Capacitors shorted? Is this normal?

[xboxgamer]

Hello everyone. My first post.

I have Xbox 360 console, I was trying to replace fans to more quiet units so I wanted to measure the fan header voltage to check which one is ground and which one is plus. Unfortunately my test lead slipped of the fan header pin and somehow I touched negative and positive pins together, there was loud spark and console shut down. The console starts now without any issues except that fans are not spinning but I already found the culprit, the transistor responsible for fans exploded and  base is shorted to collector and I need to replace the transistor.

But since I watched tons of videos about fixing laptops, Nintendo consoles and other devices on Youtube, I decided to measure a little bit on the motherboard to see if everything is OK apart from the transistor.

Question 1 - From all the videos I understand that only one leg of smd capacitor can have continuity to ground. Is this ALWAYS correct or there may be an exception of that rule?

Question 2 - Is it ALWAYS true that touching both legs of capacitor should only cause short beep and not a long one? (by long one I mean constant continuity).

The reason I am asking is that so far I have found at least 4 caps on the motherboard that seem to me that are shorted to ground and they beep constantly when checking continuity between 2 legs. Of course as a beginner I am not confident enough to remove them from the motherboard and then measure (yes I am aware that this is suggested) so I check everything in the circuit.

Now the confusing part is the console runs fine (except for the FAN not running). When I connected fans directly to 12V just to have any cooling I am able to start the game, play for hours without any glitches, copy files etc. Except of the fan I don't see ANY issues with the console.

Please help me understand that.

I also found the schematics/Diagram so please take a look at the images. The caps that are shorted are C4C6, C5C5, C5C6, C8B8 and probably some others but these 4 should be ok for now. Please help me understand how to determine the purpose of the capacitor from the diagram. Is it normal behavior I described above or the caps should be replaced?

[srb1954]

A continuity beeper is not the best way to check for shorts in capacitors while they are in circuit. Continuity testers generally beep even there is a moderately low resistance; they don't only beep for true short circuits. There may be low enough resistance in rest of the circuit across the capacitor to cause the continuity tester to beep while the capacitor itself is still perfectly OK.

The better way to test is to use the resistance ranges of your multi-meter and check for very low resistance, say less than 1 across the capacitor. If your multi-meter has a null or zero function use this to zero out the lead resistance to get a more accurate indication of very low resistances.

The above method is not entirely infallible as there are rare circuit configurations e.g. capacitor paralleled with an inductor, where you will measure extremely low resistance even though the the capacitor is OK.

If your console runs OK it is very unlikely that there are shorted capacitors on the supply lines.  Usually you would also measure the supply voltages to verify the circuit. A shorted capacitor on a supply line would drop the supply voltage to very close to zero and disable the console completely.

[helius]

Hello and welcome to the forum.

A DMM (multimeter) in ohms or continuity mode tries to find the resistance between its leads by using Ohm's Law (V=IR).
Some of them do this by acting as a voltage source and measuring current, while others work by acting as a current source and measuring voltage. The voltage or current they source can be AC or DC. With all the possible combinations, it's easy to see how the results of measuring in-circuit are hard to predict. That being said, the continuity mode is normally set up to make it possible to find wire connections in a circuit. To do this, the meter should source a low voltage (less than a volt) and measure current. By only sourcing a low voltage, it avoids turning on semiconductors like transistors or diodes. There is still the matter of DC or AC: most meters use a low DC voltage but both are possible.

The continuity mode is intended for finding wired connections. Is there a trace between Pin 9 of U13 and the anode of D12? It can answer these kinds of questions confidently. It is not intended for measuring components either in or out of circuit. First, because there is no parameter of a component that is measured by a beep. Second, because different measurement setups are required to find the values of components like capacitors. Thirdly, most capacitors are not easily tested in-circuit at all, since they are most often used to bypass the power supply near each chip, which puts them electrically in parallel with each other.

