantly when many high performance measurement points are needed. These typically encompass disposable tips that can be soldered down to various places on a circuit board. The more expensive and sensitive probe head can then be shuttled between these tips.
Engineers may be tempted to add length to the tip wires of probes to access hard-to-reach areas. Active probes typically have a bulky amplifier near the end that may not fit into tight spaces, or the oscilloscope may be far enough away from the device that the probe cables won't reach. Some manufacturers offer probe heads or tips that extend this dimension. These tip wires can be modelled as coupled transmission lines.
Usually in probes, the termination of these wires has higher impedance than the characteristic impedance of the wires, which causes the input impedance to resonate low at the quarter-wave frequency. Increasing the length of the wire increases the length of the transmission line and therefore pushes the resonant frequency lower, decreasing the usable bandwidth of the probe. It is important to realise that spreading the tip wires apart results in the same effect of decreasing the measurement bandwidth.
A tip resistor, which can usually be found on all high bandwidth probes, helps alleviate the resonance issue by limiting how low the input impedance can resonate. It is important that this resistor be as close to the device-under-test as possible. Unfortunately, the tip resistor does not prevent the decreased bandwidth caused by longer tip wires. Although doing so may seem like a quick and easy way to probe tricky devices, extending the length of tip wires is not a practical solution for quality, high bandwidth measurements.
Understanding details of the probing system can help engineers make reasonable decisions when pushing their designs and equipment to extremes. Certain probing system architectures are more suitable for use in extreme measurements. Traditional probe architectures use a relatively high-impedance compensated divider in front of a high-impedance voltage input amplifier. Due to the high impedances, this topology does not lend itself to the use of controlled impedance transmission lines. As a result, the compensated divider must stay within close proximity of the amplifier.
Most probe manufacturers now use the concept of putting a lump in the cable. Under this approach, a minimal passive tip is connected via flexible, small-diameter coax cables to an amplifier with input impedance equal to the characteristic impedance of the coax. This system forms a terminated transmission line system, with less parasitic effects and loss of fidelity than that of traditional systems. This topology enables different probe head tips to be used for different applications.
In the case of extreme probing, the transmission line in this architecture offers a huge advantage. The length of the transmission line can be increased dramatically without a loss in bandwidth. Coaxial extension cables can provide an electrically sound solution to additional probe length, adding several feet between the probe amplifier and the probe head, for lump-in-the-cable probe architectures.
A longer transmission line between the probe head and the probe amplifier does have minor impacts on signal integrity. Longer coax cables will have some loss and may increase the equivalent input noise. It is very important to use high-quality coaxial cables for this connection. The two most important qualities in these cables are phase matching and loss. Deviations from perfect phase-match will degrade the signal integrity of the differential signal arriving at the amplifier. Minimising cable loss will help reduce in-band loss which manifests itself in the time domain by reducing rise times. It is possible to buy extension cables as low-loss cables that come in phase-matched pairs, resulting in minimal impacts on signal integrity.
Cables that enable
These extension cables allow for measurements in complicated setups, and in some cases, the measurement may even be impossible without their use. They are especially convenient when the probe amplifier does not fit near the device under test. The cables allow the amplifier to remain outside of an enclosure, away from the device. In combination with solder-in probe heads, the probe amplifiers can be interchanged among the cables without turning off the device being tested.
Testing circuits over extreme temperature ranges has become quite common. Agilent classifies temperatures into three general ranges: a standard range over which products are fully warranted (5°C to 40°C), a more severe range typically used when abuse testing consumer electronics (-25°C to 80°C), and an extremely harsh range typically used in the automotive and aerospace industries (-55° to 150°C).
Active probes, which typically have an amplifier ASIC built into the probe cable - the lump in the cable - cannot withstand temperatures outside of the standard range, because the precision electrical components that make up the amplifier will not work effectively. And there are thermal dissipation issues at high temperatures.
It would be very impractical and expensive to develop an active probe amplifier that can withstand and operate reliably in these extreme temperature ranges. Also, scope-probe calibration assumes that the ambient temperature surrounding the probe amplifier and oscilloscope does not change significantly over the course of the measurement.
