中兴ZXDU58 W121室外一体化通信电源机柜

中兴ZXDU58 W121室外一体化通信电源机柜

中兴ZXDU58 W121室外一体化通信电源机柜

概述

　　ZXDU58 W121 V4.0系列化产品是室外型120 A通信直流电源系统。默认输出直流电压-53.5 V，适用于220 V/380 V交流电网。该产品系列包含丰富的机柜系列和温控系统组合，内嵌120A电源插箱，构成各种室外电源系统。

　　ZXDU58 W121 V4.0系列室外电源系统包含机柜和电源插箱两部分组件构成。　　

　　特点

• 宽电压交流输入范围，80 V - 300 V，减少了电力不稳定对设备的影响。
• 多种温控设备可供选择，适应各种场景的温控需求
• 支持普效整流器与高效整流器混装，节省能耗
• 提供三相和单项两种可选交流输入模式
• 提供一体化机柜和分体化机柜供用户选择

　　用户体验

•  选择范围宽
• 适应性强　中兴ZXDU58 W121室外一体化通信电源机柜

Summary

The ZXDU58 W121 V4.0 series product is an outdoor 120 A communication DC power supply system. The default output DC voltage -53.5 V is suitable for 220 V/380 V AC power grid. The product series includes a rich combination of cabinet and temperature control systems, embedded in 120A power supply box, and forms a variety of outdoor power systems.

The ZXDU58 W121 V4.0 series outdoor power supply system consists of two components, a cabinet and a power supply box.

Copy of _MG_6971 copy.Jpg _MG_6941 copy 1.jpg _MG_6935 copy 1.jpg

Characteristic

The wide voltage AC input range, 80 V - 300 V, reduces the effect of power instability on the equipment.

A variety of temperature control devices can be selected to meet the temperature control needs of various scenes

Support mixed universal effect and high efficiency rectifier rectifier, energy saving

Two kinds of optional AC input modes for three phase and single item

Provide integrated cabinet and split cabinet for user selection

User experience

Wide range of selection

Adaptability

随着近年来数据中心的大规模建设，传统供电系统在大规模部署和运营中暴露的可靠性、维护性等问题日益突出，推动着用户、设备商和方案设计公司合力进行供电系统的创新和优化，供电系统的建设思路逐步从传统上关注可靠性转移到保障可用性上来。那么为何要建设高可用供电系统，如何建设高可用供电系统，本文对此做出了一些探讨。

可用性综合反映用户的真实需求,可靠性是影响可用性的因素之一

可靠性通过可靠度来衡量,可靠度的定义为:“给定系统在规定的工作条件下和预知的时间内持续完成规定功能的概率"。平均工作时间MTBF ( 又称平均故障间隔时间) 是决定电源系统可靠度的重要指标，MTBF可通过定量定时的工业试验或理论计算的方式获得。可用性是指产品在任一随机时刻需要开始和执行时，处于可工作或可使用状态的程度。可用性计算公式是：

式中，MTBF(Mean Time Between Failures)是平均故障间隔时间，MTTR（Mean time to repair）是平均修复时间。

可靠性的高低代表了电源系统是否容易故障。但是从实际应用的角度来说，任何设备都不可能保证在生命周期内不出故障，用户希望的是设备尽量不出故障，即使故障了也不要因故障导致业务受影响；如果业务受到了影响，那么应尽快消除故障。相比之下可用性的定义相比可靠性范围更加宽泛，对于可修复系统而言，它不仅涵盖了设备是否容易出错的问题，还涵盖了设备是否容易从故障中恢复。很明显可用性更加真实地反映了用户的需求。

在UPS行业，通常用几个“ 9"来代表系统可用性的高低。它是指一年内，系统在线运行及可进行生产的时间比例。比如6个“9"(可用性可达到 99.9999%)，即每年可能存在的宕机时间少于 32 秒。UPS系统的目标是尽量提高 UPS 电源系统的可用性，减少来自市电的影响。

提升供电可用性的途径

提高供电系统可靠性

从可用性计算公式可以看出，提高可靠性是提高可用性的一个重要途径。提高供电设备可靠性分四个层次：

，设计标准级。在产品规划设计阶段，应充分考虑产品的可能应用环境，选定相应的设计标准。对产品使用时可能的电气隔离、EMI/EMC、防雷、防浪涌、防噪干扰等电环境，防湿、防尘、防震、防腐等自然环境，及操作、维护、管理、搬运、安装等的人环境有充分的评估，从而构建产品合理的设计框架。

第二，器件级。在产品设计阶段，严格筛选器件，配合电路设计，并反复模拟各种恶劣环境测试器件应力裕量，保障各类元器件的可靠运行。对于关键器件如电解电容，如果电路设计不够优化，纹波电流过大，芯温过高，寿命将大大缩减，从而导致设备可靠性降低。散热风扇也要选择稳定性好性能优异的厂家提供，防止风扇故障导致功率模块温度上升，影响正常供电。

第三，部件级。部件的可靠性主要体现在它的稳定性和冗余性，在保证部件故障率降至的前提下，关键部件采用冗余设计是提高部件级可用性的方法。

第四，方案级。通过优化系统设计，使供电系统运行可靠稳定，并且具备容错能力，整个供电路径无单点故障点。图1展示了一个无单点故障的冗余系统架构图。该方案由两套系统组成，在每套系统中，A4环节中兴ZXDU58 W121室外一体化通信电源机柜做到输入冗错，A5环节做到双回路互为备份，A6使用模块化UPS或者并机， A7为单电源负载提供双路保障，如果有条件A1和A2环节采用双路市电输入，单供电系统做到可靠冗余设计，然后方案采用2N容错设计，基本做到无单点故障点和在线维护。