What is the resistance of a capacitor? The question is badly formed, because it's asking to make a measurement of the wrong type of component. When a DMM in resistance mode is put on a capacitor, it can't find the resistance because the relationship between V and I doesn't act like a constant resistance. You can see anything from 0, to rapidly changing values, to a reasonably steady (yet wrong) value. If the capacitance is quite large (or there are many caps in parallel), it is rather likely to be 0, since it behaves closer to an ideal capacitor. Remember, an ideal capacitor has no series resistance at all!

The ceramic capacitors you have circled are not commonly subject to failure. It is possible for this type of component to fail if subjected to physical impact or excessive board flex, but not from shorting a 12V supply. My recollection is that the fans respond to temperature, so the driving circuit may have a voltage or frequency coefficent of temperature.

Finally, I would have sensible expectations about the probability of a successful repair. Any kind of learning process begins with simple objects of study and progresses to more advanced topics. When you have developed the requisite knowledge and experience, it becomes possible to conquer complicated problems like repairing highly-miniaturized circuits. To minimize disappointment you should write off the xbox now, since the approach you're taking is not working (the orange wire in the picture above reminds me of cat playing the keyboard).

[xboxgamer]

Thank you both for the answers, I am not saying I am understanding everything perfectly but it gave me a better picture of how my testing was wrong.

But that unfortunately creates even more confusion now because there are many channels on Youtube of people who repair laptops/motherboards professionally and they all show the method described by me in the first post so they simply check if the capacitor is beeping and if it does then they either replace the cap itself or they replace the chip close to the cap.

Let me post 2 videos showing what I am talking about.
https://youtu.be/lpRgRDbhOPI?t=387
https://youtu.be/Xix7qzbOT6Q?t=327

So I find it contradicting that you gentlemen are saying that the methodology of testing is wrong or at least can generate false positives but the method is used by professional technicians who repair computers every day. Or maybe they use this method because they simply have more experience and know what to expect?


Regarding the Orange wire please ignore it. It was installed by professional who installed a modchip in my console called RGH which allows me to run Homebrew applications but this was years ago and it has nothing to do with my issue with Fans.

[srb1954]

Thank you both for the answers, I am not saying I am understanding everything perfectly but it gave me a better picture of how my testing was wrong.

But that unfortunately creates even more confusion now because there are many channels on Youtube of people who repair laptops/motherboards professionally and they all show the method described by me in the first post so they simply check if the capacitor is beeping and if it does then they either replace the cap itself or they replace the chip close to the cap.

Let me post 2 videos showing what I am talking about.
https://youtu.be/lpRgRDbhOPI?t=387
https://youtu.be/Xix7qzbOT6Q?t=327

So I find it contradicting that you gentlemen are saying that the methodology of testing is wrong or at least can generate false positives but the method is used by professional technicians who repair computers every day. Or maybe they use this method because they simply have more experience and know what to expect?


Regarding the Orange wire please ignore it. It was installed by professional who installed a modchip in my console called RGH which allows me to run Homebrew applications but this was years ago and it has nothing to do with my issue with Fans.

The reason they use the beeper is that is it quick and with "professional" repairers they want to get the job down in as few hours as possible. I put the work "professional" in quotes because I don't view technicians who carry out repairs in such a fashion as being all that professional. The procedure may actually successfully repair the fault in many cases but they might also disturb a latent fault so that it appears to have been repaired only for it to return at a later date.

There is a trade off between being able to quickly find a potential fault and the wasted time unnecessarily replacing false positive components but parts are generally cheap and if you have the right soldering gear and technique you can quickly change the part to check whether the fault has been resolved. The most important thing is to have the right equipment and necessary soldering experience because there is always a risk that the PCB is damaged during the repair process. Without sufficient experience and the right gear it is better to test, test and test again before taking the steps to replace a component.

Having a quick look at the first video I would suggest you don't follow his example for desoldering a capacitor. There is significant risk of lifting the pad off the PCB the way he is doing it.

