A troubling trend in our technological society today is when good capacitors (caps) go bad. Perhaps this is a phenomena that you have heard of somewhere in passing, but do not know as much about as you'd like to?

In which case you have come to the right place! Because in this article I will endeavour to discuss when good caps go bad, and how you might spot the bad caps for yourself too.

Step 1: What Are Bad Caps, and Why Should I Care?

A bad capacitor is an electronics component that over the course of its life has turned to the dark side. It is evil now and is no longer serving its intended purpose in life. It is a hazard to all other electronic components that are relying on it functioning properly now too. In short, it is broken. We will soon learn it is a short. Here is one that my tester thinks is a diode. Good caps will never read as diodes.

Step 2: Let's Try to Read That Cap Again, Shall We?

In this image I have flipped the cap around in the test clips and read it again. Now it is coming up as a capacitor, and we can see readings related to capacitance too. Let us take a moment to decrypt the readings the meter is displaying now.

Reading across we see; 1-||-3 Vloss=34%

On the next line is shown; 943μF ESR=2.2Ω

1-||-3 are the test points the meter has found a component on

Vloss is the amount of voltage the meter has measured that the capacitor has lost.

943μF is the capacitance the meter has measured of the capacitor. 943 micro Farads.

ESR=2.2Ω is the measured Equivalent Series Resistance (ESR) of the device under test.

Well, you may ask, Why is any of this so bad? To begin with it is a 2,200μF capacitor so 943μF is not even close to its marked capacity. But that is not the worst of it. A voltage loss of 34% is terrible! This part is leaking electricity like a sieve. Which is not what capacitors are supposed to do at all. Quite the opposite in fact. Capacitors are supposed to be able to store electricity. They are supposed to have the capacity to store an electrical charge. Hence their name. The ESR measurement is pretty bad too, but not nearly as out of spec as the Vloss.

There is a lot of talk about ESR, so if you are interested in that I suggest you browse the web for additional information on that topic.

Step 3: How Can I Spot a Bad Cap?

You may not have one of these fancy meters at your disposal. But the good news is bad caps often appear bad visually! So you can just see that they're bad. OK so what are the telltale signs of a bad cap? Well, bad caps typically have a domed, or swollen top. Sometimes really bad caps can leak their electrolyte out of themselves too. Then you may see this brown crust around the capacitor, or perhaps on it. It often looks somewhat like a dried coffee stain.

In this image I have tried to photograph the slight bulge on the top of this bad capacitor. It is easier to see in person than in my picture. A good cap will have a very flat top. Any bulge is a bad sign.

You can also use some deductive reasoning to find the more likely places in circuits where caps may go bad. Capacitors that are placed under heavy stress are more likely to go bad than caps worked less hard. So power supply filter caps go bad often. The power supply is often where the power comes into equipment. You may see buck boost coils there, a diode bridge, heavy switching MOSFETs, and things like that. All of those items are a bad influence on impressionable caps, and can make them go bad.

Step 4: Why Should I Care About Bad Caps?

A fair question. Realize that a few inexpensive capacitors can be the difference between your expensive electronics working, or not. I had several caps go bad in my flat screen TV, and when they did it would no longer come on! $4 of parts replaced brought a $1,400 TV back to life. I bought my TV when they were more expensive than they are today.

Yes, I was extremely ticked when my TV seemed to die at such an early age too. But when I resurrected it I was happy then!

So you too could suffer from early infant mortality of your expensive electronic devices, just like I did. Bad caps cause a lot of electronics to fail long before they should today. This is a huge issue. Any electronics that you have that plugs into a wall outlet has filter capacitors in it. When they go so does your whole piece of equipment then too.

If you can spot bad caps for yourself you can fix this equipment for yourself too. Saving you hundreds, if not thousands of dollars. Over what repair, or replacement costs may add up to.

Step 5: Enough About Bad Caps Are There Any Good Caps?

