Capacitor Plague

As Capacitor Plague ( German: Kondensatorpest ) or " Badcaps " ( German: bad capacitors) was premature massive precipitation of aluminum electrolytic capacitors called with liquid electrolyte, which began in late 1999, mainly electrolytic capacitors Taiwanese manufacturer concerned and in the years 2002-2007 led to high failure rates in PC motherboards, power supplies, LCD monitors and many other electronic devices. The failures were caused by a water-driven corrosion and could be attributed to a faulty electrolyte. The troubled Taiwanese capacitors thus prepared were failed by statistical calculations in total by the end of 2007. The problem is now (2013 ) probably actually stopped because the last release of failures with similar error characteristics from the year 2010 comes. If now electrolytics occur with failures of this type, they are most likely to other causes, be such as congestion due.

Possible reason: industrial espionage

Called The massive failures of aluminum electrolytic capacitors with non-solid electrolyte, or " electrolytic capacitors ", in the years 1999 to 2007 based on a million times incorrectly manufactured capacitors from Taiwanese productions more producers mainly in PCs motherboards, PC power supplies, in LCD and monitors were used in power supplies and operating prematurely failed after a few months. Many of the electrolytic capacitors used were specified with the life (load life) from 2000 h/105 ° C. At an average internal temperature of a computer of 45 ° C and a ripple current corresponding to the data sheet specifications, these capacitors have a life expectancy of about 15 years of continuous operation. In case of failure by 1.5 to 2 years can really be said to "prematurely ".

The images were quite spectacular failure, broken Elko - cup, pushed out rubber stopper and spilled electrolyte were found on countless boards. Many well-known equipment manufacturers such as Dell, Cisco, Intel, Asus and Abit had to perform product recalls or accept repair costs because of these capacitors. Many repair instructions for self-help can be found on the internet.

As a probable cause of the faulty Elko - production industrial espionage is considered in connection with the theft of an electrolyte formula. An electrolyte developer has probably the change from Japan to Taiwan took the chemical composition of a new low-resistance, low-cost, water-containing electrolytes and then tried to recreate these electrolytes in Taiwan in order to then sell it cheaper than the Japanese can. But obviously the formula is copied only incomplete, lacking important substances to secure the long-term stability of the capacitors.

Development of capacitors with electrolyte water-based systems

• Construction of an aluminum electrolytic capacitor with a liquid electrolyte

Elko principle and cut through the layer sequence of Elko Construction

The first electrolytic capacitor at all was an aluminum electrolytic capacitor with a liquid electrolyte, invented by Charles Pollak in January 1896. Principle, the " capacitors " have remained the same. On an aluminum anode is the grown by formation of very thin dielectric alumina. The liquid electrolyte, the structure of the dielectric adapts, forms the cathode of the capacitor and thus makes the thin layer thickness of the dielectric loss. A spacer made ​​of paper prevents a direct contact of the oxide layer with a second aluminum foil ( cathode foil ), the current supply to the liquid electrolyte, and stores it. Shut up and provided with connections it gives the billions used in electronic equipment, inexpensive and generally quite reliable " electrolytics ".

The electrolyte as an ion conductor causes a large portion of the resistive losses in Elko. Great efforts have been made over the years to reduce these losses, so that the " ampacity " ( ripple current ) can be increased, because without such improvements, size reductions can not be realized.

In Japan the mid-1980s new, low- water-based electrolytes have been developed in the context of these developments, the conductivity of which was to electrolytes significantly improved with organic solvents. Water, with its relatively high permittivity of ε = 81 is an effective solvent for electrolytes. As such, it dissolves salts which only give its conductivity electrolyte in high concentration. The high concentration of the dissociated ions in the electrolyte, the better conductivity. But water reacts with unprotected aluminum quite violent. It converts this into a highly exothermic reaction of aluminum ( Al ) in its hydroxide (Al ( OH) 3) around. This is accompanied by a strong heat and gas development in the Elko and can lead to bursting of the capacitor. Therefore, the main problem was to get in the development of new water-containing electrolyte, the aggressiveness of the water compared to aluminum in the handle so that the capacitors have a sufficiently good long-term stability.

The Japanese manufacturer Rubycon was the late 1990s, a leader in the development of new water- electrolyte systems with improved conductivity. In 1998, Rubycon the "Z- series" with the two series ZL and ZA the first capacitors on the market that worked with an electrolyte with a water content of about 40% and for a wide temperature range of -40 to 105 ° C were suitable.