A much more practical approach is to leave the amplifier at ambient temperature, while exposing just the probe head to the extreme temperatures. Probe heads are made up of only passive elements and are not necessarily as sensitive to extreme temperature swings. Agilent has tested a variety of InfiniiMax probe heads over the three temperature ranges; check with your probe manufacturer to learn the absolute temperature limits of your equipment. With extension cables, it is possible to subject the device under test and probe heads to drastic temperature swings, such as in a temperature chamber, without affecting the probe amplifier or oscilloscope.
Although it is possible to simply use heat guns and freeze spray to heat and cool your circuits, these methods provide very little control or consistency over the heat applied or removed. It is very easy to exceed the material limitations - that is, the melting points - of probes and circuits using heat guns. A temperature chamber provides a controlled environment, resulting in much better temperature accuracy.
At extreme temperatures, particularly in the most aggressive range, many material properties of components change drastically. Plastics can become brittle and fragile when cold, and most approach their glass transition temperature or melting points around 150°C. Special attention should be paid to cable jacketing as it is especially vulnerable to high temperatures. Kinking probe head cables and device cables should be avoided if they will experience very high temperatures.
Fixturing and securing probes is important when they are connected to a device in an environmental chamber, since it is nearly impossible to make careful positioning adjustments once the device and chamber is at an extreme temperature. Most glues and tapes do not handle extreme temperatures well. It is often good practice to test adhesives' unknown temperature properties in a safe area. Probe-head cable jacketing can be considered a safe area; even if the adhesive disintegrates or flows, the cable jacketing should protect the centre conductors and signal integrity of the probe head. Another consideration for drastic temperature changes is the formation of condensation on the probe components as well as the device under test. Water can cause shorts and result in electrical damage to the probe heads or the circuit electronics. If condensation begins to form, disable power to all devices and wait for the water to vaporise before proceeding with additional testing.
It is important to note that lifetimes of all components, circuits, and probes are significantly shortened when used in the most extreme temperatures. Repeated temperature cycling, in particular, is very harsh, which is why lifetime testing of designs tends to be performed with extreme temperatures in highly-accelerated lifetime testing chambers. This means that, even though probe heads can be used in these extreme ranges, they will have a shortened lifetime. Check with the probe manufacturer for reasonable cycle limits to expect for qualified probe heads.
It is important to realise that probe head failures are more likely at extreme temperatures. Depending on the measurement, a probe head failure due to extreme temperature cycling may not be immediately obvious from the measurement on the oscilloscope. For instance, the symptom of an Agilent probe head failure due to temperature is a DC open circuit on either the positive or negative side; this will vary with probe manufacturer. If a failure is suspected, it is important to measure the continuity from the probe tip to the connector on each side. The DC resistance should be 25kΩ.
It is a good idea to keep probe heads used in extreme temperature situations isolated and labelled differently from normal probe heads. Input resistance of the probe heads can be measured before extreme tests to make sure they will still make a reliable measurement. The cable jacket may stiffen and discolour during extreme temperature testing. This is normal and should not be a cause for concern - 150°C is near the reflow temperature of most PVC-type plastics such as that on Agilent probe head jackets.
As long as the coax cable is completely covered by the jacketing, it is functional. Most high-performance RF cables are made using a Teflon dielectric; depending on the manufacturer, this can be either extruded or tape. Tape dielectric cables are usually more flexible, but they can be more fragile as well. Kinks and excessive bends will cause permanent deformation of the dielectric in cables, which can affect electrical performance. Care should be taken to not drape probe head cables or extension cables over sharp objects, such as a heat sink, that could slice through a heat-softened jacket. A cable draped over an electrically active area could potentially melt the jacket and short that area to the cable shield.
Another aspect of lifetime testing and real world product testing involves mechanically vibrating and shocking designs. This can simulate the dropping of a cell phone or the installation of electronics in automobiles. Probing considerations for this case are similar to that of extreme temperatures and extreme geometries: protect the sensitive probe amplifier, use solder-in probe heads or tips to avoid open circuits when connecting to your circuits, and strain-relieve probe heads and cables to prevent movement from causing any disconnects.
For the most part, common sense and a simple awareness of the limitations of probing equipment can guarantee long equipment lifetimes and quality measurements. Most oscilloscope manufacturers make a variety of accessories that can help with special applications as well. These tips should help make accurate measurements, while prolonging the life of the probes and accessories in extreme environments.
Jason Swaim is a mechanical design engineer and Ned Brush is a hardware design engineer at Agilent Technologies