Yes there are! Not all caps are created equally. The Japanese make a superior capacitor than what is available from some other parts of the world. When you buy replacement caps it pays you to spend the few pennies extra and get these better capacitors for yourself too.

In my capacitor testing I have found that the absolute best capacitors I have run across are manufactured by Rubycon. Rubys are some fine capacitors. Although Nichicons are no slouches either. These are called Jap caps. No slur intended. Because they are the very best caps you will find today. They generally do not cost much more than the inferior ones do either.

But if you are a manufacturer buying tens of thousands lots of capacitors then pennies add up I suppose. Still what is going on is a grave injustice to all of humanity. Expensive equipment dying prematurely over such an insignificant cost difference is a travesty.

Step 6: For Folks Curious About My Meter

Someone is bound to ask, so I might as well say something about the meter I used in this post. I am using something called an "AVR TransistorTester by Karl-Heinz Kübbeler". There are a number of models available from the Far East. The first one I purchased was called an LCR-3 (now going under the name of Mega328 Transistor Tester). Sometimes it is referred to as the "graphical meter". Because it has a larger screen that it draws the parts out on graphically. Unfortunately the designers of it took some license with how they made the circuitry. This makes the meter less accurate as a result. (I have thought about trying to rectify that but I am not sure if I can due to a lack of firmware availability)

I learned all of this later as I researched the topic more myself. So then I bought a kit that I assembled my "AVR TransistorTester by Karl-Heinz Kübbeler" clone from myself. I might as well supply a link to that here now too, because people are likely going to ask for that as well.


I am in no way endorsing that company, I am merely providing information people may want. But honestly I think the whole deal is pretty sweet. I am extremely satisfied with my meter today too. Here is a link to the original open source project all of these meters were derived from


Which is in German, but you can translate it. Anyhow, the meters are pretty neat and anyone with even a passing interest in electronics would be wise to get one for themselves. They are an awesome deal for what they are. They will help you pinpoint bad caps even better too. Because not all bad caps look bad.

