Tuesday, September 18, 2007

A “Beginner’s Guide” to Fault Tolerant Power Supplies

The effectiveness of having a fault tolerant power strategy was demonstrated after hurricane Katrina hit the Gulf Coast in 2005. A financial news television station interviewed the heads of two telecom carriers to find out when their telephone services would be operational again. The interview was very short – “we never lost service” they replied.

The telephone systems we take for granted have expensive and complex back up systems. Fault tolerant power supplies are supported by battery banks, generators and uninterruptible power supplies. Large Industrial complexes have also implemented similar systems - having an oil refinery stop production can result in enormous sums of money being lost!

For those with less extensive budgets, this brief article will explain the benefits, terminology and tips on how to implement a relatively low cost, but effective system.

Why have redundant power supplies?
Imagine a 24VDC 10A power supply driving motors and sensors on a conveyor based production line. For two or three years everything works fine, then one Friday (always at the end of the month), the power supply fails causing the conveyor to stop. Even if a spare part is in stock, it could still result in 30 minutes of expensive lost production.

If two identical power supplies had been installed in a fault tolerant, redundant mode, the remaining (good) unit would have continued to power the production line. The failed power supply could then be replaced at a more convenient time during routine maintenance.

Frequently Used Terminology

N+1
An expression where N is the number of power supplies needed to run the system. The simple two power supply system mentioned above would be considered 1+1. A triple redundant system (where two failures would have to occur to shut the system down) would be designated 1+2.

Hot-swap
Some equipment is operated 24 hours a day, 7 days a week, allowing no time to bring the system down for maintenance. In this case the failed power supply must be “swapped” out and a new one inserted without disruption to equipment operation.

ORing diodes
In the rare event of a power supply failing with a shorted output, low voltage-drop ORing diodes block that short from bringing down the system power.

Current share
Some power systems employ a method of balancing the current between the power supplies to increase field life. This can be an electronic signal wire that links the power supplies together or a switch* on the power supply that initiates a slight drop in the output voltage as more current is drawn. (*Common on high power DIN rail units)

Two Ways of Implementing Fault Tolerance

DIN Rail mount
For the example listed above, the simplest off-the-shelf solution is to use a diode “ORing” module and two power supplies. Here we are using Lambda’s DIN rail mount DLP-PU module and two 24V 10A DLP240-24-1/E power supplies.


Tip: When wiring the system, ensure that the cable lengths from the output of the power supplies to the ORing module are equal. This will help optimize the performance and life of the power supplies.
Inside the diode ORing module are two diodes and two alarm relays. Even in the event of one power supply failing with an internal short circuit, the remaining unit will continue to deliver power. See below.


Tip: - It is important to identify power supply failure using the relay alarms to flag the need for maintenance. Engineers sometimes overlook this which can result in a second failure unexpectedly bringing the system down!

Rack Mount
System Engineers requiring more power are turning to the communications style racks. These sophisticated low cost systems allow power supplies to be hot-swapped and come completely self contained. An example of such a product is Lambda’s FPS series.
Advantages of this solution include:
  • Easy mounting into a standard 19” rack
  • All in one solution
  • Hotswap capable (ORing diodes or MOSFET switches built-in)
  • No tools are required for replacement of a supply
  • High density, low profile (1.75”)
  • Off the shelf parts
  • Fully safety approved
  • All necessary warning signals included
  • 12V, 24V, 32-36V and 48V outputs

Click on http://www.lambdapower.com/products/fps-series.htm for an animated example.

Finally, one important note
A company wanted to ensure that in the event of a power supply failure their system would continue to operate. A battery was installed across the power supply output to give 24 hours uptime in the event of a power supply failure.

Unfortunately no thought was given to how anyone would know that the system needed maintenance! The power supply did eventually fail and the battery kept the system up for 24 hours before it discharged resulting in a system shutdown. A simple alarm circuit could have prevented that.

If you take Lambda’s recommendation to invest a little extra money up front to make your power system more secure, test your system to make sure you have it right!

Friday, September 7, 2007

Advances in Power Supplies for Automated Electrochemical Mini-Plants

On-site and on-demand production of disinfectants, biocides and water purification chemicals including sodium hypochlorite, chlorine dioxide has been substantially improved via the use of advanced switchmode power supplies that provide the power for automated electrochemical generators.