Improving the conductivity of the electrolyte obtained results from a comparison of two capacitors, both = have a capacitance value of 1000 uF at the rated voltage 16 V with the size DxL 10x20 mm. The capacitors of Rubycon series YXG, which are provided with an electrolyte based on an organic solvent, can be loaded with a ripple current of 1400 mA at an impedance of 46 milliohms. Capacitors of the ZL series with the new water- based electrolyte, however, can be loaded with an impedance of 23 milliohms with the ripple current of 1820 mA, an increase of 30%.

Other manufacturers such as NCC, Nichicon, Elna, etc. then followed a short time later. The new series have been touted as "low ESR " or "Low Impedance " "Ultra -Low- Impedance " or "High - Ripple Current electrolytics ". The highly competitive market in the digital data technology and the power supply handle this new development quickly, because by improving the conductivity of the electrolyte, the electrolytic capacitors could handle not only a higher ripple current load, they were even cheaper, since water is quite cheap. Better and cheaper as well, for the bulk products such as PCs, DVD players and power supplies, the cost argument was decisive.

Alumina - stable dielectric and corrosion protection

The electrolyte in an electrolytic capacitor is in electrical terms, the actual cathode of the capacitor. He's on the other side but also a chemical, an acid or an alkali, which must be chemically inert, so that the capacitor whose components are made of aluminum, stable long-term stays. But aluminum is a very ignoble, easily connecting with oxygen metal. And hydrous acids behave quite aggressively against the aluminum. Only a stable aluminum oxide Al2O3 layer on the surface of the metal and so-called inhibitors or Passivators in the electrolyte can protect aluminum from the aggressive reactions with the water and prevent corrosion driven by the water. The problem of water-containing electrolyte systems therefore is to control the aggressiveness of the water compared to aluminum.

This problem runs through the development of electrolytic capacitors have over many decades. Because even the first technically used electrolytes middle of the 20th century were mixtures of ethylene glycol and boric acid. But even in these so-called glycol electrolyte entered an unwanted chemical reaction on crystal water, according to the scheme:

In an initially anhydrous electrolytes an esterification occurs with a water content of up to 20 %. These electrolytes were originally only for a limited temperature range of -25 to 85 ° C. They had a voltage-dependent life because of corrosive effects of the residual current increased exponentially by the aggressiveness of the water, especially at higher temperatures with increasing voltage and was faster desiccation on the associated increased electrolyte consumption. Even newer " Glycol electrolytics " who have no corrosive effects more, still have a low dependence of the lifetime of the applied voltage.

Water but on the other hand also provides the oxygen in the self- healing of the electrolytic capacitor, the Reforming of Anodenoxids, the dielectric of the capacitor. Since depending on the pH and the temperature of the electrolyte can diffuse into the oxide of the dielectric and alter the crystal structure during storage periods ions. It can therefore on the one hand electrical defects in the oxide and on the other hand, occur by the dielectric are weakened, which leads to a reduction of the insulation resistance. Both effects result when applying a voltage to an increased leakage current, which, however, goes back again by formation of aluminum oxide. This normal operation of the Nachformierung occurs in two reaction steps. First, it converts in a highly exothermic reaction of aluminum ( Al ) in its hydroxide Al ( OH) 3 to:

This reaction is accelerated by a high electric field and high temperatures and is accompanied by a pressure build-up in the capacitor by the released hydrogen gas.

The gel-like Aluminiumorthohydroxid Al (OH ) 3, aluminum hydroxide, aluminum hydrate or aluminum trihydrate ( ATH) called, usually slowly changing in the second reaction step at room temperature in a few days, at higher temperatures more quickly, in the crystalline form of aluminum oxide Al2O3 to:

Only the newly formed during the Nachformierung stable aluminum oxide, with which the defects on the anode of the electrolytic capacitors are annealed, forming the stable dielectric of the capacitor. It also protects the capacitor against the aggressive reactions of aluminum in the presence of water. However, the electrolyte is taken in the conversion process, a small amount of water. It is electrolyte "consumed".

Aluminum hydroxide - loss of self-healing

The aluminum oxide layer in the electrolytic capacitor is resistant to chemical attack, as long as the pH of the electrolyte is in the range of pH 4.5 to 8.5. However, the pH value of the electrolyte should ideally be slightly acidic (pH = 5 ) to about 7 (neutral ). Measurements of the residual current, which have already been carried out in the 1970s, have shown that the residual current was larger, ie occurring chemically related defects as soon as the pH left this ideal range.

