When and Why a Power Supply Fails
Why is it that petrochemical disasters are etched more in our minds than when an installation is operating without any issues? How many times has a radio station been taken off the air or a factory shut down by a lightning strike? In a morbid sort of way, how is it that we find it so fascinating when Category 5 hurricanes, ice storms or 200 MPH tornadoes shut down a city’s infrastructure or telecommunication system? Simply put, a community, an installation or facility without issues - is not news.
For those that are in the know, there are not enough fingers and toes to count the times when critical equipment malfunctions. A calculator is required to add up the times when it is not business as usual or when power generators or other power providing systems fail. But, putting aside meteorological events for a moment, too often a variety of other issues will cause equipment malfunctions. There are the ongoing situations when aging equipment disrupts operations or products when they are subjected to harsh environments. Often, surge protection systems are not properly designed for factories. Worst of all, there are cases where an improperly trained individual throws the wrong switch or a floor technician mistakenly plugs in a power supply the wrong way.
In the best of all worlds, a “shutdown” will stop equipment for only a few minutes or, perhaps, an hour or so. But, suppose there was “a catastrophic event” or even an apparent innocuous incident translating into shutting down operations for days or even weeks? And, suppose the effect of these “calamitous conditions” could have been minimized or prevented entirely?
When customers and power supply vendors kick-off exploratory product design meetings, there is the typical discussion about volts, amps, watts and basic options – “Power Good,” “Power Fail” and “Over Voltage Protection.” Interestingly, during the initial Q&A conversation between the OEM and the power supply vendor, “Quality” is rarely discussed since it is assumed that the workmanship standards are exemplary and the finished product meets all of the design and testing constraints. Yes, there may be a question or two about DMTBF (Demonstrated Mean Time Before Failure). Numbers such as 250,000, 500,000 or even 1 million hours might be bandied about.
Surprisingly, rarely does the OEM drill down to determine what those numbers mean. Is it based on one type of power supply model or a family of similar products? What actually is the field population used to generate the statistics? What was the application that led to the DMTBF assumption? What was the field return percentage for this group of products? And, how were these numbers actually calculated?
While discussing a future program and the search for the “best” power supply, in addition to DMTBF, often other aspects critical to a successful program need to be addressed. Examples would be manufacturing processes, quality systems, MRP controls and how line personnel interface with each other and upper management. Are employees empowered to make low-level decisions? Are they encouraged to communicate with all management levels? What type of relationship does the power supply company have with its supply chain partners? Will the power supply company’s vendors “jump through hoops” to support an unplanned requirement? Does the Human Resources Department maintain up-to-date records documenting the skill level of each employee? Is there an on-going safety program in place?
Equally important for the power supply company is to have a clear understanding of the specific application and how the product will actually be used. Is the physical location of a plant or platform and/or if the power supply will be airborne, in a marine environment or in areas where there will be little or no technical support. Are the customer’s applications calling for product to be situated in (very) hard to get to places? There is even a “hands off” application where the product is located where once everything is installed, “the door is closed and the key is (so to speak) thrown away.”
Particularly important, for the power supply company is to know what peripheral equipment will be used and where the power supply will actually be physically installed in a cabinet or system. If the power supply is improperly interfacing with other electronic equipment, mounted too close to cabling, if system cooling is inadequate, the performance and life of the product may be compromised.
Yet, too often a procurement specialist is not aware that the power supply will be adjacent to or interconnected with other electronic devices, which could change the efficiency, interrupt or even harm the unit. For instance, components may fail if the power supply is exposed to input voltage fluctuations beyond the product specification. The good news is that with precise information from the customer concerning the application, a power supply can be easily designed or modified to handle voltage issues or most other matters beyond the normal specifications.
But, suppose a system has gone down and an on-site field tech is sent to resolve the problem? Typically, the first step is to remove the perceived failed power supply and replace it with a new unit. What happens to the removed power supply if the system continues to malfunction? Normally, the power supply is returned to the manufacturer or to a third-party repair house with the statement, “Broken,” “It doesn’t work!” or “Fix as required!” It is rather disconcerting for PMI to see that >20% of the product returned “for repair” is categorized as NPF (No Problem Found!).
Putting aside NPF issues, suppose an older technology product (or even a new power supply for that matter) is returned due to some sort of undefined failure mode. After a quick confirmation that the power supply doesn’t work, it is fascinating for the power supply company to discover that often the root cause of the failure is corrosion, contaminated dust or some other foreign matter preventing the unit from working! The obvious question: Was the power supply company informed that the product would be subjected to a harsh application? How was the basic specification or application defined?
Recently, power supplies in the field for < six months were returned with a complaint that the units had “failed” or that they were “overheating.” A quick observation revealed that the units were filled with a thick oily substance. The root cause was actually rather amusing.
The field application called for the power supplies to activate a camera device used to select fruit size when the items move along on an outdoor conveyor belt. It turned out that the conveyor equipment was slowing down. All of the conveyor’s moving parts were getting hung up from all of the dirt, twigs and other debris when a truck would drive up to the conveyor system and dump the fruit into a hopper which would then, in turn, transfer the fruit onto the belt.
The maintenance staff at the fruit sorting facility determined that the solution to resolve the slowing conveyor was to spray heavy lubricants into all of the equipment’s moving parts. However, the equipment was not turned off resulting in the power supply fan motor sucking in the oily substance, which, in turn caused the fan to stop running and the power supplies to malfunction. This is a typical example where the end user should have trained its staff to turn off the conveyor unit before cleaning and lubricating the equipment. In addition, we could have conformal coated the power supplies to minimize other issues.