图1 无单点故障的冗余系统架构图

提高UPS供电设备的可维护性

降低维护时间是提高可用性的另一重要途径。模块化设计可以有效改善易维护性，降低维护时间。UPS设备各个功能单元模块化之后，故障之后只需更换上相应备件即可，大幅降低了维护的技术门槛，运维人员可自行更换维护。不但维护成本可有效降低，故障修复时间也可大幅缩短，从而将业务损失降到。另外，模块化易于实现在线维护，即故障修复期间负载可以不断电。如果需要断电才能维护，就需要拉备用电源为负载供电，这样维护非常复杂，而且维护时间很长。

提高UPS供电设备的易用性

易用性是供电设备“可用性"的升华，直接影响用户的产品体验。从用户的角度看，需要从以下几个方面改善：①易搬运、易安装。这需要产品体积足够小，重量足够轻，并且是模块化可分解，从而降低搬运和安装的难度。此外UPS是否支持上下进线，是否支持并柜安装等都将影响安装的难度。②易扩容。数据中心一般都有未来的扩容计划，以匹配未来的业务增长需要。而现网的UPS供电设备为了确保可靠性通常供电路径非常复杂，牵一发而动全身，扩容非常不便，即使条件满足也有负载断电的重大风险。这样的供电现状显然是不易用的。如果能够像通信电源一样，功率模块可以热插拔，扩容只需采购功率模块在线插进去，那么扩容的易用性就可大幅改善。③易管理。UPS设备要高度智能化，各个供电节点做到可视化管理，便携化管理。比如，可以开发手机APP进行随身监控和管理。

UPS供电系统可用性发展的历程

代UPS——动态UPS。其利用机械惯性储能以及电动机、发电机的能量传输机制以提供短时间的不间断供电，体积庞大、造价昂贵、噪声巨大，犹如一个小型电厂。动态UPS的特征是占地面积较大，噪音大，不易维护和使用，接近一套工程设备。

第二代UPS——工频机。相比于动态UPS，其可用性提升主要体现在以下几个方面：，体积变小，搬运和安装难度降低；第二，备电时间可以由后备电池决定，从动态UPS的秒级备电上升到小时级；第三，可以对较差电网优化，如果一旦电网波动比较大，可以给后端设备提供相对稳定的电力供应。但是，工频UPS依然存在一些问题：，运输与安装问题。工频机因为体积庞大无法通过门和内置的升压用变压器重量太重无法使用电梯运输等，导致安装此类UPS经常要打墙安装、吊车运输；第二，维护问题，UPS主机类似黑盒设计，有任何故障或者异常都只能依托原厂家维修，运维人员不敢直接打开操作，时间响应慢，对业务影响大。

第三代UPS——高频机。高频机的出现进一步提升了功率密度，体积减小了50%，从功能模块上提升了维护性，缩短了MTTR时间，可在数小时内完成修复。重量较工频机进一步降低，有效提升了工程的可安装性。同时，高频机也大都采用了全数字化的高集成化设计，在维护性方面也有较大改进。THDi可以做到5%以下，明显减少电网的谐波污染，效率也进一步提升到92-96%，体现出其节能优势。但是，对设备可用性的追求探索并未停止：单点故障是否可以排除？故障修复时间是否可以缩短至分钟级？维护中兴ZXDU58 W121室外一体化通信电源机柜技术门槛可否降低至可以自行维护？

第四代UPS——模块化高频UPS。高频机技术的发展为UPS的模块化架构提供了技术可能，结合类似通信电源的模块冗余技术的供电架构，模块化的高频UPS得以实现。①可靠性大幅提高，常态工作的功率模块、控制模块实现全模块化冗余，消除单点故障点。②经济效益显著，模块化技术使得UPS效率上了一个新台阶，同时采用了通信电源成熟的智能休眠功能，让UPS系统始终处于效率点。③可维护性方面揭开了历史崭新的一页，维护技术门槛也大幅下降。对于单模块容量50KVA以下的小系统模块化UPS，采用模块热插拔技术运维人员可以自行在线维护和扩容，故障修复时间和扩容时间也缩短至分钟级，，对于单模块容量200KVA以上的模块化UPS，采用模块隔离技术，虽然重量较重无法热插拔，但运维人员可以自行在线分、合模块来维护和查找故障，大幅度缩短修复时间，同时剩余模块自行保证用户的容量可用性。④在安装、运输上也体现出了模块化的明显优势——各单元模块化可拆卸。模块化高频机UPS的功率密度比上一代产品更高，占地面积更小。据考证华为的小系统模块化UPS可以做到单柜300KVA以上。

结语中兴ZXDU58 W121室外一体化通信电源机柜

UPS供电设备的核心价值是保障高可用性，为用户提供高品质、易用的不间断供电从而确保业务的稳定运行，因此，对于数据中心UPS供电设备而言，我们需要转换设计理念，从可靠性的点向可用性的面演进。而模块化UPS相比传统UPS在可靠性、易维护性、易用性等各个方面均有优异的表现，可更有力地保障业务的连续性与稳定运行，更契合用户对于高可用供电的需求。

With the large-scale construction of the data center in recent years, the traditional exposure power supply system in large scale deployment and operational reliability, maintenance and other issues have become increasingly prominent, pushing users, equipment manufacturers and design companies ability to innovate and optimize the power supply system, power supply system construction ideas gradually focus from the traditional reliability to guarantee availability up. So why to build high availability power supply system and how to build high availability power supply system, this paper has made some discussion about this.

Availability reflects the real needs of the user, and reliability is one of the factors that affect availability.

Reliability is measured by reliability. The definition of reliability is "the probability that a given system can perform the specified function continuously under prescribed working conditions and predicted time". MTBF is also an important index to decide the reliability of power supply system. MTBF can be obtained by quantitative timing industrial test or theoretical calculation. Availability is the extent to which a product is in a working or usable state when it is required to start and execute at any random time. The formula of availability is:

In the formula, MTBF (Mean Time Between Failures) is an average fault interval time, and MTTR (Mean time to repair) is the average repair time.