It is also known that the "normal" course of the formation of the aluminum oxide can be broken by the aluminum on the intermediate step of the aluminum hydroxide to aluminum oxide through a stable alkaline electrolyte. The chemistry of this interruption, the following reaction is an example:

In this case, it may happen that the hydroxide formed in the first step separates from the aluminum surface, not transforms into the ( desired ) alumina, and the cause of the formation of oxide, a flaw or weakness in the dielectric, is retained and the defect is not cured. Then there will be a further formation of aluminum hydroxide at this point, without conversion to the stable alumina. The self-healing in Elko ( Nachformierung ) no longer takes place. The reactions do not come to a halt, the pores in the anode foil proliferate with the hydroxide to and created by the forms in the reaction of hydrogen gas in Elko mug an ever increasing pressure.

Evidence of false electrolyte

• Pictures of an open Ausfallelkos

Unsettled Elko wraps, the films adhere firmly together

This situation of unbridled hydroxide (English: hydration ) and the associated formation of gas was it that led to the above as " capacitor plague" or "bad caps" incident with the masses of precipitating aluminum electrolytic capacitors. Proved it was the investigation of failed capacitors of Taiwanese manufacturers by C. Hillman and N. Helmond.

These two scientists from the University of Maryland presented initially determined by ion exchange chromatography and mass spectrometry, that it is indeed hydrogen gas, leading to bulging of the Elko - cup and open the valve cup. Thus it has been proved that the oxidation of the first stage of the formation of aluminum oxide to take place.

Because it is common in electrolytic capacitors, to bind the hydrogen formed with the aid of reducing compounds to reduce the generated pressure, and then searched for compounds of this type. To this end, mostly aromatic nitro compounds or amines. Although the above research methods used are very sensitive, not pressure-reducing compounds could be detected. For capacitors, the internal pressure build-up was so great just that the cup was bulged but the valve had not opened yet, could then be detected in the analysis of the electrolyte, the electrolyte of the faulty capacitors an alkaline, alkaline pH ( 7 < pH < 8) had. Comparable Japanese electrolytic capacitors on the other hand had an electrolyte having a pH in the acidic range (pH ≈ 4). Since it is known that can be solved in alkaline liquids aluminum in acidic media but not then the electrolyte from the faulty capacitors was investigated by means of an EDX fingerprint analysis and it could actually dissolved aluminum are detected.

To protect the aluminum against the aggressiveness of the water are phosphate compounds, known inhibitors or deactivators, used when it comes to long-term stable, aqueous electrolytic capacitors. Since in the investigated Taiwanese electrolyte phosphate ions were missing, the electrolyte also was still alkaline, so they lacked the protection against the water and the more stable alumina formation was stopped. There was no brakes, only aluminum hydroxide.

It was underlined the chemical outcome of the investigation by the electrical measurement of the capacitance and the leakage current in a long-term test over 56 days. Due to the chemical attack on the oxide layer of the capacitors it has been weakened so that after a short time, the capacitance and the leakage current increased before both parameters after the opening of the valve fell off rapidly.

With the report of Hilmann and Helmond was proof that the cause of failure was actually a term used by the Taiwanese manufacturers of defective electrolyte, it lacked the necessary for long-term stability of the electrolytic capacitors ingredients that should ensure the stability of the capacitors.

The fact that the electrolyte with his lying in the basic range pH then the fatal unchecked hydroxide had the effect can be both photographically as demonstrated with an EDX fingerprint analysis of the chemical constituents of the surface oxide.

Even with a micrograph with only a 10-fold magnification, as shown in the pictures to the right, a significant change in the structure of the anode surface is visible. On the surface of the "fresh" unused anode from an electrolytic capacitor, the grooves in the running direction of the anode, resulting in the production process, is clearly visible. However, the increase was insufficient to show the openings of the pores in the anode. On the anode, which comes from a failed electrolytic capacitor from the Taiwanese production, the surface is overgrown transversely to the running direction with a placke like substance. EDX fingerprint analysis then showed chemical difference in surface oxide. The surface of the "fresh" Elko was covered with the stable alumina. The surface of the failed Elko was that proves the significantly higher oxygen peak, covered with aluminum hydroxide.

Electrical effects

A slightly different electrical behavior in nearly all electrolytic capacitors with aqueous electrolyte can be measured with electrolyte systems based on organic solvents with respect to the electrolytic capacitor. The power-on leakage current on the storage time is at a higher level. However, if a water- electrolyte ie still not yet caught in the basic pH range in the steady state, then the residual current is set after a few minutes to a low value. The defects which had been formed by the aggressive behavior of water on aluminum are healed quickly. However, the electrolyte lost its stability during operation and has drifted in the basic range, then ends the process of self-healing after the formation of aluminum hydroxide. Subsequent conversion to the stable alumina is inhibited by the basic environment. The defects that cause the residual current, remain unprotected obtained only by hydroxide covered and can be attacked by water.