Many municipal water, food processing, and wastewater treatment plants are switching over from the use of chlorine disinfectants and biocides to safer and more environmentally friendly point-of-use and on-demand generated chemicals. The primary reasons for this change are that conventional chlorine agents require transport by tankers on accident-prone highways or railroads, ever increasing safety and environmental regulations regarding toxic gases and chemical spills, and the required bulk storage of these hazardous materials at the sites where they are used. Safer and in many cases more effective chemicals have been developed that can replace chlorine. For example, after many trial and error attempts to find a way to effectively control Legionnaires’ disease, it was found that chlorine dioxide (CIO2) was one the few chemical agents that could consistently and safely disinfect Legionella bacteria (see References). Add to this the ability to manufacture these safer chemicals at the locations that use them, and only when needed, and the advantages in total become obvious.

Two popular substitute chemicals for chlorine are sodium hypochlorite (NaOCI) and chlorine dioxide (CIO2) both of which can be manufactured via mini-plants (aka, generators) that are delivered to the end users’ site as a complete package and provide the disinfectants on-demand and as needed. In many cases, these mini-plants operate automatically and can be employed in unmanned locations such as municipal water treatment sites.

These electrochemical generators use the process of electrolysis as the basis for the production of these disinfecting and biocide chemicals. Recalling our science classes, electrolysis is a common method of separating bonded elements and compounds by passing an electric current through them. It involves applying a voltage between two electrodes (anode and cathode) which are submerged in a conductive solution (electrolyte). When a voltage is applied to the electrodes, electric current flows and in turn breaks down the molecules within the solution into its components (Figure 1).


Figure 1 shows part of the process that is used to produce sodium hypochlorite (NaOCI), which is more commonly known as household bleach when sold as a solution containing 5-6% of NaOCI. However, instead of a static vessel as shown in Figure 1, modern electrochemical generators pump the electrolyte solution continuously through one or more tubes that have the electrodes mounted within them. As the electrolytic solution flows through these tubes (electrolytic cells), the electrolysis process continuously separates the molecular components. In some instances, the solution is run through the dual-electrode electrolytic cells more than once to further refine and separate the resulting chemicals. (Note: Batteries operate by a reverse process from electrolysis; they generate electricity by means of galvanic or voltaic cells that contain anode and cathode electrodes that are in contact with an electrolyte solution or gel.)

Historically, the power supplies that provide the driving force for electrochemical generators have evolved from basic transformer and diode rectifiers, to transformer and SCR (silicon-controlled-rectifier) power sources, to modern and more sophisticated power sources. The development of the switchmode power supply greatly reduced the size and substantially improved the efficiency of these power sources. In addition, switchmode power supplies have the ability to provide electronic signals for status information (volts, amps, temperature, etc.), remote control, and communications to and from a PLC (Programmable Logic Controller) or a local/remote computerized controller.

The vast majority of switchmode power supplies are designed to operate as regulated voltage power sources. These supplies regulate the output voltage very precisely regardless of the amount of current drawn from the supply, up to its design limit. For example, a 1500-watt supply can provide a 12-volts output while providing from 0 to 125 amperes of current. Once the maximum current of 125-amps is reached, the supply is designed to go into a current-limit mode (where the output voltage is automatically reduced or the supply shuts down).


Figure 2 above shows the loop diagram of a typical sodium hypochlorite generator. The part of the system shown above that is called the Electrolyzer consists of multiple electrolytic cells (tubes containing electrodes), connected in series, through which the electrolytic solution is pumped and in turn separated into its primary chemical components (e.g., sodium hypochlorite solution and hydrogen gas) via electrolysis. The electrochemical process for manufacturing chlorine dioxide is similar to the above except it starts with a solution of sodium chlorite.

It has been found that in many electrochemical processes, including the production of disinfecting agents, that standard voltage-regulated power supplies do not always provide the ideal power profile for these processes. In fact, in many instances, the power supplies are being forced to operate at a fixed voltage and at close to their maximum current rating. If these operating conditions are maintained for long periods of time, the supply will internally heat-up and prematurely fail, thus shutting down the production of the disinfecting agents.

Why does this happen and how can it be avoided? As described above, during the electrochemical process, in order to keep up with the continuous electrolysis process with constantly flowing electrolyte solutions, the power supply must provide a high enough voltage to overcome the impedance between the two electrodes and the solution surrounding them, and, more importantly, to provide a high enough current density (amperes) to effectively separate the molecules during the short time (determined by the flow rate) that the solution comes in contact with the electrodes. By using a switchmode power supply that is designed to operate in a “constant-current” mode (instead of constant-voltage, as is the norm) the electrochemical process has been found to produce chemicals much faster, with consistent high quality, without forcing the power supply into an overload state.

