When a designer is concerned about the distance between the power supply and the load, it is usually related to IR drop (V = IR) or losses due to noise pickup. The IR-drop problem can be solved by various strategies. These include using heavier wires, four-wire Kelvin sensing at the load (although it presents its own potential problems due to feedback loop oscillations), or a more distributed architecture with an intermediate bus converter providing Multiple point-of-load (PoL) DC/DC converters nearby. Noise is attenuated by ferrite beads on the power lines and bypass capacitors placed close to the load.
However, the same physical law that characterizes voltage drop is also called power dissipation P = I 2 R. Of course, this power is dissipated as heat, which is not a problem in most designs because the power loss in the cable and the heat generated are negligible compared to the overall system dissipation.
This is changing as data cables and even DC power cables are increasingly routed alongside other cooling cables, often in places with little or no convection cooling. In many commercial and industrial environments, a large number of AC cables run through risers and plenums, so there is little airflow.
In fact, the National Electrical Code (NEC) in the United States and similar codes around the world define the maximum allowable free air dissipation for these cables and then add derating factors for cables with little or no air flow. It gets worse when these power cables run alongside other cables, so in addition to cutting off airflow, there are adjacent heat sources.
Of course, placing cables carrying high currents in enclosed spaces is nothing new. Industrial cabinets used for motor control as well as power cables in data center racks can have hundreds of amps. But these units are designed to support the associated heat loads, and their operating environment is limited. No one is going to casually "throw" another cable into a well-designed server rack that handles kilowatts.
By contrast, pushing an ethernet cable or two into an air chamber already filled with AC cables seems harmless. However, the heat from these cables can degrade the insulation and electrical performance of the Ethernet cable.
The increasing use of higher power Power over Ethernet (formerly known as PoE++, now officially designated as IEEE 802.3bt) is exacerbating this situation, as it allows more than 100 watts of power to be delivered to the load. While most of this is dissipated at the load, some will be dissipated along the Ethernet cable itself. For free-air cables or relatively free spaces, this is not a problem; however, many of these PoE cables end up being laid under carpet, or in narrow risers, with the narrow cables running alongside the AC cables (Figure 1 ).
Figure 1 Power over Ethernet (PoE) provides flexibility in powering remote devices, but also allows for temporary cabling.
Compared to the problems associated with charging electric vehicles (EVs) at high-current charging stations, these self-heating problems are minor.
Electric vehicle chargers experience severe thermal stress
In the case of EV charging, we see hundreds of amps or more flowing through the charging cable despite the relatively short distance between the charging station and the car. In fact, this risk of overheating of the cable is one of many limitations limiting the charging rate (Figure 2). This is a problem that power system designers have already addressed, but more can be done.
Figure 2. Key components of a typical DC EV charging system, using the combined charging system Type 1 standard (J1772 AC + CCS) connector as an example.
One team analyzed, designed, and tested a method to increase the current-carrying capacity of ultra-fast EV charging cables from a little over 500 A today to over 2,400 A—a nearly fivefold increase. They developed a method to predict the heat transfer and pressure drop characteristics of laminar and turbulent flow through concentric rings with uniformly heated inner walls and adiabatic outer walls. By capturing heat in both liquid and vapor form, liquid-to-vapor cooling systems can remove at least ten times more heat than pure liquid cooling (Figure 3)
Figure 3. Schematic diagram of the circulation geometry and boundary conditions used by the researchers for the fundamental thermal model.
Their scheme involved pumping the highly subcooled dielectric liquid HFE-7100 through a concentric ring simulating an actual cable segment, with a uniformly heated 6.35mm diameter inner surface representing the electrical conductors and an insulating 23.62mm diameter outer surface for the exterior catheter. At these power, heat, and fluid flow levels, the test and measurement "plumbing" is complex, as are the various sensors required to control and measure it, as well as the parameters of interest (Figure 4).
Figure 4 identifies key components