The withstand voltage of this hydroxide is, however, only reach the level of the applied operating voltage, which is usually much lower than the rated voltage. This impaired withstand voltage of the anode can also be measured. In the picture, such a measurement result is reproduced. The reduced dielectric strength compared to the original anode voltage strength shows that a chemical process has damaged the oxide layer of the dielectric term.

This damage is the thickness of the insulating layer, the effective dielectric smaller. This means according to the formula of the plate capacitor,

In ε, the permittivity, A is the electrode surface area and d is the distance of the electrodes to each other, which increases with thinner dielectric of the capacitance value. But an increase in capacity to a higher value as the trace of the capacity curve in a life test in the picture Indeed, at the electrolytic capacitors in which the unrestrained hydroxide has already begun, the Elko has not yet burst, measured, showing the top right.

The final stage of this process is achieved when has risen so high by the constant and ever more rapid unchecked hydroxide, the pressure in the condenser, causing the valve to open or to express the rubber stopper. Simply put, it bursts. If the capacitor is then open, it dries out very quickly loses its capacity down to a minimum value and the ESR increases significantly up to the area kOhm. Since the ripple current continues to flow through the remaining capacitor will become hot at higher currents of the capacitor winding and color the paper of the package clearly brown.

Reasons for Elko failures after 2007

The first press releases about the mass occurrence of failures appeared in September 2002., It can be assumed that by the mid of 2003 the producers concerned changed the production and have resorted to a "right" composite electrolyte. With a decreased life span of about 1.5 to 3 years, so you should actually be down by mid-year 2006, the last of the faulty capacitors. The Internet is called 2007 as the end point of the failures with improper electrolyte often the year. After this date should actually occur no further incidents more. But even after 2007, the year in which the failure of the Taiwanese capacitors with the wrong electrolytes should be actually past, the messages failed electrolytic capacitors are available on the web. The problem of bursting capacitors is so still, because the described failure images with ruptured Elko, expressed cups and rubber stoppers are identical with the former failures.

Affected by these failures are Elko series for rated voltages from 6.3 V to 100 V, which have one thing in common, they have a water-based electrolyte with a very high water content of up to 75%. In the catalogs of the manufacturers they are identified under "low- ESR " capacitors or even with low- impedance "," Ulta -Low- Impedance "or" High - Ripple -Current Capacitors ". These capacitors must not be confused with aluminum -polymer capacitors, which are often called "low- ESR electrolytic capacitors ". Affected are only aluminum electrolytic capacitors with liquid electrolytes.

If so failures in the above- described type of capacitors with a newer date of manufacture, so it can not be due to the incomplete electrolyte of the former Taiwanese production. If detected in an SEM and EDX analysis of the failed capacitors still aluminum hydroxide, the electrolyte is not automatically wrong, because it can also be the circuit design have been wrong. Therefore first two questions need to be answered:

It should be noted that the commonly used 10 - degree rule ( Arrhenius rule RGT rule) to estimate the Elko - life for the electrolytic capacitors with aqueous electrolytes not ( per 10 ° C decrease in temperature, the lifetime doubles ) is often considered. The 10 - degree rule applies only when it is confirmed by the respective manufacturers Elko. Because some manufacturers specify other life calculation formulas, sometimes even different formulas for their various series. Although a graphical method for estimating the Elko - life is specified by a manufacturer, and seem to follow the 10 -degree law, these curves should not be fooled. The gradient according to the 10 -degree control in this example is different than the specified curve.

The following two examples of Elko - life calculations are intended to show the difference in results between the 10 -degree formula and the formula that specifies the manufacturer Rubycon for provided with water containing electrolytes series ZL series.

For a PC power supply, an electrolytic capacitor is selected with the lifetime specification 1000 h/105 ° C. The average operating temperature of the capacitor as measured in the metal region of the predetermined breaking point is 45 ° C. The ripple current load corresponds to the first example, the data sheet value (100 %), in the second example the double data sheet value.

In a ripple current load of twice the data sheet value is:

The calculated life under the double value of the ripple current is significantly smaller than by using the 10 -degree formula as rubycon formula. This expresses that the operation of electrolytic capacitors can be problematic in case of overload with aqueous electrolyte. This also comes in a warning note many Elko manufacturers expressed in the operation with a higher ripple current is warned than the specified value:

• Do not apply a ripple current exceeding the rated maximum ripple current.