中兴ZXDU58 W121室外一体化通信电源机柜

The reliability of the power system is a representative of the power system is easy to fail. But from a practical perspective, any device can not guarantee full of trouble in the life cycle of the equipment is the user wants to try not to fail, even if the fault is not caused by malfunction affected business; if the business is affected, so the fault should be eliminated as soon as possible. In contrast, the definition of availability is broader than that of reliability. For repairable systems, it covers not only the problem of device's error, but also whether the device is easy to recover from failure. It is obvious that availability reflects the user's needs more truly.

In the UPS industry, a few "9" are usually used to represent the availability of the system. It refers to the proportion of the system running online and the time that can be produced in one year. For example, the 6 "9" (availability can reach 99.9999%), namely every possible downtime is less than 32 seconds. The goal of the UPS system is to improve the availability of the UPS power system as far as possible and reduce the impact of the power from the city.

Ways to improve the availability of power supply

Improving the reliability of power supply system

From the calculation formula of availability, it can be seen that improving reliability is an important way to improve availability. Improving the reliability of power supply equipment is divided into four levels:

First, design the standard level. In the stage of product planning and design, the possible application environment of the product should be fully considered and the corresponding design standards should be selected. The electrical isolation, the use of the product EMI/EMC, lightning protection, anti surge, anti noise and electrical environment, moisture resistance, dustproof, shockproof, anti corrosion and natural environment, operation, maintenance, management, transportation and installation of the environment have been adequately assessed, so as to construct a reasonable framework for the design of products.

Second, device level. In the stage of product design, we must strictly select the device and cooperate with the best circuit design, and repeatedly simulate all kinds of harsh environments to test the device's stress margin, so as to ensure the reliable operation of all kinds of components. For key devices such as electrolytic capacitors, if the circuit design is not optimized enough, the ripple current is too large and the core temperature is too high, and the life will be greatly reduced, resulting in the reduction of device reliability. The fan also should choose the manufacturers with good stability and good performance to prevent the fan fault from rising the power module temperature and affecting the normal power supply.

Third, component level. The reliability of components is mainly reflected in its stability and redundancy. Under the premise of minimizing the failure rate of components, the most effective way to improve the availability of components is redundancy design of key components.

中兴ZXDU58 W121室外一体化通信电源机柜

Fourth, program level. By optimizing the system design, the power supply system is reliable and stable, and has the ability of fault tolerance. There is no single point of failure in the whole power supply path. Figure 1 shows a diagram of a redundant system architecture without a single point of failure. The scheme consists of two systems, in each system, A4 link to achieve input redundancy, A5 link to achieve double loop mutual backup, A6 or UPS and the use of modular machine, A7 provides dual protection for single power load, if the conditions of A1 and A2 link using dual mains input redundancy the design to achieve a single power supply system, and then the scheme uses 2N fault tolerant design, basically no single point of failure and online mainten中兴ZXDU58 W121室外一体化通信电源机柜ance.

Figure 1 redundant system architecture diagram without single point failure

Improve the maintainability of UPS power supply equipment

Reducing maintenance time is another important way to improve availability. Modular design can effectively improve the maintainability and reduce the maintenance time. After the modularization of each functional unit of UPS device, the replacement of the corresponding spare parts can only greatly reduce the technical threshold for maintenance, and the operation and maintenance personnel can replace and maintain themselves. Not only the maintenance cost can be effectively reduced, but the time of fault repair can be reduced greatly, so the business loss is reduced to a minimum. In addition, the modularization is easy to maintain online, that is, the load can not be cut off during the fault repair. If power is needed to maintain, it is necessary to pull the backup power supply to the load, which is very complicated and has a long maintenance time.

Improve the usability of UPS power supply equipment

Ease of use is the sublimation of the "availability" of power supply equipment, which directly affects the user's product experience. From the point of view of the user, it is necessary to improve the following aspects: 1. Easy to handle and easy to install. This requires that the product is small enough, the weight is light enough, and it is a mold.

近年来，随着互联网、云计算、移动互联网和物联网等技术的快速发展，数据中心系统规模不断增大，重要性越来越高，其对系统弹性、可用性、运营效率、可运维性等提出了更高的要求。作为数据中心供配电系统的关键组成部分，UPS无疑需要匹配这种要求。在此背景下，UPS模块化已经成为业界的共识。与传统塔式机相比较，模块化UPS具有以下优势：

1）投资有效性：随需扩容，节省初期投资；

2）模块冗余高可靠性：避免出现重大断电事故；

3）易维护性：在线热插拔，维护简单快速，无须转旁路；

4）节能环保性：对电网污染小，高效率及模块休眠等技术减少能源浪费。

正因为具有如此众多的优点，目前大多数UPS厂商都已发布模块化UPS，越来越多的用户已经或正在考虑使用模块化UPS建设新数据中心。但现今市场上的模块化UPS所采用的技术不尽相同，客户在选用过程中有一定的困惑，本文将基于笔者的应用实践与理解对两种主流架构的模块化UPS进行剖析，希望能给各位读者一些帮助及启发。

2. 模块化UPS的两种典型架构

1)分布式架构

图1中展示了分布式模块化UPS的系统架构。

中兴ZXDU58 W121室外一体化通信电源机柜

 
       图1 分布式结构的模块化UPS架构

分布式是早期模块化UPS经常使用的一种架构。此类模块化UPS系统层面上等价于数立的UPS直接并联，其功率模块利用小型UPS改造而成，可自主独立工作，其特点是：①除整流、逆变的控制外，均流与逻辑切换也由内部控制单元控制；②内置容量与功率模块容量一致的静态旁路，在旁路模式时，由每个模块内的静态旁路共同承担负载。

2)分布+集中式架构

与之相对应，图2展示了另一类架构的模块化UPS。

中兴ZXDU58 W121室外一体化通信电源机柜

 图2 分布+集中式结构模块化UPS架构

分布+集中式结构的模块化UPS设备所有的功率模块内置控制单元用于本模块的整流器与逆变器控制，而将整个系统的均流及逻辑切换等功能从模块内部控制单元中提取出来，由一个集中的控制模块控制。为了消除可能引入的单点故障，该控制模块及相应通讯总线均进行1+1冗余。当一个控制单元出现故障时，整个UPS系统中功率模块可由另一处于热备状态的控制单元无缝接管系统控制，保障系统不间断运行。同时，功率模块内不再内置静态旁路，系统配置一个静态旁路模块，其容量即为系统容量

