Don’t Make the Cable the Weakest Link in Delivering Power Supply Efficiency
Amateur cyclists can spend considerable sums on the very latest, ultra-light-weight, high-performance bikes in order to shave seconds off their personal best times. And, while investment in the newest tech is no doubt going to enhance performance, it may be that in some cases a rider could do even better if they simply lost those few extra kilos that they have been carrying since the holiday season!
But what does this have to do with delivering energy-efficient power designs?
Well, the situation with the cyclist and the bike is not so different to that of the power engineer who puts effort and resource into creating the most efficient AC/DC or DC/DC power converter, only to find that the in-the-field efficiency is reduced by ‘real-life’ factors. One of those factors is losses from the cable that connects the supply to the circuit it is powering.
And, in the same way as losing weight can improve a cyclist’s personal best, changing a cable can make a difference to optimizing overall system performance of the latest ultra-efficient power architectures.
Factors to Consider When Choosing Cables
Consider the example of a modern 60W or 100W power adapter in which a highly efficient AC/DC power supply is used to quickly charge a mobile phone or other portable device via USB.
If the power supply is built around a topology such as Eggtronic’s QuarEgg and SmartEgg, designers can expect conversion efficiency of up to 96% at full load and 92% at light load.
However, those efficiency benefits can quickly be eroded by compromising on the USB cable. The voltage drop resulting from the wrong choice of cable may mean that the device does not work as originally intended - charging more slowly than expected or, in extreme cases, not working at all.
Of course no cable technology can be completely loss-free and even on a good cable it is not unusual to see a power dissipation of around 2W within a 100W charging scheme. The key to mimimizing this dissipation is to keep the I2R loss as low as possible, which is why cable resistance is fundamental.
Because a conductor’s resistance is determined by the equation R = ρ*L/A (where ρ is material resistivity, L is conductor length and A is cross-sectional area) then resistance will depend on cable length, diameter and the type of material used. Though this does not automatically mean (as people sometimes assume) that the more expensive the material the lower the resistivity. Gold, for instance, with a resistivity of 2.45 x 10-8 ohm-meter, has a higher resistivity than both silver (1.64 x 10-8 ) and copper (1.72 x 10-8 ).
When it comes to the USB cable in our example, there are certain minimum specifications that need to be met with respect to voltage drop. For a USB Type-C cable, for example, USB Type-C Cable and Connector Specification Release 2.0 (2019) states that the maximum allowable cable IR drop for ground is 250mV and 500mV for the VBUS. The diagram below shows the factors that contribute to these IR drops.
However, these are only minimum requirements and, in reality, the best solutions should deliver better performance than the basic specification. At Eggtronic, for example, when it comes to high current capability we avoid cables longer than 2m to reduce the power lost. through the cable length.
Let’s consider a simple example and suppose that you have an Eggtronic 45W adapter capable of 95% efficiency. When compared to a standard efficiency of 91% you have saved 1.8W power. However, if you then have a poor quality cable exceeding the USB limit - e.g. 750 mV on VBUS and 500 mV on GND resulting in an extra drop of 500mV (compared to the maximum limit) - then at 3A output full load power loss (I*V) will be 1.5W. As a result the improved results in adapter efficiency enabled by the advanced converter technology are almost all eliminated thanks to poor cable performance.
|Cable Quality||Cost||Possible Voltage Drop (VBUS)*||Possible Voltage Drop (GND)*||Total Expected power loss for 3A|
|Poor Quality||Cheap||827 mV||706 mV||1.533W|
|Mid-range Quality||Mid-range||494 mV||247 mV||0.741W|
|High Quality||Expensive||361 mV||187 mV||0.548W|
* The maximum allowable voltage drop is 500mV for VBUS and 250mV for GND.
The table above shows the results of some research that Eggtronic has conducted on various cables in its own engineering labs. As this illustrates, saving money on cables is a false economy when it comes to delivering high-efficiency adapters.
Beyond 3A - E-Markers
Unsurprisingly, an important consideration when specifying cables is the current that needs to be carried. Not only does a higher current contribute to higher losses (thanks to I2R) but once you get above 3A the cables also need to include additional hardware.
The USB-PD (power delivery) specification allows a device to be charged at a maximum current of 5A. The 3A rating is sufficient for the charging of mobile phones and other devices with chargers of up to 60W (20V/3A) and all USB-C cables should be able to handle this level of current. However, the latest adapters for the fast charging of larger devices such as laptops deliver 100W (20V/5A). The more the current rise the more the resistive cable losses increases.
All USB Type-C cables that support 5A (or exceed 60W) must incorporate an ‘E-Mark’ protocol controller IC to ensure safe power and data transfer. Communicating between the power source and the load, the E-Mark chip provides all relevant characteristics of the cable, including length, the maximum current and voltage that can be supported and all protocol and transmission requirements.