<p>As a novice electronics tinkerer, this is valuable info! Thank you for clearing up a lot of question marks. I also heard through many websites, that offer kits or schematics, that when given a choice, buy the components from Japan. This is a nice confirmation. Thank you again!</p>
<p>My tried and tested method of putting capacitors on trial are as follows:</p><p>Connect both leads to the L/N of a mains plug, turn it on and see what happens; if the capacitor explodes spectacularly, then it was a capacitor of pure heart, if it doesn't explode, then it is an evil capacitor.</p>
&quot;If she drowns and dies she wasn't a witch. If she floats and lives she's she's a witch!&quot;
Only buy Japanese capacitors! All Chinese capasitors are rubbish.
<p>Nice tut!<br>Good to know about this things, but a question arises to me, how do I take out the cap, and how do I replace it back? Whats's your best method to do so?<br>In advance thanks for your reply ;)</p>
<p>Desoldering components is beyond the scope of this article. But generally heating one lead, and levering the part out works. Usually after going back, and forth a few times they come up. It helps to add solder to the joint too. The new lead free solder they use today is horrible. That is another conspiracy being perpetrated on the consuming public. Did you know the military, and aerospace are exempt? Strange how only we have to suffer, isn't it? I only use 60/40 solder myself. Don't bother with that lead free stuff, it is trash.</p><p>To put a new cap in you need to clear the board holes the cap was sitting in. I use a solder puller for that. Electrolytic capacitors are usually polarized, so observe proper orientation when installing one!</p>
<p>very lovely and useful clues to some not that excellent but have records in doing some good trouble shooting. My Samsung 37 inch TV has white rings shown on LCD screen with no problem in the input signal. In the power supply one 820 microfarad has a slight bulge on the top. 1500 microfarad 35 volts also slightly bulged. Do you think changing the caps will cure the problem.</p>
<p>Rather late, but YES definitely change that cap and any which are close to it. Just make sure you take note of the voltage, going over is OK but not under. You can often get away with something close in value, eg a 1680uF instead of a 1500uF (uF = Microfarad.</p>
<p>People should be aware that capacitors, especially the variety found in older appliances may have a brown or whitish gunge looking substance around their bases. This isn't always a sign of problems, it could be simply the goo manufacturers slap around caps to anchor them to the PCB. Also be very careful not to go prodding capacitors, especially the larger types. I've seen one decide to spurt a fountain of substance, something like spider web when ejected and if it lands in your eyes it could prove nasty. By the way, I have one of the BangGood test meters and it's excellent for all common components. If you buy one opt for the plastic case made for the unit, it's worth the extra few dollars.</p>
<p>Yes the thicker goo is some kind of a sealer, or glue, that manufacturers use, for an unfathomable reason. Maybe the foreman's kid needed a job at the factory? I don't honestly know (perhaps to keep parts stable while the board is being wave soldered?). When a cap squirts we often see just the dried residue, and it looks a lot like someone spilled coffee on the circuit board, and it dried up. It always looks like a brown crust to me. The stain is quite thin. It can often be very subtle, the residue left over from a leaked capacitor. Like residue left from just a small drop of fluid.</p><p>But folks looking closely for such things should be able to notice it if peering carefully enough. When troubleshooting any telltale sign is a valuable clue. Although every clue does need to be investigated thoroughly, to determine precisely what it means.</p><p>When I strip boards for parts I hate that glue goo that is on them sometimes. I was just peeling some off a pair of capacitors yesterday here. It was some better translucent silicone based glop. The brown stuff is like cheap hot melt glue. Worst industrial quality.</p><p>The fumes that come off the brown stuff when heated probably reduce lifespan. Some electronics are doped to smell terrible when they burn, Just so folks realize it is burning. Maybe they'll unplug it then too? One can always hope!</p>
<p>it's a very professional touch and something I sure appreciate - when a manufacturer uses Silastic or other elastomeric compound to stabilize and dampen parts against the circuit board.</p>
<p>A brown coffee-like stain could also be (burnt?) flux, used during the soldering process.</p>