[CaptDon]

It comes down to this, those caps are likely not shorted
but are sitting across a buss with a bunch of chips you
blew up with your screwdriver slip. Now the entire buss
and everything on it appear shorted. In terms of repair
cost Snap=$10 Crackle=$50 Pop=$100 and with all of
the unknowns that may be connected across a capacitor
no sensible technician 'beeps out' capacitors in circuit
unless you have a small fault in an isolated area, not the
entire power rail!! Yup, there are a lot of dumbass "here's
how to do it" videos on YouBoob etc.!!! A lot of bad info
from clueless people and posting as if they are the authority
on the subject. I always assume that I am wrong until my
testing and conclusions can prove I am right. 98% of the
time I find the correct failed component the first time unless
some want-to-be technician let the magic smoke out of
several components at once. Noob-Tube should be called
how to ruin any electrical device by watching this two
minute 'repair it yourself' video.

[xboxgamer]

It comes down to this, those caps are likely not shorted
but are sitting across a buss with a bunch of chips you
blew up with your screwdriver slip. Now the entire buss
and everything on it appear shorted. In terms of repair
cost Snap=$10 Crackle=$50 Pop=$100 and with all of
the unknowns that may be connected across a capacitor
no sensible technician 'beeps out' capacitors in circuit
unless you have a small fault in an isolated area, not the
entire power rail!! Yup, there are a lot of dumbass "here's
how to do it" videos on YouBoob etc.!!! A lot of bad info
from clueless people and posting as if they are the authority
on the subject. I always assume that I am wrong until my
testing and conclusions can prove I am right. 98% of the
time I find the correct failed component the first time unless
some want-to-be technician let the magic smoke out of
several components at once. Noob-Tube should be called
how to ruin any electrical device by watching this two
minute 'repair it yourself' video.

So looks like I am back to where I was. I know in my original post I asked about shorted capacitors and the response I received is that it's very unlikely that the caps are shorted because my console would not work. But after seeing this response I started to worry that the short may still exist on the power rail instead.

So let me repeat my question.

If Black cable of my multimeter is connected to ground and Red cable "beeps" on both sides of the capacitor, does that definitely mean there is a short to ground on the rail or (since my console seems to be running fine) this may be completely normal depending on the schematics and I am just trying to fix non-existing problem? In other words is it possible that the continuity to ground is acceptable in some instances. Because @CaptDon concluded that there is a short on the bus and this sounds scary.

[jdutky]

It is also possible that your meter is beeping only momentarily, which does NOT indicate that the capacitor is shorted, only that, for a small current, and for a small period of time, the capacitor "conducts" (as it "fills up" with charge). If you hold the probes on the cap for more than a second or two it is likely that the beep will stop.

Measuring capacitors with a multimeter is unlikely to produce meaningful results. Even if the meter has a capacitor setting, you really need to understand how it's doing the measurement in order to understand what the readings actually mean.

[xboxgamer]

It is also possible that your meter is beeping only momentarily, which does NOT indicate that the capacitor is shorted, only that, for a small current, and for a small period of time, the capacitor "conducts" (as it "fills up" with charge). If you hold the probes on the cap for more than a second or two it is likely that the beep will stop.

Measuring capacitors with a multimeter is unlikely to produce meaningful results. Even if the meter has a capacitor setting, you really need to understand how it's doing the measurement in order to understand what the readings actually mean.

The beep lasts far more than 2 seconds and doesn't stop until I remove the probe from the cap. I know about the charging capacitor because most of them on the motherboard act this way, they beep for a moment and then I can see a value increasing but the 4 mentioned caps in my first post just beep all the time.

But sorry to say this still this doesn't answer my question so I will copy it again.

If Black cable of my multimeter is connected to ground and Red cable "beeps" on both sides of the capacitor, does that definitely mean there is a short to ground on the rail or (since my console seems to be running fine) this may be completely normal depending on the schematics and I am just trying to fix non-existing problem? In other words is it possible that the continuity to ground is acceptable in some instances.

[wraper]

Low voltage high power rails of big chips may have normal resistance in single digit ohms or even fraction of ohms. Not something powered from linear Vreg, though.

