My main strategy to save power is to use a intermediate DC bus (running at 19V DC) to provide power to the lot of standby boxes (like internet router, NAS box, ...). Those require various supply voltages from 5V DC to 15V DC. I’m currently using a variety of home made DC/DC converters to convert the 19V DC bus to the required voltages. Since the 19V DC bus must be supplied from the mains power, there’s a centralized power supply unit using redundant sources.
To achieve my main goal, the efficiency of the involved power supplies (mains power to 19V DC, 19V DC to whatever is required) is very important to know and to be able to optimize.
Your typical wall wart or power brick supply (from mains to 5V / 12V DC) in the low watts range (2W ... 20W) has still a not so good efficiency, even regarding the latest energy star regulations. So it is possible to achieve better total efficiency using the 19V DC bus approach.
I’m using comercially available 19V DC supplies with good effiency here, but I’m building my own DC/DC converters to provide the local lower voltages. To achieve better total efficiency than the standard off-line power supply, these must have efficiency better than 95%.
Evaluating the efficiency of such a DC/DC converter isn’t a simple task if one needs accurate results. Moreover, variation of input voltage and output load have influence on the efficiency, and the DC/DC circuit must be optimized for the typical load.
All of the devices I plan to optimize fall into the available output power range of the E36312A supply. Typically the range is from 1W at 5V DC to 20W at 12V DC. The accurate voltage and current readback together with the logging and exporting capabilities will be a great help to find out the typical power supply requirements of my devices. The supply promises to provide an easy way to make measurement of consumed DC power over time.
Second, after knowing the power requirement of a device, I can go for the optimization of the DC/DC converter circuit. Doing so requires a precise measurement of input and output power of the converter, and taking measurements at varying input voltage and load conditions.
The E36312A supply eases the task of measuring the input power to the DC/DC by using the precise readback, remote sensing and logging.
An additional useful feature is the sequencing / ramping capability: This will be used to semi-automatically measure input power over input voltage at constant load.
In case one output channel provides enough power for the DC/DC converter, a second channel would be used to read back the output voltage of the DC/DC, with current limit set to zero (or a near zero voltage).
Now one can read back the output voltage of the DC/DC using one channel of the power supply, only one additional amp meter will be required to read the DC/DC output current and finally calculate the efficiency.
Imagine now, three of four measurements are done by the bench power supply unit at constant DC/DC output load, and only one measurement left: efficiency over load variation.
If one uses a constant current sink or a simple power resistor as the load to the DC/DC converter, and has the DC/DC converters output connected to the bench power supply to read back the output voltage, one could also supply some current (in CC mode) from the bench power supply into the load. This reduces the current drawn from the DC/DC output by the amount supplied from the bench supply.
Now it is possible to evaluate (and semi-automate by using the remote control capabilities) the efficiency of my (or anyone elses) DC/DC converters by using just three components: The E36312A bench power suppy, a precise ampmeter and a suitable load resistor / current sink.
Go back