I. Introduction
For a long time, in the design, construction, and application of power supplies for data centers in China, "zero-ground voltage" has been touted as something extraordinary, even becoming a primary indicator of power supply quality. This trend has intensified in recent years, and incredibly, this unscientific concept of "zero-ground voltage" has even been incorporated into some national standards. For example, a certain GB-level data center design specification requires that "the effective value of the zero-ground voltage of the UPS power supply system be controlled within a range of less than 2V." Many manufacturers and users habitually attribute various problems in data systems to zero-ground voltage. Currently, the impact of excessively high "zero-ground voltage" on IT equipment, such as mainframes, minicomputers, servers, disk storage devices, network routers, and communication equipment, based on "statistical data," can be summarized as follows:
■ This could cause "mysterious" and fatal damage to the microprocessor CPU chip in IT equipment;
■ This may increase the probability of IT equipment crashing;
■ This may lead to an increase in network transmission error rate and a decrease in network speed;
■ May cause damage to storage devices, data errors, etc.
■ Some well-known IT manufacturers stipulate that the machine will not be powered on if the neutral ground voltage is greater than 1V.
However, a review of international IEC and UL power supply standards reveals no mention of the term "zero-ground voltage," and a thorough search of IEEE articles yielded no literature on the impact of zero-ground voltage on IT loads. Interestingly, I once accompanied power supply experts from Europe and America to visit some Chinese data center users. Some users raised the question of zero-ground voltage, and these experts, who had decades of experience in power supplies and participated in drafting the US UL power supply standards, couldn't understand it at all. After repeated explanations, they finally grasped the basic meaning of "zero-ground voltage," but one of them asked in surprise, "In China, is there any conclusive evidence of the impact of this voltage on IT loads?"
Although there is no conclusive scientific evidence regarding the impact of neutral-to-ground voltage on IT loads (most discussions confuse ground potential with neutral-to-ground voltage), many users in China are willing to invest heavily to address this dreadful and mysterious issue. For example, one telecommunications data center purchased dozens of transformer cabinets and installed them at the input terminals of various floors to reduce neutral-to-ground voltage. This not only resulted in significant resource waste and reduced the reliability of the data center's power supply system, but also substantially increased the data center's operating costs, further exacerbating the already unprofitable IDC business.
Therefore, the author believes that it is essential to systematically discuss the generation and transmission mechanism of "zero-ground voltage" in the power supply system of the computer room, especially its impact on IT load, so that the design, construction and users of the power supply of the computer room and the computer room have a scientific understanding of the "zero-ground voltage" problem.
[Full Text Download] Scientifically Understanding the "Zero-Ground Voltage" Problem in Data Center UPS Power Supplies