3. 分散控制与分布+集中控制逻辑模式对比

分布式架构的模块化UPS采用分散控制逻辑模式，系统中每个模块都含有一个完整独立的控制单元，系统的主控模块会通过一定的逻辑规则从系统内所有模块中选出，其余模块作为从控模块听从主控模块调度。当UPS系统中的一个从控模块出现故障时其余模块仍正常工作，当主控模块出现故障时可通过一定的竞争规则来使得另一个模块作为主控模块，保障系统继续正常工作。

分散控制逻辑模式的优点在于每个控制单元都可以完成对系统独立控制的工作，故不存在这方面的单点故障点。但缺点也很明显，首先因为主控模块既要处理本身的信号，又要协调各模块之间的信号，所以控制逻辑比较复杂，软件逻辑可靠性不高。其次各主控模块故障后，会在剩余模块中竞争产生一个模块作为主控模块，该过程中也容易发生竞争失败导致系统故障。

分布+集中式架构的模块化UPS功率模块内整流、逆变的控制是分布的，而均流逻辑等控制则是集中控制模式，即采用独立集中的控制模块（如图2中控制模块）来检测市电的频率和相位，然后向每个模块发出同步信号，各个功率模块接受到此同步信号后通过自身的控制环输出相应频率相位的正弦波。当市电丢失时，集中控制模块会自激产生同步信号发送给各个UPS模块来保证各单元的输出同频同相。同时在均流的控制实现形式方面，集中式架构的模块化UPS依靠控制模块来检测整个系统的负载电流，然后除以系统模块数量来作为各个UPS模块的均流参考值，进而与各模块输出电流比较后求出偏差值来不断调整各模块的输出电流，以保证系统内模块间良好的均流度。分布+集中控制逻辑模式的优点在于采用独立的均流与逻辑控制单元，均流度更好，且控制逻辑层级清晰，各功率模块之间不存在竞争关系，软件逻辑可靠性较高。为了保证集中控制单元的可靠性，避免单点故障，一般采用该架构的UPS控制单元及通讯线路均会做1+1备份。1+1热备份是的备份方式，其可靠性在各类系统长期运行实践中已得到验证。
综合来说，集中式冗余架构具有的优势是明显的。

4. 集中旁路与分散旁路对比

正如本文中两种架构图所示，目前大容量模块化UPS系统的旁路控制技术主要有两种模式：1、系统集中旁路模式（UPS系统内只有一套旁路系统，如图2所示）；2、系统分散旁路模式（UPS系统内每个功率模块都有一套旁路系统，如图1所示）。集中旁路系统具有过载能力强，可靠性高的优点，而分散旁路具有可扩容，成本低的优点，但可能存在一定的可靠性风险。

对于分散旁路模式，表面上看因分散布置，在UPS模块冗余时类似于冗余设计，一处旁路故障，其它旁路仍可工作。实际上此种分散与冗余有本质不同。旁路的主要器件为SCR。因为器件的离散性较大，系统工作在旁路模式时，各个旁路基本不可能处于均流状态；而为了保持旁路输出的电压波形完整，在旁路模式时不会进行开关动作，难以电流进行控制，仅依赖自然均流不均流度很难控制在25%以内，电流大的模块很可能因旁路过载而关机，影响系统供电连续性。

除了稳态的均流问题，在瞬态时分散旁路系统也具有一定的风险。在系统控制器发送切换旁路模式的信号之后，因为信号传输路径、模块控制器响应速度、器件一致性等各方面原因，各个旁路很难同步切换，而先切换导通的SCR将承担大部分负载甚至所有负载，极易导致该SCR失效。

静态旁路是主路模式的冗余，作用非常重要。而分散旁路的设计方式大大降低了旁路的可靠性。实际上，在传统塔式UPS应用中当并机数超过四台时，一般为了避免旁路不均流问题，都需要采用集中静态旁路系统。因为旁路系统的限制，采用分散旁路系统的UPS很难具有较好可扩展性。

5. 总结

如上所述，模块化UPS因其高可靠、易维护、易扩容等优点，大大地节省了客户运营维护成本，为业务的长期稳定运行提供了保障。两种典型架构的模块化UPS都能提供较好的维护性与扩容能力，比起传统UPS的可用性大幅提升。但从技术角度分析，集中式结构的模块化UPS具有更高的安全性，更优异的可靠性。

In recent years, with the rapid development of technology such as Internet, cloud computing, Internet and Internet of things, the size of data center system is increasing, and its importance is higher and higher. It puts forward higher requirements for system flexibility, availability, operation efficiency and operation and maintenance. As a key component of a data center for distribution systems, UPS is undoubtedly required to match this requirement. In this context, UPS modularization has become a consensus in the industry. Compared with the traditional tower machine, the modular UPS has the following advant中兴ZXDU58 W121室外一体化通信电源机柜ages:

1) investment effectiveness: to expａnd the capacity with the need to save the initial investment;

2) high reliability of module redundancy: avoid major power failure;

3) easy maintenance: online hot plug, simple and fast maintenance, no bypass;

4) energy conservation and environmental protection: small pollution of the power grid, high efficiency and module dormancy technology to reduce energy waste.

Because of so many advantages, most UPS manufacturers have released modular UPS at present. More and more users are considering the use of modular UPS to build new data centers. But the UPS module on the market today the technology is not the same, customers have a certain confusion in the selection process, the practice and understanding of the author based on the modularization of the two mainstream architectures of UPS are analyzed, the hope can give you some inspiration and help readers.

The two typical architecture of 2. modularized UPS

1) distributed architecture

The system architecture of distributed modular UPS is shown in Figure 1.

