DC-DC power conversion challenges in transportation applications
Transportation vehicles pose a challenging environment for the power supply designer. DC-DC converters are used to power on-board sensors, communication radios, positioning and location sensing, lighting, and information systems. Huge voltage variations and disturbances occur from load dumps, jumpstarts, conducted and radiated switching noise, and spikes from other equipment.
Andrew Hutton discusses some techniques and solutions to meet these engineering challenges.
Today, vehicles are like mobile data centres; the computing, communication and sensing power in a typical car is astonishing. With upcoming autonomous, driverless features, even human participation in vehicle control is disappearing with the driver now cocooned in an environment where he or she expects to be entertained and internet-connected in climate-controlled comfort. It’s not just cars of course; trains, buses and planes and even forklifts can have all these features as well. The transportation environment can be a headache to product designers though, with high temperatures, shock, vibration and extreme electrical disturbances to cope with. The cabin of a parked car in the sun can easily rise to lethal temperatures to electronics, as well as to your pet dog. Leave your smartphone in there and it will refuse to operate until cooled off. When it powers up, you really have to believe that your USB charging outlet is an effective protection barrier to electrical transients. It’s a pretty hostile environment.
The automotive electrical environment
Figure 1. LV124 cold cranking conditions
Figure 2. Summary of 12V DC-DC converter input range requirements
Cars have particular electrical supply characteristics with their typical 12V nominal DC bus swinging over a wide range from as low as 3.2V with cold cranking to 42V with load dumps. The automotive standard LV124 is often applied, established by the German car manufacturers in 2013. Part 1 is for electrical requirements and tests and is quite severe. Figure 1 for example shows the test voltages under cold cranking conditions for a 12V system: the red limit being for start with a degraded battery. LV124 defines different allowable outcomes depending on the tested equipment ranging from functional status A, where there should be no effect, to functional status E requiring repair work.
At the other end of the voltage spectrum, ISO 7637-2 specifies various high voltage transients at different severity levels. Different car manufacturers have their own particular interpretations and requirements but transients are applied up to -220V for 5ns along with higher energy pulses at lower voltages, for example +101V for 400ms. Negative pulses are specified, as this is what results from parallel inductive loads being de-powered. Series inductance, such as in cabling, causes positive-going spikes on switch-off, generally of lower energy. Reverse connection and load dumps need to be withstood as well, high-energy surges up to 27V for 300ms being typical. Because the source impedance set during these load dump tests is very low, it is often impractical to absorb the energy and DC-DC converters on the rail are expected to include the peak voltage in their normal operating input range. A summary of the static voltages that a 12V nominal converter should typically withstand is shown in Figure 2.
A feature of these specifications for automotive applications is that there really is no universal standard. Car manufacturers often set their own limits, often at a more severe level than the generic standards. A saving grace though is that there will usually be some centralised transient suppression, starting with Transient Voltage Suppressor (TVS) diodes or similar embedded in the alternator.
Automotive DC-DC input filtering
Figure 3. Automotive DC-DC input filter
Figure 4. Reverse polarity protection options
For an isolated or non-isolated DC-DC converter to operate effectively on an automotive voltage rail, it should have an input range as wide as possible to withstand the surges and ride through the lowest dips. It will need either internal or external filtering for the higher voltage spikes and some sort of reverse polarity protection as shown in Figure 3.
Here the series diode provides reverse polarity protection, the Metal Oxide Varistor (MOV) provides an initial ‘soft’ voltage clamp then the TVS, after an EMI suppression inductor, forms a harder clamp at a lower voltage, acting like a zener diode. A disadvantage of the series diode though is that it drops some voltage and dissipates power. A parallel diode is a lossless option, which conducts with reverse polarity and blows the fuse, but a neater self-resetting solution is to use a series P-channel MOSFET that only conducts when the input is positive. The other options are shown in Figure 4 without the EMI filtering components.
Railway power specifications
Unlike in automotive, there is no centralised transient and surge limiting in rail applications guaranteed, so electronics at the system level often have to withstand extremely severe stress. The standard generally applied is EN50155, although British standards RIA12 and RIA13 are still sometimes seen. An extra requirement over automotive is for the equipment to withstand regular input dropouts that can last for up to 20ms in ‘Class S3’ applications. Equipment for rail applications tends to be larger in scale than automotive, so DC-DCs in DIN rail and chassis mount configurations are often seen. There are different categories of installation, though, ranging from axle- to body-mounted with different shock, vibration and bump requirements. The category effectively defines the degree of environmental sealing required, up to full encapsulation. An added complication is that the nominal DC system voltage can be anything from 24V to 110VDC.