There are a number of ways of providing current-mode power supplies for enhanced electrochemical applications. One method is to use Programmable Power Supplies. These supplies are designed to be manually or remotely programmed to operate in a voltage-mode and/or a current mode, at a specific voltage and current range, along with other specified parameters. As an added bonus these supplies usually include a serial digital communications port that allows it “talk” to local or remote computer controllers. Additionally, these supplies can be connected in parallel to the electrodes, or to groups of electrodes, when an electrochemical process requires more current than one supply can provide. For example, Lambda Americas’ model ZUP10-80/U programmable power supplies is adjustable from 0 to 10-volts with 0 to 80-amps (800 watts total). This type of supply is being used in its “constant-current” mode to efficiently produce disinfectant and biocide chemicals at many unmanned, non-air conditioned, municipal water treatment sites. In some applications, two or more ZUP supplies are connected in parallel to provide the necessary amount of current for the electrochemical process.


Another method of providing a “constant-current” mode power supply is to modify the design of a voltage regulated supply. This can be done by adding circuits that monitor the supply’s output current to prevent an overload, yet maintain a “constant-current” profile from the supply. For example, Lambda Americas has produced modified versions its HWS-CC 1500-watt supply to do exactly this. In electrochemical applications that produce disinfectant and biocide chemicals, a number of these “current-mode” supplies are connected to different sets of electrodes, and/or, in parallel, to support different output current requirements for various models of electrochemical generators. Obviously, generators that produce higher output rates of chemicals require higher current levels.

Lambda’s HWS-CC Series Power Supplies

This paper has focused on the techniques and benefits related to advanced switchmode power supplies for mini-electrochemical generators (self-contained plants) that produce disinfecting and biocide chemicals on-site. It should be noted that electrolysis processes are used extensively in many other chemical and industrial areas, some of which are listed below.
  • Production of aluminum, copper, sodium
  • Anodizing
  • Production of hydrogen (e.g., for the cars and fuel cells of the future)
  • Electroplating and polishing
  • Large waste water treatment plants
  • Factory and power plant cooling towers - recirculating water treatments
Many of these electrochemical processes require power levels that far exceed the range of the switchmode power supplies described above. These high power rectifier systems (ranging from 300 to 30,000 kW) are very specialized, large, heavy, and are usually comprised of huge transformers, rectifiers, thyristors, SCRs, capacitors, regulating controllers, and water cooling systems. Some of these high power sources are as large as a typical bathroom, kitchen, and larger. There is no doubt that as technologies advance, these huge power sources will see reductions in size and improvements in efficiencies.

In summary, the application of switchmode power supplies operating in a “constant-current” mode has been shown to provide significant improvements in electrochemical self contained mini-plant generators that are used to produce disinfecting and biocide chemicals. These benefits include:
  • Improved current-density control for consistent electrolysis
  • Enhanced quality of the resulting chemicals
  • Higher efficiencies and improved regulation of the power sources
  • Reduced space and weight
  • Power Supplies meet international Safety and Power Factor Correction (PFC) standards
  • Availability of digital communications, remote control, and status signals
  • Substantial reduction of downtime

References:
http://en.wikipedia.org/wiki/Electrolysis
http://www.medscape.com/viewarticle/520378
http://www.lambdapower.com/ http://www.doh.wa.gov/ehp/dw/Publications/alternate_disinfectants.htm

Thursday, September 6, 2007

What’s all this stuff about "Digital Power"?

It seems that every 2-3 weeks an article or news announcement about “Digital Power” appears in electronic design periodicals or online news links. In fact, one industry newsletter seems to be having a love affair with digital power, as it mentions it in just about every issue.

So what is all this fuss about Digital Power and what is it anyway? Well, the simple answer is that there are two basic types of digital power. These are Digital Control (used internal to the power devices) and Digital Power Management (provides external control and communications between power devices and a master controller).

Digital Control
The majority of switchmode AC-DC and DC-DC power supplies/converters use analog techniques to regulate/control the output voltage, current, and power factor correction circuits, etc. The closest that most of these devices come to looking a bit digital (On/Off states) in nature is by employing Pulse Width Modulation (PWM) in their switching regulator circuits; but even that is a bit of a stretch.

In recent years, new integrated circuits (ICs) have been developed that can replace “analog” control ICs and discrete circuits, which are used extensively in all power devices, with those that are, at least in part, “digital” in nature. These internal ICs and circuits perform such control functions as: voltage regulation (VR), power factor correction (PFC), pulse width modulation (PWM) control, internal monitoring/alarms, and external communications.