中兴ZXDU58 W121室外一体化通信电源机柜

Graph 1 modular UPS architecture of distributed structure

Distributed is an architecture that is often used by early modular UPS. This kind of modular equivalent UPS system level on the number of independent UPS directly parallel, the power module and the use of small UPS transformation, can work independently, its characteristics are: the control of rectifier, inverter in addition, uniform flow and also by the internal logic switching control unit control; the internal capacity and static bypass the power module capacity is consistent, in bypass mode, the static bypass within each module shared load.

2) distributed + centralized architecture

In contrast, figure 2 shows the modularized UPS of another type of architecture.

Graph 2 distribution + centralized structure modularized UPS architecture

分布+集中式结构的模块化UPS设备所有的功率模块内置控制单元用于本模块的整流器与逆变器控制，而将整个系统的均流及逻辑切换等功能从模块内部控制单元中提取出来，由一个集中的控制模块控制。 In order to eliminate the single point of fault that may be introduced, the control module and the corresponding communication bus all carry out 1+1 redundancy. When a control unit fails, the power module in the whole UPS system can be controlled by another control unit which is in hot standby state, which ensures seamless operation of the system. At the same time, the static bypass is not built in the power module, and the system is equipped with a static bypass module, which is the capacity of the system.

3. distribution control and distribution + centralized control logic model comparison

UPS modular distributed architecture with distributed control logic model, each module of the system contains a complete and independent control unit, main control module of the system will be selected from the system in all modules through certain logic rules, the other from the control module to the module as the main control module scheduling. When one of the UPS modules fails, the other modules still work normally. When the main control module fails, a certain rule of competition can be used to make the other module as the main control module, so that the system will continue to work normally.

The advantage of the decentralized control logic mode is that each control unit can complete the work of independent control of the system, so there is no single point of fault in this respect. But the disadvantages are also obvious. First, because the main control module not only processes its own signal, but also coordinates the signals between modules, so the control logic is more complex and the software logic reliability is not high. Secondly, after each main control module fails, a module will be generated in the remaining modules as the main control module, which is also prone to competition failure, resulting in system failure.

Control + distribution centralized architecture modular UPS power module, inverter rectifier is distributed, and the flow control logic is centralized control mode, which adopts the independent centralized control module (Figure 2 control module) to the frequency and phase detection of the power, then sends a synchronization signal to each module. Each power module receive the synchronous signal through its control loop output the corresponding frequency sine wave phase. When the electricity is lost, the centralized control module will generate the synchronization signal to each UPS module to ensure the same phase of the output of each unit. At the same time form in controlling flow, UPS modular centralized architecture relies on control module to detect the load current of the whole system, and then divided by the number of modules of the system as a flow of reference value of each UPS module, and then compared with the output current of each module after the obtained partial difference to continuously adjust the output current of each module. In order to ensure the system modules are good mobility. The advantage of the distributed + centralized control logic mode is that the independent flow and logic control unit is adopted. The flow rate is better, and the control logic level is clear. There is no competition relationship among the power modules, and the software logic reliability is relatively high. In order to ensure the reliability of the centralized control unit and avoid a single point of fault, the UPS control unit and the communication line of this architecture will do 1+1 backup. 1+1 hot backup is the most commonly used backup method, and its reliability has been verified in the long run practice of various systems.

Integrated

目前UPS产品在行业应用已有五十余年的历史，其为保障关键设备和业务的不间断运行做出了的贡献。随着信息化建设的不断推进，需要UPS保护的场景越来越多，其作用愈发重要。当前市场上存在工频机、高频塔式机、高频模块化UPS三类产品，其利弊优劣众说纷纭，令用户感到十分困惑。本文旨在通过阐述UPS的发展历史及对比各类UPS的优劣势，帮助用户识别UPS产品发展的趋势所在。

一、从工频机UPS到高频塔式机UPS的发展

工频机结构UPS技术出现在上世纪70年代，因其整流工作频率与电网频率一致而得名。受制于当时半导体技术发展，逆变器中IGBT器件耐压只能做到600V，故母线电压受限，逆变器输出电压不能做到380V；而且工频机逆变器是全桥电路，输出为三相火线，无法满足单相IT负载和三相四线制负载的需求，必须进行Δ-Y转换。为解决这些问题，厂家在工频机逆变器输出端加入了变压器用于升压和产生中线，以使输出电压满足负载的要求，这便是工频机内置变压器的真实目的。图-1所示为工频机的典型拓扑。

中兴ZXDU58 W121室外一体化通信电源机柜

图-1 工频机典型拓扑

而到上世纪90年代，第三代沟槽型IGBT面世，其耐压能力提升至1200V，促使了UPS技术的革新。通过整流侧高频升压电路将母线电压提升至700V左右，逆变器输出电压可以做到380V，输出变压器得以取消。而这种整流逆变电路都工作在高频（几kHz以上）且没有输出变压器的UPS就被称为高频UPS。图-2所示为一典型的高频机拓扑。

图-2 高频机典型拓扑中兴ZXDU58 W121室外一体化通信电源机柜

二、高频UPS与工频UPS的对比

1.工频机输入功率因数低、谐波高

工频机UPS采用可控硅半控整流，6脉冲整流UPS输入功率因数低于0.7，谐波高达30%；12脉冲整流UPS输入功率因数仅为0.8，谐波高达15%，即使加上谐波处理措施，功率因数也只能改善至0.95。相比之下，高频机采用IGBT-PFC全控整流，输入功率因数业界均可做到0.99，谐波电流小于3%。严重的谐波污染不仅可能干扰其他设备无法工作、使控制与保护器件误动作外，而且直接导致投资大幅增加：客户需要购买额外的谐波处理设备降低谐波；如果前端接柴油发电机备电，发电机的容量要配置为UPS容量的2-3倍，同时前级配电器件、线缆等均需要提升20%左右，而高频机只需前端发电机容量配置为UPS容量的1.2-1.5倍即可，配电容量和UPS容量保持一致或略高。