[srb1954]

If Black cable of my multimeter is connected to ground and Red cable "beeps" on both sides of the capacitor, does that definitely mean there is a short to ground on the rail or (since my console seems to be running fine) this may be completely normal depending on the schematics and I am just trying to fix non-existing problem? In other words is it possible that the continuity to ground is acceptable in some instances.

The beeping from your multimeter is NOT an accurate way of checking for shorts and should not be relied upon as a sole diagnostic tool.

If you look closely at the second video that you referenced you will notice that that just about every point he touches produces a beep but, if you also look at the resistance readings on the multimeter, these are often several ohms or more when the beeper is active. This resistance reading is just the load of all the other circuitry on the power supply and is perfectly normal for a large digital circuit - it doesn't represent a short circuit in a capacitor or other device across the supply rails.

The short answer is not to rely on your ears but open your eyes and look at what the display on your multimeter is telling you. A true short circuit would indicate less than 1 resistance or less than a few mV if your meter continuity test function also tests diode voltages. 

[xboxgamer]

If Black cable of my multimeter is connected to ground and Red cable "beeps" on both sides of the capacitor, does that definitely mean there is a short to ground on the rail or (since my console seems to be running fine) this may be completely normal depending on the schematics and I am just trying to fix non-existing problem? In other words is it possible that the continuity to ground is acceptable in some instances.

The beeping from your multimeter is NOT an accurate way of checking for shorts and should not be relied upon as a sole diagnostic tool.

If you look closely at the second video that you referenced you will notice that that just about every point he touches produces a beep but, if you also look at the resistance readings on the multimeter, these are often several ohms or more when the beeper is active. This resistance reading is just the load of all the other circuitry on the power supply and is perfectly normal for a large digital circuit - it doesn't represent a short circuit in a capacitor or other device across the supply rails.

The short answer is not to rely on your ears but open your eyes and look at what the display on your multimeter is telling you. A true short circuit would indicate less than 1 resistance or less than a few mV if your meter continuity test function also tests diode voltages.

Ok sorry for the delay. Today I made some measurements with my multimeter set to resistance on the lowest setting 200.

Almost all the capacitors that beep have 0.5 Ohm to ground on one leg and 2.0 Ohm to ground on the other leg.
Only one capacitor had 0.5 Ohm and 4.0 Ohm respectively.

(BTW test leads touching together also show 0.5 Ohm)

Please note that no matter how long I keep measuring them, the value does not change and I don't see them charging.

What's confusing to me is that I measured a lot of capacitors on the board and some of them do not show constant value for resistance but I see the value raising so the capacitor is charging however some of them show very high constant resistance like 0.5 on one leg and 140 on the other leg.

I am not sure now, should capacitor always charge or this is normal that the value stays at 2 Ohm or 4 Ohm or even 140?

Meanwhile I've located a bunch of capacitors very close to each other on the other side of the board where the CPU (or GPU) is located and there is about 10 of them and they ALL beep and show 0.5 and 2.0 so I really start to believe now that it should be normal because it very unlikely that capacitors so close to CPU are shorted and console plays fine. But I would appreciate some explanation if the values I measured are looking normal and why some of them are charging and some of them have very low stable resistance.

[jdutky]

As has been said, testing capacitor in circuit is problematic. It is also very difficult to test devices in circuit when you don't have a schematic to work from, because you have no idea what other paths may connect the two nodes you are trying to measure between. All components in the system will have some characteristic resistance, even unbiased diodes and transistors (it will be in the mega-ohm range), so you can't even rely on having the thing powered off to completely isolate individual components.

I agree that you should not suspect that these caps are actually shorted if the device appears to be working correctly (except for the fans?). As it stands I don't think that there is much to be gained by poking around at the Xbox any further. You would be better served, if you are trying to learn more about electronics, to experiment with discrete, leaded components (much easier to handle than surface mount devices), or muck around with something for which you have schematics that show how things are connected (and from which you can reason about what can and can’t be tested effectively in circuit). You would also be well served to get an oscilloscope, as a multimeter, even a very good one, has significant limits on what it can measure.