The surges at worst can be 3.5 x the nominal input or 385V from a 110V nominal lasting for 20ms. Transients for some microseconds are up to several kV defined by the EN61000-4 series called up in the European ‘EMC directive’. Dips can be down to 70% of nominal supply for no loss of function or 60% with some loss.
A summary of the possible voltage surges and dips applied to electronic equipment in the rail environment is given in Figure 5 for US, European and the French national standard NF-F-01-510.
Figure 5. Input voltage ranges for rail applications
Dips and dropouts
As in automotive, the high energy surges are even less practical to clamp or absorb, so DC-DCs are either designed to include the highest voltage in their normal input range or to rely on a pre-conditioning circuit elsewhere that provides a stabilised input. In any case, to include the dips the DC-DC should have the widest possible operating range, with 10:1 input ranges not uncommon. DC-DCs with auto-ranging switched inputs are possible but have to operate consistently and safely with the input voltage surging and dipping across input ranges. Often a practical solution is to precede the DC-DC with a linear regulator that drops the excess voltage during a surge. Its peak dissipation is high but averages to be low, as the surges are relatively infrequent.
Because of the energy levels involved, coping with dropouts is particularly problematical. A large capacitor on the converter input is a simple solution but is impractical for the lower nominal input voltages. Imagine trying to hold up a DC-DC input at 200W for 20ms with a nominal 24V and dropout of 16V. The capacitance needed would be 25,000µF. Worse still, the capacitor voltage rating would need to be 75V to cope with the surges on a 24V line. As of today, that is a component about 2.5” (50mm) diameter and 6” (150mm) long, comparable with the size of a 200W DC-DC! Schemes have been devised to boost the input to a higher voltage level and store energy on a capacitor which is then ‘switched’ onto the input when it is detected to have dropped. Although this does add complexity and cost, it can be traded against the high cost of a large capacitor, and if an electrolytic type can be consequently avoided there is a reliability and life gain.
Figure 6. Common mode noise filtering
Input filters similar to automotive as in Figure 1 will commonly be used in rail applications as well for transient susceptibility and reverse polarity protection. However, automotive and rail specifications also put limits on emissions generated by a DC-DC converter, which must also be dealt with. Sometimes, in both application areas, a degree of common-mode noise emission suppression is needed, particularly if the DC-DC produces isolated outputs. This type of noise is attenuated by a common-mode, or ‘current compensated’ choke inserted in the DC-DC input lines. The windings are arranged such that normal running current magnetic flux cancels, so high winding inductances can be used without fear of saturation. The choke presents, instead, a high inductance and hence high impedance to noise currents, which are common to both lines circulating to ground. Figure 6 shows the typical arrangement. L1 and L2 give some differential mode noise attenuation.
Of course, the ideal solution is to specify a DC-DC converter that already has input filtering and maybe reverse polarity protection built in. Parts are available from several suppliers that have these features and are compliant with EN50155 for rail. Delta and Bel Power Solutions, for example, offer compliance with EN 50155 for their wide input DIN-rail DC-DCs, as does Meanwell for its enclosed RSD30/300 series. Delta additionally has a chassis mount range suitable for harsh environments in transport, the B40/62/70SR series, with IP67 optional sealing. All have high levels of transient suppression and ESD immunity built in.
For a discrete filter, MOVs are available from suppliers such as Kemet, Bourns or TDK, and Transient Voltage Suppressor diodes from AVX, Bourns or Vishay. Common- and differential-mode Inductors can be obtained from suppliers such as Schaffner, TE and Premier Magnetics.
Avnet Abacus can supply parts from the suppliers mentioned from stock for quick time to market for new developments. If you need advice on power supply selection for your design, click the Ask an Expert button to get in touch with our technical specialists in your local language. Alternatively, click below to explore our range of DC-DC power solutions and design resources.
Product Manager, Power, Avnet Abacus
How can combining different types of sensors result in more accurate data?
With more and more applications combining data from multiple sensors, Alessandro Mastellari looks at...
Pressure sensors: Why you should consider going digital
To give the electronic systems we design the ability to detect changes to what is around them we nee...
Pressure sensor miniaturisation: How smaller sensors are enabling better healthcare
In the second installment of our pressure sensor series Alessandro Mastellari explores how manufactu...