2.工频机功耗大

有三个因素导致工频UPS效率低于高频UPS。一是工频UPS整流为降压拓扑，器件工作电流大，无论是内部线路无论是线性损耗还是平方损耗都比高频机高；二是因输出需要升压的原因工频机比高频机多内置一个输出变压器，致使工频机效率下降2%-3%左右；三是在实际应用中，为了提高输入功率因数至0.95以上，并降低其注入电网的谐波污染，工频机还要外置一个5次或11次谐波滤波器，效率将再次下降2%-3%。据英国某运营商与西班牙某运营商现网运行统计数据，工频UPS的效率一般在85%左右，相比高频92%左右的运行效率和模块化96%左右的运行效率，导致大量的能量损失。以400kW负载为例，工频机将比高频机年多耗电41万度，比模块化年多耗电近58万度。除此之外，工频UPS还有高谐波、低功率因数等导致配电线缆损耗增大等问题。

3.工频机体积大、重量重

因为工频机采用低频器件且配置输出变压器，致使UPS体积重量大大增加。以某品牌400kVA工频机和高频机对比，工频机重量是高频机的2.2倍，体积是高频机的1.5倍，在实际运输中可能存在机房门或者走道偏小、电梯载重不够、楼层承重不足等问题，有些情况下甚至需要用吊车装卸，然后破墙而入来安装工频UPS，大大增加了运输时间及成本。

4.工频机相比高频机在可靠性方面并无优势

工频机和高频机的主要差异体现在整流器和变压器上。工频机整流器采用SCR器件，电压应力小，电流应力大，高频机主要采用IGBT器件，电流应力小，电压应力大。SCR与IGBT目前均为成熟器件，只要应用得当，可靠性并不会有差异。事实上，工频机的逆变部分也是使用IGBT，并没有因此而降低工频机的可靠性，也没有证据证明逆变器是工频机的薄弱环节。从拓扑上讲，工频机用的是相控整流+全桥逆变，高频机一般采用高频整流+半桥逆变。这些拓扑均为电力电子技术上非常常用的拓扑，并不存在谁原理上更可靠的问题，其可靠度取决于设计的水平。

而对于变压器，业界经常可以听到其很多所谓的优点，比如抗冲击能力强、降低零地电压等，然而真的是这样吗？

，过载能力强，抗负载冲击能力强。过载能力是IEC62040-3中要求标称的关键指标之一，其强弱可通过实际数据来衡量。表-1所示为同一厂商的工频机与高频机过载能力，由表-1可知，两类机型过载能力并没有区别。

表-1 某厂商工频机与高频机过载能力对比

输出变压器并不会增强工频机的抗冲击能力，中兴ZXDU58 W121室外一体化通信电源机柜对于变压器可以增强抗冲击能力的想象来源于变压器的电感特性，电感平滑电流的能力在负载电流激增时可以平滑电流波形延缓电流冲击。但实际上电感平滑电流的能力与其本身感量成正比。工频机输出变压器变比小，变压器输出绕组的励磁电感也不会太大，在大电流冲击下极易饱和，很难对逆变器的冲击有明显的缓冲作用。而按照传统变压器传递能量的特点与磁性器件原理分析，当后级负载也就是变压器输出侧出现能量冲击时，在变压器能量传递能力达到饱和上限之前，后端的尖峰励磁电流会直接反射到前端对UPS的IGBT产生冲击，并且由于变压器的变比问题前端所受到的冲击电流会比输出端更大，同时造成的损害也更为严重。而且，工频系统由于变压器的磁滞特性，难以实时监测后级动态响应。当变压器后端出现突变并反馈到前级时，系统采取相关动作较无变压器的高频机来说会延迟几十甚至几百个ms，此时流过IGBT的冲击电流已经足够损坏UPS甚至引发火灾。

第二，在逆变器IGBT管直通故障时隔断直流危险电压。工频机变压器确实可以避免直流传递至副边，但高频机通过快速检测与保护措施一样可以避免直流危险电压对负载造成危害。当高频机逆变某IGBT出现直通故障时，UPS控制器可立即检测输出电流异常，并通过整流单元关机及输出端口熔丝保护等措施快速隔断直流危险电压到输出端口的路径。在保护过程中，输出到负载端口的电压约为持续几个ms的400V直流。对于使用开关电源供电的IT负载来说，其输入允许电压可以达到276Vac，整流之后电压也在400Vdc左右，器件选型等均依据母线电压选型。此时输入端口的400Vdc不会超出器件耐受范围，不可能对设备造成伤害。而对于工频机而言，其原边加载直流电压，将导致电流急剧增大，温度快速上升，可能引发火灾等更严重故障。

第三，可以降低零地电压。许多服务器等设备都有零地电压的要求，尽管这样设计的原因已无法考证，因为从理论上来说零地电压的大小并不会影响IT设备的正常工作。在数据中心中，IT设备只允许使用TN-S或TN-C-S供电制式，那么IT设备输入端口的零地电压主要由零线接地点(TN-S系统)或零线与地线分离点(TN-C-S系统)至IT输入端口的零线阻抗与零线电流及系统中三次谐波电流决定。在相同的系统中，无论是工频机还是高频机均不会影响零线阻抗，而零线电流及三次谐波电流主要是与三相负载配置与负载特性有关，即UPS的类型不会对于零地电压不会有明显的影响。真正决定零地电压的是配电系统的设计。如果需要改善零地电压，是从配电系统入手，着手减少线路阻抗与零线电流。减少线路阻抗的方式即在负载的列头柜内置隔离变压器。需要注意的是在应用时有将工频机变压器副边直接接地的做法，这是一种不规范的做法。工频机变压器N线并未隔离，对于TN-S系统和N与PE已经分开的TN-C-S系统，N线重新接地也将导致PE线有电流流过，可能干扰设备正常工作。国标还是IEC标准均不允许此种不规范做法。

而第四，工频UPS的变压器可以起到隔离作用，可以保障人身安全。为了保障主旁平稳切换，工频UPS输出N线由旁路引入，也即工频机的变压器并不能起到电气隔离作用，也不能重新接地。在需要隔离场合的场景，即使使用工频UPS，其旁路也必须加一变压器用于隔离N线，以实现真正的隔离。