With regard to conformal coating, this process often reduces the ramifications of a harsh environment on the power supply. On the surface, this seems like a reasonable approach to prolonging the power supply life. Yet, OEM and end user budgets tend to come into play. A procurement decision-maker may try to outguess statistics. How long will the power supply last with conformal coating or without it? Is the cost to protect the power supply worth the expense?
It’s interesting that these important power supply specification (criteria) discussions quickly evolve into the bottom line. Is the cost of the power supply “reasonable” or “expensive?” Does the product’s uniqueness (as it relates to location, function and specifications) translate into increased costs? How long will the power supply last if the product is not designed to withstand a harsh environment? What are the ramifications for an installation if it is not fully operating and/or if (fortunately) a spare power supply is available and will be used while the other product is being repaired? With all of these questions in mind, most remarkable is the “communication process” (or lack of communication) that occurs between a field operator and the supplier when a power supply has malfunctioned.
Typically, a customer or end user contacts the customer service department indicating that a unit has failed, equipment is shut down, there are no spares and an immediate turn around is required. Or, an import-export company sends an email requiring a quote for a replacement power supply “with a sense of urgency” – the unit needs to be shipped the next day. Obviously, the supplier wants to help. As a first step, the end user is asked a myriad of questions. Often, it is learned that the failed power supply in question is actually manufactured by another company – that the power supply vendor is no longer servicing older technology products or the company is no longer in business.
However, putting aside the manufacturer, when a field service person is asked about ratings, dimensions, wattage, output voltage, etc., it sounds incredulous but there have been too many times when the response is “there was a power glitch…it’s in a silver box… the unit was dropped…the voltage rating on the label can’t be read . . . but it weighs about 10 pounds.” Through a process of elimination, it’s usually possible to determine what’s inside the “silver box” and whether or not we will be able to assist the end user.
Hopefully, from the customer’s perspective, the power supply manufacturer is still in business and is supporting the repair of the older-type products. If not, there may be third-party repair houses that can try to repair the defunct product. Or, as an alternative, there might be a few power supply manufacturers that continue to build older technology power supplies. There might be a possibility that these power supply companies, supporting the legacy technologies, may have a drop fit solution or a product that would be “close enough” to be used as a replacement.
In the event that servicing a power supply is deemed the best approach to resolving an inoperative system, repairing a power supply in an emergency situation (from a system perspective) may only be a short-term fix. There may be many older technology products in the field (having been manufactured and shipped out at the same time) that will soon reach their end of life. Therefore, why not be proactive and establish a preventative maintenance program? Why not obtain information and, particularly, good information that can be used as a foundation for years to come? The usual answer is that “…preventative maintenance programs are too expensive and that the thought of recycling power supplies installed in systems throughout the world would be a logistical nightmare!”
Given that a preventative maintenance program is not an option, there are several key (inexpensive) steps that installations or facilities could start doing – immediately:
- Put together a list of all essential equipment and those devices, in particular that interface with the power supply
- Determine if operating manuals are on site
- Check if troubleshooting guidelines are available
- Establish a list of vendor company names, model numbers, specifications, phone numbers and e-mail addresses. Perhaps a contact name such as the vendor’s customer service agent would be useful.
- Initiate a communication process with the power supply vendor with regard to the availability of material required to repair the power supply. Often, key components that were originally designed into the power supply ten or twenty years ago are no longer available. What is the power supply company doing about this issue? Are equally good components being used as an alternative to the originally designed-in parts?
- Does the power supply vendor have a solution if the older technology products can no longer be repaired? At times, power supply manufacturers will have a newer technology product that is a drop fit and, perhaps, a product that is smaller, more efficient and, in some comes, less costly! However, “being better, newer and smaller” does not always mean the “best” solution!
- The end user may want to consider a “last time buy” program whereby spares can be procured and put into stock for a just-in-case situation. It seems that having “a few thousand dollars” of product sitting on a shelf is a needed insurance policy as compared to having a plant down with hundreds of thousands of dollars of lost revenue!
In most cases, when we shop for a flashlight battery, the battery manufacturer is not a consideration – it's price! Shopping for a light bulb is based on wattage – not whether or not the bulb will screw into the socket. Many of us ask the “local mechanic” to fix a problem with our car rather than going to a more expensive dealer. Why? Most of the products we buy today or the service we expect are based on a “standard.” Therefore, “price” often becomes the driving force rather than size or shape.
Much like a battery or a light bulb, with all of the different applications, one would think that a power supply is a power supply is a power supply – particularly the more high-end, switching types designed to operate everywhere 24/7 and which are one of the most essential aspects of so many industries. Wouldn’t it be nice if there were a common type of power supply that serves all industries?
Unfortunately, there is no such thing as a standard power device or universal product, per se, crossing over to all aspects of the myriad of industries. To complicate matters, the creativity of a power supply design engineer or an OEM specification attempting to address all intervening variables will not translate into an all-purpose product. Essentially, it is almost impossible to imagine a cutting edge, broad-spectrum product that supports all applications. To try to anticipate every variable imaginable actually evolves into difficulties not considered during the initial design process, and, if it were conceivable to design a product that was totally universal, the costs would be prohibitive.
However, with clear-cut information from the OEM, the power supply company will determine if the power supply initially proposed is designed to (only) operate in a benign, office situation or (perhaps) if an alternative product is more suitable in order to withstand a harsh environment. With free-flowing and honest dialogue, aggravation can be significantly minimized if not entirely eliminated. And, if this occurs, all systems will be – Go!
Provided by Jerry Rosenstein,
President & COO, Pioneer Magnetics
©2013 Pioneer Magnetics