<p>Not on top of a capacitor can it can't be. Plus flux is shiny, and cap leak goo is not. When you see flux you should clean it off anyways. Flux left on circuit boards is just sloppy. But in consumer electronics that is common. Because cleaning rosin flux is pretty time consuming. Most mass produced electronics are assembled by machines and they use water soluble flux. But often some parts need to be hand soldered too. I haven't seen completely hand soldered circuit boards in decades. Like old electronics that came out of Japan in the 60s were hand soldered. No one is doing that kind of thing today though.</p><p>Well, I'm sure someone is, but it is extremely rare.</p>
<p>The glue is used for one of two reasons. </p><p>1) It holds the cap in place during assembly, prior to soldering. </p><p>2) It provides mechanical strength in case there is vibration or limited width or thickness PCB copper holding it down.</p><br>
<p>The glue is also used to prevent the can from oscillation. Many times over the years I have had customer complaints about hums or squeals coming from consumer appliances and electronics, only to find that a cap is making the noise. Most manufactures are aware of it and so they put the glue in place to prevent the stray oscillations. In the mid-80's to about the mid 90's when the transition from transformer / diode pack with very large capacitors gave way to switching power supplies, it was common to carry a hot-melt glue gun in the tool box to &quot;quite&quot; noisy power supplies from the first-gen switching power supplies.</p>
<p>PhenomenON. not 'phenomena' (which is plural) - just my two cents. professionalism please.</p>
<p>Thanks for the meter recommendation. I've found one on Aliexpress which comes with a plastic case at a good price. I've ordered one.</p><p><a href="http://www.aliexpress.com/item/86-Plastic-Shell-DIY-Multifunction-Transistor-Tester-Kit-For-LCR-ESR-Transistor-PWM-Signal-Generator-M328/32372954591.html" rel="nofollow">http://www.aliexpress.com/item/86-Plastic-Shell-DI...</a></p>
<p>That looks like the same as my kit. I just ordered the graphical kit. Because I do like the graphical display for checking transistors. I have a prebuilt graphical without the precision voltage reference. So I just want it all in one meter.</p>
<p>Is this the one you ordered? Looks like it measures capacitor voltage loss, which I gather the pre-built graphical versions don't.</p><p><a href="http://www.aliexpress.com/item/DIY-M12864-Graphics-Version-Transistor-Tester-Kit-LCR-ESR-PWM/32423936601.html" rel="nofollow">http://www.aliexpress.com/item/DIY-M12864-Graphics...</a></p>
<p>I saw one of these available on Amazon (not a kit) for about $20.00. Someone compared readings from it to a Fluke for various good capacitors and the cheap unit was not reliable. I also saw one for around $40.00 which had no complaints. <strong>Has anyone tried such an expeeriment with this unit?</strong></p>
<p>That same someone may have also not calibrated their instrument properly too. Plus there are so many AVR tester models available it is impossible to make a generic statement about them. The best ones appear quite accurate though when used properly.</p>
<p>You have a Multimeter which can measure 200mV AC above 80kHz?</p><p>You have almost everything to measure caps directly on the board. No just one thing missing: a SPS with a Transformer where you can fit one(!!) additional winding of wire around. I got 700mv AC from a 10Watt 10 Volt SPS. Connect parallel to the Multimeter probes an connect to the leads of the soldered in Capacitor. Simply compare to a known good one of approximately same size and voltage.</p><p>I use it all the time with suspicious and unsuspicious caps.</p><p>PS: 90 % of not good caps looked good for me, and bulged does'nt mean automatically damaged.</p><p>It's symply not economical to put 10-20$ worth of caps in an thing that nobody will even bid on at Ebay! Only replace real bad ones even with used good ones of junk electronics., because the other parts of jour device to repair are even as old as the bad caps you exchange.</p>
<p>The component tester you suggested looks great - and at an unbelievable price. I've just ordered one :)</p>
<p>Yeah I am getting into all of the kits lately myself. I just ordered the crystal tester and frequency meter kit. It was $7.45</p><p><a href="http://www.banggood.com/1Hz-50MHz-Crystal-Oscillator-Five-LED-Display-Frequency-Meter-Kit-p-959592.html" rel="nofollow">http://www.banggood.com/1Hz-50MHz-Crystal-Oscillat...</a></p><p>It is also based on a German hobbyist design. In the reviews there is a link to the site.</p><p> I am working my way up to one of the oscilloscope kits eventually. I've also built the DDS signal generator now too. It is pretty cool if you have an oscilloscope.</p><p><a href="http://www.banggood.com/DDS-Function-Signal-Generator-Module-DIY-Kit-Pulse-Sine-Wave-p-958215.html" rel="nofollow">http://www.banggood.com/DDS-Function-Signal-Genera...</a></p><p>I was building a homebrew function generator but this kit is way better than what I can make on my own.</p>