实际上，变压器的设计反而增大了环流的风险。图-3所示为两类机型的环流路径。工频机UPS的并联就是变压器的直接并联，整条回路上没有器件限制，电压的偏差很容易产生环流。而高频机的环流路径上具备多个二极管，小于2V的电压差根本形不成环流。

图-3 工频机与高频机并机环流路径

5.工频机增加用户投资

由于工频机整流工作在市电频率，需要更大的电感储能。其更大体积的电感与无法省掉的变压器均由铜和磁性材料组成，成本难以下降，价格一般比高频机要高30%以上。

中兴ZXDU58 W121室外一体化通信电源机柜

综上，从性能、可靠性、价格上讲，高频机比工频机均具备优势。从各主要厂家的系列来看，业界主要厂商均已不推出新工频机型，部分厂商已全面转向高频机的研发与销售。工频机被高频机取代已是大势所趋。

三、从高频塔式机UPS到模块化UPS的发展

模块化UPS早在上世纪九十年代即已出现，但因为技术能力沉寂了很长时间。而自2000年起，由于DSP、数字控制等技术的发展，多功率模块并联均流控制问题得以逐步解决，模块化UPS技术开始蓬勃发展。2009-2010年中国电信对模块化UPS展开深入测试，根据各地实际使用单位的反馈，中国电信认为业界主流模块化UPS已满足通信行业的使用要求，并于2011年底开始对模块化UPS进行集中采购。中国移动模块化UPS也以单独标段进行集采。

四、模块化UPS与高频塔式UPS的对比

1.模块化UPS系统可用性高

供配电系统作为现在信息系统极为重要的一环，对其一个基本的要求就是该系统必须能连续工作。而要达到连续工作这一目的，首先是系统应具备较高的可靠性，其次该系统必须做到能够快速修复。如果不能快速修复，就可能面临二次故障导致整个系统瘫痪的风险，客户的负载就不能保障连续工作。

在快速修复方面，模块化UPS具备天生优势。首先，在修复时间上，由于快速插拔这一特性，模块化UPS现场即可完成更换，平均的修复时间在半小时之内，相比于传统塔式机典型修复时间24小时，修复速度明显提升。其次，在修复质量上，模块化UPS的修复形式是将故障模块更换，而传统塔式机需要原厂派专业工程师到现场进行故障定位，然后拆机修复故障电路、单板，修复周期长，而且存在沟通和定位过程，易造成重复工作，影响故障处理效率。

可能有的用户会质疑，认为模块化UPS的N+1体系结构不如1+1并机系统稳定。确实，从理论上来讲，N+1并机系统中1+1的可靠性肯定是的。但是实际的场景中往往不是这么简单：

首先，此结论忽略了负载率这一情况，作为1+1并机系统，最多只能允许一台UPS损坏；而对于模块化UPS体系，以4+1为例，99%负载的时候可靠性要低于1+1，但是75%负载率的时候，模块化体系实际就变成了3+2，50%的时候就变成了2+3，可靠性要远大于1+1并机。在常见应用场景中，UPS负载率是在20~40%左右的，在这种情况下模块化的优势具有非常明显的优势。

其次，不同于传统单机，模块化UPS可以轻易实现N+2、N+3这种冗余模式，仅需增加1-2个模块即可实现，而塔式机要做到此模式不仅仅是增加1台主机，机器运输、场地安装、走线设计以及相应的配电、电池都需变更，导致投资大幅增加。

综上，UPS模块化在实际场景中可靠性远高于传统塔式并机；再加上UPS快速维护、扩容的特性，模块化UPS的可用性更是大大高于传统塔式机。

2.模块化UPS的扩展性更好

塔式机扩容需要购买整台新机、将机器安装到位、将系统中其他UPS转旁路后把新机接入系统，整个步骤中不仅投资高、安装时间长，而且在并入新机时由于整个系统处于旁路状态，存在市电中断导致负载掉电的风险。

而模块化只要初期规划好配电系统，就可以通过增加模块来匹配负载的提升，且在扩容过程中保障对原有负载的不间断供电。

3.模块化UPS运输安装难度低

塔式机UPS需要作为一个整体来安装、运输，大型单机就会比较困难。如容量400kVA的UPS重量一般为1500kg左右，体积超过3m3，塔式机UPS会受到运输通道不足、重量高难运输的困难，而模块化UPS一方面可以将模块、机架分开搬运，另一方面多数机型机架之间可以分开运输，塔式UPS可能遇到的问题将迎刃而解。

4.模块化UPS实际运行效率高

目前高频塔式UPS与模块化UPS均可做到96%的效率值，但这是在负载率在50%以上才能达到的。而前面提到，因为系统冗余及超前规划，常见工况下UPS负载率在20~40%左右。高频塔式机在此工况下只能做到94~95%的效率，而主流模块化UPS普遍具备“模块休眠"特性在保证一定系统冗余的基础上，可以休眠一定数量的模块（可以手动或者设置自动），让UPS系统工作在效率比较高的区域，即保持在高点96%附近。图-4即展示了休眠提升负载率中兴ZXDU58 W121室外一体化通信电源机柜与运行效率的原理。

图-4 休眠可有效提升UPS负载率与运行效率

而且有些厂家考虑到模块老化时间可能不同，更进一步开发了“轮换休眠功能"：即每隔一段设定好的周期，休眠模块进行轮换，以平均每个模块的老化时间，提升整体UPS系统寿命。图-5展示了轮换休眠的典型过程。

图-5 轮换休眠技术

五、结束语

自其诞生之日起，模块化UPS就旨在满足用户对于供电系统的可用性、可靠性、可维护性及节能等方面的需求。经过长期的运行验证，模块化UPS在这些方面相较传统UPS系统确实具备很大优势。随着能源成本持续增加及用户对供电系统的灵活性、可用性等要求的进一步提高，模块化UPS必将得到更广泛的应用。

At present, UPS products have been used in the industry for more than fifty years. It has made outstanding contributions to ensure the uninterrupted operation of key equipment and business. With the continuous advancement of information construction, more and more scenes of UPS protection are needed, and its role is becoming more and more important. There are three kinds of products in the current market, such as power frequency machine, high frequency tower type machine and high frequency modular UPS. The advantages and disadvantages of these products are different, which makes the users feel very confused. This article aims to help users identify the trend of the development of UPS products by explaining the history of the development of UPS and comparing the advantages and disadvantages of various kinds of UPS.