The advantage of these digital ICs is that they can be programmed by engineers with digital or analog electronics training. And, since the Universities are pumping out more digital (e.g., computer science) than analog engineers these days, these digital ICs are becoming attractive. However, at present the cost of these digital ICs (along with NRE for the equipment needed to program the devices) is still higher than for the mature analog ICs. Nonetheless, some predict that these IC costs will become equal within the next 12 months or so. A potential disadvantage of these digital ICs is that, by their nature, they require a high speed clock to operate, which can add to the radiated and conducted noise coming from the power supply or converter. However, advanced functions such as fault diagnostics/prevention and improved power efficiencies are among the promises of the new digital control ICs.


Digital Power Management
As mentioned above, Digital Power Management (DPM) involves the external control and communications between power supplies (or converters) and a master controller. Currently, many analog-based power solutions already have the ability to communicate with an external computer or controller via digital communications links (e.g., RS232, RS485, GPIB, or I2C bus).

Newer DPM control and communications formats have evolved that are designed to operate with the new digitally-controlled power devices. These include DPM technologies such as PMBus (Power Management Bus) and Z-One. Sadly, these technologies are not compatible or interchangeable. In fact, currently there are lawsuits between the backers of both of these technologies.

If I were a potential user of these Digital Power Management schemes, I would stay clear of them until the lawsuits are settled (expected to occur within the next 12 months), rather than find out later that the cost of these DPM ICs or controllers have substantially increased due to royalties that must now be paid to the company that won the lawsuit.

The potential advantages of the DPM and digital power technologies in general, include enhanced bidirectional communications, fault diagnostics, remote programming of the linked power supplies/converters, automatic compensation of dynamic input and output load changes, and overall improvements in efficiencies that relate to green-power.

In Summary
Although “Digital Power” is a popular buzzword these days, especially by those companies who have developed or adopted the technologies, the bottom line will always be: “What do I get for my money?” At present, there are hardly any power supply or converter applications that “must have” digital power when compared to the many lower cost and field-proven analog solutions that exist.

For example, during their new power-product design and development process, Lambda has designed in-parallel devices that employed both analog along with those that use digital control ICs and technologies. In all cases, the final decision on which technology ultimately goes into production has been based on comparative price/performance factors; which is the dominant decision factor for their customers.

When the time comes that digital power products offer the same or better performance and reliability, along with the ”needed features”, at the same or lower price as analog-based products, that is when Digital Power will become the winning technology. Realistically, someday digital power will provide a price/performance advantage over purely analog power devices. Who knows for sure when that time will come.

Tuesday, September 4, 2007

Can I Operate my AC-DC Power Supply with a DC Input?

The answer is yes, sometimes.

Many standard AC-DC switch mode power supplies (most of Lambda’s products) specify a high voltage DC input range in addition to the more common AC input range of 90-264VAC. We receive many questions about how and where to connect the DC input to an AC-DC supply that is spec’d to operate off of DC as well as an AC inputs.

Where and why is high voltage DC power used? It turns out that many power generation facilities provide a high voltage DC to power the plant’s equipment rather than the regular 115VAC or 208VAC power grid. This high voltage DC (typically 120 or 130-330VDC) can be easily used with batteries to provide a secure source of power rather than using expensive centralized or local UPS systems.

Now back to the subject. The topology of many switch-mode power supplies actually lends itself to operation from either AC or DC input. Important Note: Always check your power supply’s Operations Manual or spec sheet to confirm that it is designed to operate from either an AC or DC input.

Referring to the simplified power supply schematic below:


When powered by an AC sine wave, during the first half cycle the current flows from the Line terminal through the input filter and charges capacitor C1 through diodes D1 and D3. During the second (negative) cycle, current flows from the Neutral terminal and capacitor C1 is charged through diodes D2 and D4.

When powered from a high voltage DC source, the polarity of the connection is not critical as far as the operation of the power supply is concerned. If the positive connection is made to the Line terminal, C1 is charged through diodes D1 and D3. If the positive connection is made to the Neutral terminal, then C1 is charged through diodes D2 and D4.
An important note of caution to insert here is about the protective fusing of the power supply. Internally most power supplies have a fast acting AC rated fuse in series with the Line terminal. It is recommended that a DC rated fuse be installed external to the power supply. If one side of the high voltage DC buss is connected to ground, then the fuse is usually positioned in series with the “hot” side (the ungrounded side). It is recommended that you consult with your local safety engineer to be sure.

Share it

Popular Posts

TDK Corporation - Americas | Press Release