<p>In addition to Rubycon and Nichicon, United/Nippon Chemi-con capacitors are also very good. In my opnion, none of these three is any better or worse than any of the other two. </p>
<p>In testing a lot of my vintage capacitors here I have found that Rubycons spec the best. So there you have it. Rubys are the best.</p>
<p>I found an easy way to check a capacitor. It might work on caps with value 1mfd or higher. Connect the leads of a multimeter in DC voltage mode to the appropriate leads of the capacitor. Charge the capacitor within the rated voltage. If the capacitor is good there will be a sudden voltage drop and gradual rise as the capacitor charges. Also when the battery is disconnected it won't hold the voltage for at least a few seconds.</p>
To test my caps, I use a Fluke tester.
<p>Agreed-japanese capacitors are the best by a very long way- I doubt you'll ever see one with a domed top or leaking. The electrolyte is high quality where the taiwanese/chinese is poor and sometimes contaminated.</p><p>The good ones are: Chemicon (Marked with a Black empty rectangle, KZE, KZJ, or KZG on the body of the capacitor), Hitachi (hard to describe logo but a little like across between 3 concentric circles and the viewfinder of a rifle- sort of), Panasonic ( [M] for Matsushita ), Nichicon (marked NICHICON), Sanyo (marked SANYO), and Rubycon (marked Rubycon).</p><p>All are good- choose any and you'll be ok, but remember that there are different sizes and lifetimes available for them- try to match the original as far as possible.</p><p>Avoid Toshin Kogyo (TK marking) ones because although they are a Japanese company the capacitors are made in Taiwan.</p>
<p>I've saved my brother-in-law's tv by replacing a single cap, and also his old surround sound receiver (which finally kicked the bucket a couple years ago).</p>
<p>When you replace one cap, if there are others like it nearby, it is usually a good idea to replace them as well. Even if they are not completely shot the day you're in there. They're probably going to go pretty soon.</p>
<p>a bad cap was caused by a power surge on my Marantz 2275 amp.. caused a hum in the preamp, so I figured it was in the power supply... I found similar value caps, and started adding them in parallel to the existing caps and suddenly the hum went away. Another thing I noticed in this instance is the voltage measured across the cap (with a DC digital multimeter) was only about 10V and the cap was rated at 35V, the reason for this is the cap wasn't filtering the wave, the meter reads RMS (or close), but when a cap is good, RMS and the DC readings are the same</p>
<p>good pic of the samsung tv, i fix tvs and iv fixed hundreds of those lcds with just 1000uf caps.</p>
<p>Yeah Samsung had to do a recall over that issue. The service center I was supposed to deal with was hours away, so I just fixed my set myself.</p>
I fixed a Samsun I found on te side of the road with a simple capacitor from Radio Shack for under $8.00. Using in the spare bedroom now! Easy fix.
<p>I had a similar fault on my old HAMEG scope and it was a ROE Germany paper/poly capacitor on the EHT tripler The fault manifest as the signal on the screen to be able to show only the positive going part of the signal and the negative absent. Here the cap was short circuited. I have seen Elna electrolytics on a Sony pa amp to burn out a power IC that was dear at about $100.00 ten years ago!I had many well known makes that just go dead whilst the set was working that one was a job to diagnose the cap as it was healing itself time and time again</p>
<p>ELNA capacitors are widely counterfeited. So what you saw very well may not have been a failed ELNA at all. Although I would think that Sony could source good parts. But who knows? No electrolytic capacitor lasts forever. So even the best caps die eventually.</p>
<p>If you come across CapXon capacitors in any device then its best to replace them.Many millions were made with the wrong chemical formula and have been the cause and are still being the cause of many a power supply failure (Especially in Sky boxes) Sometimes when they fail all you can see is 2 legs sticking out of the pcb, the body of the capacitor has shot off somewhere like a rocket. Another type that can be quite easy to spot when they fail is the Tantalum variety. When I was working in electronics manufacturing years ago, tantalum caps would either go bang and or catch fire.</p>
<p>I think all of the first wave capacitor plague equipment is long dead, and buried by now. That was due to some failed industrial espionage I hear. I call Tantalum capacitors tantrum capacitors, because they lose their cool, and blow a fit. When a tantalum cap goes, unless it is written on the board, you may never know what you lost. Because they completely blow apart. Before they go they sweat silver blood. I've watched them do it.</p>
