A Simple Question about Capacitors

Thread Starter

AMVanquish

Joined Jul 7, 2012
5
My 42" LED tv stopped working a few days ago, after a storm caused a power surge. The tv is now stuck in standby and will not turn on.

Since I don't have an ESR/capacitor meter, I have decided to replace all of the capacitors (total cost plus shipping: $19) on the mainboard to see if doing so will make the TV work again. (In case anyone is really interested, I'm replacing the OEM Lelon capacitors with new Nichicon capacitors, which are rumored to last 7-15 years.)

As an admittedly limited test, I used a multimeter (ohms test) to see if any of the capacitors died. A few of them did what I read they were supposed to do: they charged when I put the probes one way, and discharged when I reversed the probes. Four of them, though, seemed to hold a charge regardless of how I positioned the probes. Neg to Pos, Pos to Neg did not make a difference. These four caps simply charged up to a specific value and remained there.

My question: are capacitors supposed to behave in this manner, or are these four capacitors dead? Any and all help is greatly appreciated.
 
Last edited:

MrChips

Joined Oct 2, 2009
30,708
I use an analog meter set to the resistance scale for testing capacitors.

Of course, the response of the meter needle will depend on the value of the capacitance.

For anything under 1μF, basically you should be looking for an open circuit (∞ resistance) on the highest resistance scale.

Between 1 and 100μF you should see the needle bounce up everytime you reverse the leads, again settling down to ∞ resistance.

Above 100μF the time constant becomes noticeable and it takes the current a bit longer to settle down to zero (high resistance reading). You may want to switch the meter to a lower resistance scale.
 

WBahn

Joined Mar 31, 2012
29,978
As an admittedly limited test, I used a multimeter (ohms test) to see if any of the capacitors died. A few of them did what I read they were supposed to do: they charged when I put the probes one way, and discharged when I reversed the probes. Four of them, though, seemed to hold a charge regardless of how I positioned the probes. Neg to Pos, Pos to Neg did not make a difference. These four caps simply charged up to a specific value and remained there.

My question: are capacitors supposed to behave in this manner, or are these four capacitors dead? Any and all help is greatly appreciated.
By "charged up to a specific value" do you mean that the needle when from indicating a very high resistance (just prior to connecting the problem) to a low resistance and then back down to a higher resistance and then stayed indicating that resistance? Or does the needle go pretty much straight to the resistance it stays at?

The fact that it is staying at a constant resistance value means that there is leakage current flowing through the capacitor. Unless these are quite large capacitors and unless the indicated resistance is also quite large, these caps are probably bad (large electrolytic caps can have noticeable leakage currents even when new).
 

Thread Starter

AMVanquish

Joined Jul 7, 2012
5
By "charged up to a specific value" do you mean that the needle when from indicating a very high resistance (just prior to connecting the problem) to a low resistance and then back down to a higher resistance and then stayed indicating that resistance? Or does the needle go pretty much straight to the resistance it stays at?

The fact that it is staying at a constant resistance value means that there is leakage current flowing through the capacitor. Unless these are quite large capacitors and unless the indicated resistance is also quite large, these caps are probably bad (large electrolytic caps can have noticeable leakage currents even when new).
MrChips and WBahn, thank you both for your replies.

WBahn, the meter indeed went straight to a single resistance value and remained there regardless of how I connected the probes. One of the four caps, a 100uf 25v, reads 00.9 regardless of how I connect the probes. The three other caps do exactly the same thing: I connect the probes, they "charge" to a specific value, and then remain there. I am ignorant of the values that constitute a large capacitor--is it correct to assume that capacitors in the 100uf-220uf range are small?
 

WBahn

Joined Mar 31, 2012
29,978
MrChips and WBahn, thank you both for your replies.

WBahn, the meter indeed went straight to a single resistance value and remained there regardless of how I connected the probes. One of the four caps, a 100uf 25v, reads 00.9 regardless of how I connect the probes. The three other caps do exactly the same thing: I connect the probes, they "charge" to a specific value, and then remain there. I am ignorant of the values that constitute a large capacitor--is it correct to assume that capacitors in the 100uf-220uf range are small?
Reads 00.9 what? Ohms, kilohms?

Keep in mind that your meter is not reading the amount of "charge", but rather the current that is flowing. In a capacitor with no leakage, there is a strong relationship between the two (coupled with knowledge of the meter's output voltage and impedance).

I seem to recall that the leakage current of a typical aluminum electrolytic is about 10A/F, so a 220uF cap might have a leakage current of 2.2mA. What that equates to with your meter depends on its output characteristics and scaling factors, but it almost certainly going to equate to a fairly low resistance. For example, if your meter puts out 6V open circuit and 3V results in half-scale deflection, then if a leakage of this amount resulted in a half-scale deflection it would equate to about 1.4kΩ.

You can either measure the leakage current directly or you can back into it by looking up the specs on your meter's resistance ranges or measuring them with another meter. Do NOT measure the meter's output impedance with an ohmmeter! Instead, measure the open circuit voltage and the voltage across a resistor that puts the meter at half scale.
 

Thread Starter

AMVanquish

Joined Jul 7, 2012
5
WBahn, many thanks again for your replies. I had the multimeter (a Sperry DM-350A) set to 200k. According to its specs, the meter's accuracy is rated 1.5%, with a 100m OHM resolution.
 

WBahn

Joined Mar 31, 2012
29,978
This doesn't help me too much. Can you find (or measure with another meter) what the open circuit voltage is when on that range? And then what would the meter read if the needle were deflected half way?

Also, I still don't know what you meant by 00.9. What is that as a resistance?
 
Top