Behind the scenes: VR-Zone's PSU Testing Station Mark II

[siopao1984]

---

In VR-Zone we continuously strive to provide you with the highest quality and most accurate product reviews. Thoroughly and accurately testing a power supply is a complicated matter, not only because the testing has to be done in a controlled environment and with full control over the power load but also because laboratory grade instruments have to be used for taking readings.

For that purpose, we performed major hardware and software upgrades on the testing station that we have been using for the past couple of years. In order to be able to fully control our testing procedures and emulate real world conditions as closely as possible, we upgraded our reading instruments and replaced our resistive loads with fully electronic loads, thus achieving quality and accuracy in line with the most expensive of power supply testers currently available.

In order to create a real-world environment for our testing we decided to modify and use a common SECC steel tower case for housing the power supplies during our testing. Inside the case we installed three Watlow Firerods, rated at 500W power each, the output of which is being controlled by a common AC rheostat.

The stock rheostat and its axle have been replaced with a multi-turn precision rheostat for accuracy, driven by a DC motor with a modified speed control circuit which is controlled by our CompuLAB interface and custom software. In layman's terms, we only have to enter our desired ambient temperature in our software and it automatically adjusts the power of the Firerods to match it throughout our testing.

Read more: Behind the scenes: VR-Zone's PSU Testing Station Mark II by VR-Zone.com

[siopao1984]

Power supply units are nowadays made to work on almost every power grid and their efficiency differs depending on the input voltage. We did not want to restrict our testing on a 230VAC power grid, therefore we acquired a common 3KVA laboratory VARIAC. The VARIAC allows us to adjust the input voltage between 0 and 260VAC, giving us the ability to test the performance of all power supplies at both 110VAC and 230VAC input.



Between the VARIAC and the power supply we are using an Extech 380803 True RMS power analyzer, which replaced the older phase Lutron DW-6091 we have been using in the first version of our testing station. The Extech 380803 is a class (G) 0.5 instrument, meaning that the maximum error is less than 0.5%. Furthermore, the Extech 380803 allows us to take readings of the input apparent power, real power, input voltage, input frequency and of the power factor of the power supply, all of which are transferred directly to our software via its RS232 interface.

[siopao1984]

Although our proprietary automated load served us well for nearly two years, we decided that it was the time to bump up the quality level of our reviews to equal that of professional testing equipment. Thus, we replaced the load we fabricated ourselves with eight Vantec CA5030 electronic loads. The Vantec CA5030 electronic loads allow extremely precise control over all voltage lines, permitting current control down to two digits. One of the loads is reserved for the 3.3V line and a second one is reserved for the 5V line, leaving 6 loads to be utilized by the 12V line(s), thus allowing us to place a maximum power drain of 1800W on the 12V lines alone.



For the interconnection of the power supply with the reading instruments and the electronic loads, we had a PCB manufacturer print us our own professional test fixture. The test fixture allows for the professional and quick interconnection of the test units and equipment, plus the integrated high precision current shunts aid in the accurate extraction of voltage readings.

[siopao1984]

Voltage measurements are easy to make; even if we did not have our extremely accurate CompuLAB USB interface which is responsible for all of our voltage measurements, a good multimeter with an USB interface would do the job. However ripple can only be measured by using an oscilloscope meeting certain specifications. We decided to acquire an USB Instruments Stingray oscilloscope simply because it is the most popular device used for that purpose today and, like every other measurement instrument we are using, its results are being transferred directly to our PC.

[siopao1984]

For temperature measurements we are using yet another laboratory USB interface, the Novus myPClab, in conjunction with two PT100 sensors. One PT100 sensor is installed at the intake of the power supply inside the case and the other sensor is installed horizontally over its exhaust. This particular interface is specifically designed to work with temperature sensors and runs a linearization algorithm to increase the accuracy of our PT100 Class A sensors up to 0.2%. We are also using a calibration box which allows us to occasionally check and fine-tune the accuracy of the system.



A SL-5868P digital sound level meter with an USB interface is being used to monitor and record the sound level of the power supply under test. During our acoustics tests all cooling fans of our testing station and PC temporarily cease to operate in order to produce the most accurate results possible. The sound level meter is mounted one meter away towards the front side of the case in a straight line and with no obstacles in-between.

[siopao1984]

Most of our measurement devices came with their own computer software; however that was just not good enough for us. Even for the first version of our testing station, we needed and did develop a software able to monitor and control all of the devices simultaneously along with our DC load, as well as take measurements and create graphs altogether. We developed that first version of our software to be fully customizable for whenever it would be deemed necessary, thus adjusting it to control the Vantek CA5030 electronic loads instead of our relay-controlled load was a very easy procedure.