The difference in the reduction of the service life at a higher load of electrolytic capacitors with aqueous electrolyte to such electrolytes based on organic solvents can be found on the one hand in the aggressiveness of the water towards the aluminum. Electrolytic capacitors with solvent electrolyte such as GBL have a significantly better leakage current behavior, so that less electrolyte is consumed during operation and therefore may affect the operation of a capacitor even to the specification to estimate the service life. Water-containing capacitors consume then against because of the higher leakage current more than Lösungsmittelelkos electrolyte, thereby increasing the life of the capacitors is reduced.

On the other hand, a high ripple current load also have a consumption of the electrolyte to result because especially when discharging of the capacitor occurs, a physical effect that can lead to the cathode foil in the capacitor Aufformierung circumstances. The cathode foil is covered with an oxide layer, which is caused by the contact with the air in a natural way on the aluminum surface. This oxide layer has a dielectric strength at room temperature of about 1 to 1.5 V, at 105 ° C, these withstand voltage decreases to about 0.7 to 1.2 volts. If now a charged capacitor is discharged, then the polarity is reversed in the capacitor: For the cathode, an anode, the current flows from the capacitor out. About the stress distribution at the transition and line resistances then a voltage builds up reverse polarity at the cathode foil with a ripple current load only has no Aufformierung the cathode foil with the formation of a thicker oxide layer results in up to specified data sheet value when the cathode capacitance CK very is large relative to the anode capacitance CA. Typically, this is achieved when the cathode capacity is larger by a factor of 10 than the anode capacity. However, at higher ripple current levels, especially if the operating temperature is still too high, it may cause Formiervorgängen. The formation of new, more voltage-stable oxide layer is associated with electrolyte depletion.

The life of the electrolytic capacitors with a high water content in the electrolyte is thus determined not only by the operating temperature and the resulting gradual evaporation of the electrolyte, but also by the residual current and a possible behavior Aufformierung the cathode foil at a high ripple current load. All factors, ambient temperature, over the operating time higher residual current and the Aufformierung the cathode foil with ripple current overload " consume " liquid electrolyte. From a certain point onward, the dissolved in the electrolyte to the saturation limit salts crystallize out. The electrolyte is changed, first reduces the conductivity of the electrolyte, the ESR increases. The change of the electrolyte has in some product lines but also influence the pH, the pH may change, and indeed the greater, the closer he gets to the end of service life of electrolytic capacitors. Device then the pH of the electrolyte at the end of its life, the basic region, which is hydrated with electrolytes found in electrolytic capacitors often, then stops, as described above, the regeneration of the defects and the fatal, unbridled aluminum hydroxide formation begins. If this behavior now with the calculated end of life of the capacitors together and bursts of capacitor after the end of life, then it looks as if a new case from the " capacitor plague" occurred.

The failure pictures, burst Elko, pushed up rubber stopper, released electrolyte are identical to those of the production with the "wrong " electrolyte., So it can if now after the year 2007 still experiencing failures of this type, be determined only by a careful recalculation of the entire circuit with all boundary conditions for the capacitor such as temperature and ripple current load, if it is an error in Elko or an error the design of the circuit is. For only when the capacitor is prematurely failed ie before the end of its useful life, it can be assumed that a faulty electrolyte. Perhaps it is also the case of a well-known manufacturer of computer hardware with an electrolytic capacitor manufacturers, such as a publication in " techreport " can be seen, attributed to an incorrect calculation performed the Elko - load and the Elko life.

However, a burst of an electrolytic capacitor, even if it has reached its end of life, have an extraordinary process. Date could be assumed that while electrolytic capacitors dry out over time, but also in the dried state exhibited no external irregularities. It seems that water-containing capacitors some manufacturers have a different behavior at the end of their life: they burst! At ruptured Elko turn can no longer possible to determine whether thermal or electrical overload has led to the opening of the valve or whether the failure was prematurely due to poor coordination of the electrolyte with subsequent hydroxide. For a correct assessment of newer Elko failures so it is essential to determine the exact operating conditions of operation and perform a life assessment in accordance with the manufacturer's specification.

Coding of the manufacturing date

Many Elko manufacturers use a two-digit code to encrypt the date of manufacture ( date code)

• First digit: year of manufacture, S = 2004 T = 2005 U = 2006 V = 2007 W = 2008 X = 2009, A = 2010, B = 2011, C = 2012, D = 2013 E = 2014 F = 2015
• Second place: Herstellmonat, 1-9 = Jan. to Sept., O = of Oct, N = Nov. D = December