1. The development of UPS from UPS to high frequency tower machine

The power frequency machine structure UPS technology appeared in the 70s of last century, because the frequency of its rectification work is consistent with the frequency of the power grid. Subject to the development of semiconductor technology, the inverter IGBT breakdown voltage can only 600V, so the bus voltage is limited, the output voltage of the inverter can achieve 380V; and motor inverter is the full bridge circuit, output voltage line, unable to meet the load demａnd of IT single-phase four wire and three-phase load, must be a -Y conversion. In order to solve these problems, the transformer is added to the output terminal of the power frequency inverter, which is used to boost and generate the midline, so that the output voltage can meet the load requirements, which is the real purpose of the transformer built in the power frequency machine. The typical topology of the industrial frequency machine is shown in figure -1.

Figure -1 typical topology of power frequency machine

In the 90s of the last century, the third generation of trench type IGBT came out, and its pressure resistance increased to 1200V, prompting the innovation of UPS technology. Through the high frequency boost circuit of the rectifier side, the bus voltage is raised to about 700V, the output voltage of the inverter can be 380V, and the output transformer can be cancelled. This rectifier inverter circuit works at high frequency (a few kHz above) and UPS without output transformer is called high frequency UPS. Figure -2 is shown as a typical high-frequency machine topology.

Figure -2 typical topology of high frequency machine

Comparison of two, high frequency UPS and power frequency UPS

The input power factor of 1. power frequency machine is low and the harmonic is high

中兴ZXDU58 W121室外一体化通信电源机柜

The power machine UPS using thyristor half controlled rectifier, 6 pulse rectifier UPS input power factor is less than 0.7, up to 30% harmonics; 12 pulse rectifier UPS maximum input power factor is only 0.8, harmonic as high as 15%, even with the harmonic treatment measures, the highest power factor can only be improved to 0.95. In contrast, the high frequency machine adopts the IGBT-PFC full control rectifier, and the input power factor industry can achieve 0.99 and the harmonic current is less than 3%. Serious harmonic pollution not only may interfere with other equipment can not work and the malfunction of control and protection devices, but also directly lead to a substantial increase in investment: customers need to purchase additional equipment to reduce the harmonic harmonic treatment; if the front end is connected with a diesel generator power generator, the capacity to be configured as 2-3 times the capacity of UPS, at the same time before distribution device cable, etc. need to increase by 20%, while the high frequency machine just in front of the generator capacity configuration for the 1.2-1.5 times UPS capacity, distribution capacity and UPS capacity is the same as or slightly high.

2. power frequency machine power consumption is big

There are three factors that cause the efficiency of the power frequency UPS to be lower than the high frequency UPS. One is the frequency UPS rectifier for buck topology, device working current, both internal lines of both the linear loss or square loss than high frequency machine; two is the reason for the need to boost the power output than the high-frequency machine with an output transformer, the power machine efficiency decreased about 2%-3%; three is in the practical application in order to improve the input power factor, to more than 0.95, and decrease the injection of harmonic pollution, but also a power machine external 5 or 11 harmonic filter, the efficiency will drop again 2%-3%. According to the data from a British operator and a Spanish operator, the efficiency of power frequency UPS is generally around 85%, which is much higher than the operation efficiency of 92% and the efficiency of modularization is about 96%, resulting in a large number of energy losses. Taking the 400kW load as an example, the power frequency machine will consume more than 410 thousａnd degrees more than the high frequency machine, and will consume more than 580 thousａnd degrees more than the modular year. In addition, the power frequency UPS and the high harmonic, low power factor and so on lead to the increase of distribution cable loss and so on.

3. power frequency machine is large in volume and weight weight

Because the frequency machine uses low frequency devices and the output transformer is configured, the volume weight of UPS is greatly increased. In contrast to a certain brａnd of 400kVA frequency machine and high frequency machine, power machine is 2.2 times heavier than the high frequency machine, high frequency machine volume is 1.5 times, maybe the problem is too small, the room door or aisle elevator load is not enough, the lack of floor bearing in actual transportation, even in some cases need 中兴ZXDU58 W121室外一体化通信电源机柜to use the crane loading and unloading, and then break through the wall to install in the frequency UPS, greatly increasing the transportation time and cost.

4. frequency machines have no advantages in reliability compared to high frequency machines

The main differences between the frequency machine and the high frequency machine are reflected in the rectifier and the transformer. The power frequency machine rectifier uses SCR device, the voltage stress is small, the current stress is big, the high frequency machine mainly adopts the IGBT device, the current stress is small, and the voltage stress is big. SCR and IGBT are all mature devices at present. As long as they are properly applied, there is no difference in reliability. In fact, the inverter part of the power frequency machine also uses IGBT, which does not reduce the reliability of the power frequency machine. There is no evidence that the inverter is the weak link of the power frequency machine. From the topological point of view, the frequency machine uses phase controlled rectifier + full bridge inverter, high frequency machine ge中兴ZXDU58 W121室外一体化通信电源机柜nerally adopts high frequency rectifier + half bridge inverter. These topologies are all very common topologies in power electronics, and there is no problem that is more reliable in principle, and its reliability depends on the level of the design.

For transformers, the industry can often hear a lot of its so-called advantages, such as strong impact resistance, reduction.

中兴ZXDU58 W121室外一体化通信电源机柜

中兴ZXDU58 W121室外一体化通信电源机柜