<p>I had a computer power supply just fail about an hour ago on my test machine. I plugged a card in the computer and hit the power switch. It ran up for a second and shut down. I pulled the card thinking it was defective and tried again. Same thing. I pulled the 24pin connector and plugged it into my power supply tester. Now the power supply worked, but after 2 seconds &gt;&gt;&gt;***POW***&lt;&lt;&lt; then it shut down. I pulled it and put another in its place and took it apart to see if I could find the bad cap before I recycle the thing. I found a few bulging, but nothing blown.</p>
<p>What is troubling is even before a PC power supply fails it can generate damaging ripple, that can in turn cause harm to the rest of your PC. You got lucky if your PC PSU just stopped working, as opposed to continuing to output dirty power. The simple fact is all of those capacitors are in products for good reason. By the brand choices hardware manufacturers make themselves they show just how resentful they are of putting any caps into their equipment. So if they could get away with using no caps at all they most certainly would. But they cannot, so instead they choose to use the cheapest ones they can find.</p><p>Bulging caps are blown caps too. Once the top gets round a cap is shot then too. Although caps can be shot and show no external visible signs.</p><p>So, what kinds of caps did you see in your PSU? Were they Capxon, or Teapo? Those are like some of the worst. They blow all of the time. There are people that recap new equipment and replace those with more reliable brands.</p>
<p>A &quot;POW&quot; is usually a shorted, then internally blown open, switching transistor... one of the big ones on the primary side heatsink.</p>
<p>I forgot to add in my last post, that Panasonic capacitors are a good make, up there with Rubycon and Chemicon. When choosing a capacitor for something like a power supply in your TV or monitor choose not only a low ESR type but also look for one that can function at 105 deg C or more. You won't find these at Maplin, but they can be easily obtained from CPC/Farnell, Digikey or other major stockists. Avoid the cheaper 85 Deg C ones if at all possible.</p>
<p>Don't just choose &quot;low ESR&quot; but instead only &quot;very low ESR&quot; or equivalently worded. All very low ESR caps are rated for 105C or higher, but higher than 105C is pretty rare for a very low ESR cap... not that they can't stand higher but it gets harder to determine lifespan.</p><p>Frankly, if you can find sufficient capacitance values at the voltage you need, you'd be better off using solid/polymer capacitors instead which last multiple times as long at elevated temperatures.</p>
<p>Very helpful! It just so happens I need to replace leaky capacitors in my old digital piano, which served me well for 20 years. Being a bit clueless when it comes to electronics, your Instructable was very informative. I will order those Rubycon caps to get another 20 years out of my piano.</p>
<p>For them to last that long they must not have been in a stressful circuit. You should do fine with one of several major brands like Nichicon, Rubycon, Panasonic, Elna, Sanyo, Mallory, Nippon (Chemcon (NCC)), etc.</p><p>Keep in mind that if they are in the audio path as coupling capacitors, that there are different parameters important to consider for best sound quality. Web search for best audio coupling capacitors to get some general info and recommendations.</p>
<p>Samsung LCD TVs from 7 to 8 years ago were notorious for <br>capacitors failing on the power supply board.The reason was they were using 15V electrolytic capacitors on a 15V <br>board.This is like driving your car on <br>the interstate all day in first gear. Mine failed so I replaced them with 35V caps and it&rsquo;s been working great <br>ever since.The same thing happened to <br>my Samsung computer monitor made around the same time.</p>
<p>Electrolytic capacitors have a high voltage tolerance, one seeing 15V and rated at 15V (which by itself is very weird because they're usually rated higher or lower like 12V or 16V) would not reduce its lifespan.</p><p>Lifespan reduction is generally due to exposure to heat along with a high ripple current leading to the ESR causing self heating.</p><p>FYI, capacitor failure is one of the leading causes of death of &quot;almost&quot; every TV and LCD computer monitor which uses electrolytic capacitors.</p>
<p>Capacitors can also be bulging at the bottom, the can lifts slightly from the legs. Basically the lower the rated voltage, the smaller physical size and more heat around it, the more likely the capacitor is to fail. </p>
<p>LoL. Thanks!</p>

About This Instructable




Bio: I was pfred1 but moved, changed my email address, and lost my password. I suppose worse things could happen.
More by pfred2:Cheap CNC Dos and Don'ts Pimping My Arduino Hotgluing Erasers to an Arduino 
Add instructable to: