And our readers reply….
First, let me thank all who graciously responded – there are simply too many to publicly acknowledge!
OK – so it turns out I was not so dumb after all. Replacing the 6.3 v, 1000 uf capacitor with the SAME uf rating, but at higher voltage, was OK – overkill, but OK. If I had replaced it with a LOWER voltage rating, it could have blown up – that I figured was a No-No. Replacing it with a higher uf rating, e.g. 2200 uf, could potentially de-stabilize the motherboard. Replacing it with a BETTER capacitor of the same rating would have been a better move (Radio Shack, as many pointed it, is not known for carrying quality components).
There MAY BE some advantage to upgrading capacitors used for filtering DC power, but replacing motherboard components can be a high risk undertaking. However, some are repairing older BX boards with success, so if you have a dead board, replacing a defective capacitor is definitely an option.
What follows are excerpts from some of the many emails I received – incredibly instructive! This amounts to a short course in capacitor usage. Please forgive me if I did not use yours, but I picked out representative texts:
Manufacturers choose a capacitor with a voltage rating that is equal to or
just greater than the voltage that must be withstood. This is done to
minimize the cost and volume of the capacitor.
The two primary measures of a capacitor are it’s capacitance (commonly
measured in uF or pF) and it’s rated voltage. The capacitance is the charge storage capacity. The higher the capacitance, the more charge the device can store. The rated voltage is the maximum voltage the capacitor can withstand before it fails or degrades substantially from it’s rated performance.
The capacitor that you replaced is an electrolytic capacitor commonly used
for DC power supply filtering. The filtering ability of a capacitor used in this manner is almost exclusively dependent upon it’s capacitance. The
voltage rating only determines what voltage it can withstand. Using a
higher voltage rating will not, in most cases, improve the filtering.
Using a higher capacitance value will, however, potentially improve the filtering and, hence, possibly the stability. The reason for this is that digital circuits introduce current transients in their power supply leads as their outputs switch. If the power is not adequately filtered, these transients can disrupt other circuits, causing errors or other undesirable behavior.
As I stated, the capacitance *predominately* determines the amount of
filtering. However, higher voltage ratings usually imply a lower ESR
(Equivalent Series Resistance). A capacitor, like any electronic device, is not ideal. A first order model of a capacitor is an ideal capacitor in series with a small resistance.
The small resistance reduces the “Q” or quality factor of the capacitor, thereby degrading it’s performance relative to an ideal capacitor of the same capacitance. In DC filtering applications ,like the one you replaced, the degradation is usually negligible. However, in power supply rectifier filtering applications, the ESR can be important as a low ESR will improve the ripple performance.
So, the capacitor you put in there will work just fine. A unit with a lower voltage rating, but higher capacitance rating, may improve stability if the manufacturer really skimped on the design and the filtering was marginal, but this is unlikely. Your biggest worry should be that you bought a capacitor from Radio Shack. Most RS components are of low quality and usually don’t last as long. However, since no one uses a mobo for more than a couple years this is of no consequence.
Cliff Chase – Chief Scientist – Klein Assoc, Inc.
Here’s a link on how electrolytic capacitors work…
ESR stands for Equivalent Series Resistance –
here is a good page on ESR
Capacitors are used for a variety of tasks; most of the bigger ones you see mounted on your motherboard have one sole purpose, and that is to guarantee that all the parts they are connected to will not be affected by high power drains of other components on the motherboard.
It works as follows: In a normal situation, all of the parts on your
motherboard such as the north-bridge, south-bridge, cpu, etc. are supplied
with same 5/12-volts power coming from your power supply, or a voltage
derived from that on the motherboard by its voltage regulators.
These voltages are distributed across your motherboard to all the parts that need to be powered. On strategic locations, large electrolytic capacitors are placed to ensure that during a high power drain of another motherboard component, the other areas are not affected, as the capacitor will continue supplying that area with sufficient power until its charge is depleted.
Substituting these capacitors with higher capacity ones will not harm you in any way, provided you solder them in properly and mind their polarity. On highly overclocked motherboards with excessive power drain on specific areas, you could actually benefit greatly with bigger caps, but usually only if the manufacturer indeed cut some corners during the design of the motherboard, or when you’re really pushing the outer limits.
For the ‘extreme’ overclockers, those pushing their AXIA’s to +1.5GHz, an
all out ‘modification’ of their motherboard’s power grid and their power
supply might really help out. I’d be happy to get into greater detail about this as I’ve done it on a number of configurations already yielding good results.
Sander Sassen – Chief Executive Officer, Hardware Analysis
NOTE: I’ve asked Sander if he could write this up.
Re using higher voltage rated capacitors due to
availability issues: There are better grades of capacitors,
that can sometimes benefit a circuit, but higher voltage
ratings rarely offer any improvement in commercial circuit
card assembly. Usually, the manufacturers have calculated
the necessary voltage rating to get the job done.
They then pad it up a little for good measure; i.e., in a +5V or +3.3V
circuit on a motherboard, they will use a 6-7V rated
capacitor. Using a 35V capacitor doesn’t add any quality
factor to the circuit. Should the voltage in those example
circuits rise up above the 6-7V rating of the capacitor,
scores of other active components would be damaged also, and
the capacitor would be the least of your concerns.
John “Hoot” Hill
Physically the capacitors are different, but they are the same size if you are talking about capacity. The reason for the extra size, is because the dielectric strength (electrical strength) has to be greater for higher voltages, hence the larger physical size.
In short capacitors are used in 3 modes:
- Parallel – they filter, and store energy. Replacing with a larger capacity will help
- Series – they block. Replacing with a larger size will pass signals not designed to pass, and cause instability of loss of operation.
- Critical values for series or parallel resonance – A specific value is required to make up a frequency determining network, in which resonance occurs. Replacing with a larger capacity will change the circuit and I guarantee it will not work.
Nigel Hulse – New Zealand
The limit on stability of motherboards is mainly in the wiring (i.e. crosstalk) and the chips themselves; you could go out and buy military spec replacement chips, it might help, it might not – certainly a *lot* of work though.
Another important specification you need to consider when changing
capacitors is the Ripple Current. It is the maximum amount of current the capacitor can provide. This is due to the impedance of the internal dielectric.
This is especially important if the capacitor is used in a switching power
supply. Looking at your diagram, this appears to be what your capacitor is used for. If you use a capacitor without enough ripple current, even if the capacitance and voltage ratings are the same, you can get an unstable,
unreliable power supply.
Now, a good quality 1000uf, 6.3V, electrolytic capacitor will have a ripple current of 750 – 1000ma, while general purpose capacitor will range between 350 – 500ma.
Ripple current usually goes up (due to the larger package size) when the
voltage rating is increased. A general purpose 1000uf, 35V electrolytic will have a ripple current of around 900ma. So by going to the larger capacitor you have compensated for the ripple current rating.
Colin Lee – Canada
Switching power supplies (like
we have on the Abit mobo) use special capacitors. These special caps
have a low ESR, which means they work better with the high frequency signals associated with switching PSUs. These special caps are tricky to get hold of and I wouldn’t expect to find them in Radio Shack. Normal caps have a 85C (degrees C) rating on the side. The special caps are rated 105C. I’m not sure what the temperature rating really means, but
it is a quick way to tell which type you have.
Capacitors have different specifications – operating temperature, hours of life, low impedance, tangent of loss angle etc. Many motherboards (at least ASUS) use high quality, low impedance capacitors that have
excellent high frequency characteristics; a common brand is Sanyo OS-CON. They supposedly age well and their properties do not change with time as they have different construction (organic semiconductor – OS) than the most common cheap aluminum electrolytics.
They are commonly used in DC-DC convertors or other types of power supply. If such a capacitor needs to be replaced, it should be by another that also has low impedance, and such are usually not sold in Radio
Shack. These capacitors are so good that they’re also sought by do-it-yourself audiophiles and they fetch a high price, although higher voltages (at least 15V) are required there, as opposed to 6.3V that’s usually used on motherboards.
A reasonably good replacement would be Panasonic FC or older HFQ line, that can be obtained from Digikey.
Andrew – Vancouver, Canada.
If you wanted to help improve the ripple in your power supply you could
trace down the circuit and increase the capacitance in the final filter
caps, as long as you don’t go nuts with it! i.e., replace a 1000uF with a
2200uF is OK, but I wouldn’t go much higher. The reason is with the
higher capacitance you would also have higher inrush current needs.
To get better stability, replace the capacitor with a higher
capacitance, i.e., instead of 1000uf, replace with 2000uf. Replacing a
lower voltage cap with a higher voltage cap does not necessarily
guarantee greater stability due to the replacement tolerance &
Now if you changed all your caps with 5% polarized poly caps…
And you’d better be a consistent solderer, able to work on micro circuit
etch runs, otherwise you’ll end up lifting signal traces, breaking
soldering eyelets/pads, bridging adjuncting circuits through poor
soldering, cold solder joints, uncut leads which can short to the
chassis, destroying the capacitor through heat, shorting circuits by
putting in caps which either have no plastic insulators or whose plastic
coatings do not go all the way down or get cut because the protective
plastic coating is too thin and it rubs against an adjoining circuit,
etc., et. al.
Yes, signal wise, there will be less ‘ringing’ at resonance with a
higher voltage cap, leading to less ‘noise’ & therefore supposedly
greater stability, but most manufacturers will build it to twice the
capacitance to reduce resistance/capacitance impedence resonance.
A greater voltage cap will be able to absorb voltage spikes better, but
this is only in circuits which are used for power supplies; if the
capacitor is used for tuning circuits, the capacitance should stay the
same, and if it is used in a transistor circuit it should not have the
voltage changed, otherwise the transistor may blow; in the case of a
filter cap, you are filtering noise (ac noise, 2nd harmonic ac or
switching noise, etc) and if you want 12v out, you use a 12v cap.
Does it make sense to replace capacitors on a motherboard with higher
But it does make sense to replace them with higher quality, closer
tolerance types (and they should all come from the same batch).
A higher Capacitance may help in smoothing out ripples in signals but care should be taken not to overdo it, because you may filter the desired signal rather than the noise, especially in high frequency applications such as motherboard FSB’s.
Remember the loop rule when traversing a circuit with
capacitors in them – charge times are very important. In other words, if it takes longer to fill up the capacitor, then you could cause some serious problems. Also drain is a big issue. All in all this is stuff best left up to the manufacturer. Replace one or two and you should be ok. More than that and things could get ugly.
For proper job, you need 105C for longer life and quality. Temps on
capacitors do get heated up from warm air and smoothing the voltages due
to frequency produced by the switching regulator in number of KHz. Caps
has to give power when CPU draws very sharp demands in milliseconds.
That’s what these capacitors with this specs is for.
I have seen so many motherboards and PSUs with swelled up tops or
exploded capacitors, both from heat when fans failed or low quality
capacitors – even on Asus boards too!! 10V is easier to get and
smaller, 1000uF or more but keep dia size same to fit better. The
capacitors is ganged up to get more uF capacity by parallelizing them
across same voltage plane and ground. If fat capacitor, stand away from
board the capacitor bit higher and bend the leads over to get low
profiles or make it fit to make room.
Those motherboards with swelled up or blown capacitors bounced back
after these capacitors were replaced. Even 105C capacitors do fail in
poorly cooled computers even in regular computers due to dead air areas.
Old, unused capacitors do go bad from age too, because seals on them is
not that 100%. It’s just a rubber or plastic plug inserted into can and
crimped especially on aluminum colored small SMD capacitors. This is
worst because most palm cameras and small machines use them. They
simply die very short life in few years of use, not just from heat.
High end capacitors have epoxy filled seals.
There is a HUGE difference between one electrolytic capacitor and another. Replacing all the decoupling/bypass capacitors on a
board with Aluminum Polymer capacitors, like Sanyo OS-CONs, could very well make a large difference in stability.
You can also keep the voltage rating the same and increase the capacitance
as well (provided that that capacitor is indeed there for decoupling
purposes and not to serve as one component of a low/hi-pass filter or in a
LC resonant circuit (etc, etc.)
By carefully studying the trace layout and the data sheet for the CPU +
chipset, you can fairly easily find out which capacitors are used for power supply decoupling, and upgrading those could prove to be very beneficial.
All the electrolytic capacitors on a motherboard are there for power
supply filtering, and their values were selected to be as small as possible while still doing the required amount of power filtering, so they are as cheap as possible. You can replace them with virtually any equal or larger value and things will work just fine – unless you go crazy and put in a big enough capacitor that charging it strains the power supply. A capacitor that big wouldn’t fit in the available space though.
One caveat: Since the frequencies being filtered are now high (and
getting higher as CPU speeds increase), the leads of the capacitor should be as short as possible. A pretty short lead can represent quite a bit of
inductance at high frequencies, which can destroy the effectiveness of an
otherwise good capacitor to filter properly.
Where that capacitor is part of any circuit that is responsible for clock speeds or timing, then there would be a negative effect, and it should be replaced with an identically rated component.
I would not suggest it. While the ability to run at a higher voltage might be good for some inside of the mobo, you will find that some will allow more voltage to go through them and fry other components when it releases its charge. I think you lucked out with the one you got, and I wouldn’t suggest changing out the rest of the components on your mobo.
Re replacing the caps on the motherboard with larger units, may I suggest a few things:
(1) A capacitor that is larger vis a vis power is kind of a good idea; generally one does not want to replace a capacitor with anything much larger than about twice the rated voltage of the system the
cap is installed in, or there will be a phenomenon known as “reverse leakage.”
(2) A larger capacitor requires a longer time to react. If, for example a large number of the capacitors were replaced in the motherboard with larger capacitors, you would probably see a slowdown, not huge but measurable. It takes a finite time for a capacitor to react, either to charge or discharge. If the capacitor is application specific, such as a debounce circuit, the larger cap may move the circuit out of tolerance and the system may be more prone to errors, lockups, etc.
(3) If you want to add capacitors to the motherboard, I would look only in the power supply area.
BTW, you might see a slight increase in speed if, when replacing the larger capacitors, place a smaller one in parallel with it. For example, if you have a 1000 uf capacitor at 12 volt rating, a 1 uf capacitor with the same rating in parallel with the first will shorten the charge/discharge time and theoretically improve the circuit.
At least all this works in audio power supplies and in speaker crossover design; *and* physics is physics.
PC mainboards are usually 5 to 6 layers thick – they have layers of conductive traces INSIDE the board as well.
If your soldering job wasn’t up to snuff, the solder may not have flowed
all the way through the hole and formed connections with an inside trace.
It’s also very easy to ruin a thin trace on the board with excessive heat.
Motherboards may have more than one layer of copper
within the full capacity of the epoxy board and
therefore someone may damage the board permanently if
370C hot air is used or if a solder gun is used on the
end leads of the capacitor itself.
However, the older
method you accidentally found out about is much, much
safer. Pulling the cap out, or even cutting it with a
knife, makes sure that the leads are still soldered on
the CORRECT layer of the board; furthermore, leave a
very good potential of mounting any cap you want on
Basically, a bigger capacitor doesn’t necessarily mean it will perform better. To have any chance of gaining better smoothing performance, you want greater capacity (1000uf, 2000uf etc) rather than greater voltage rating anyway. All the voltage rating does is signify the maximum voltage you can have across the capacitor before it will fail / blow, and it doesn’t mean a higher voltage rating will give better performance.
Plus not every X capacity, X voltage capacitor is the same; there are ones which are designed for high current drain, ones designed for fast reaction etc and there are cheap generic ones. The ones on your motherboard are most likely high quality capacitors, and simply replacing a bunch of them with generic higher capacity capacitors will probably lead to worse smoothing performance.
Gavin Froude BEng
While you can generally get away with such a
replacement, it’s better if you do a little homework
and find a cap with specs that match with regard to
ESR, dissipation factor, leakage current, and
impedance ratio. If you can match those up for both
original and replacement you’re golden.
Good places to buy obscure caps for finicky
replacement purposes like this include:
The above places have better than 95% of what you’ll
ever need in terms of electronic components.
O d d O n e
You also managed to replace an electrolytic with another electrolytic cap and get the polarity right, or else it would go poof in your face. Whether luck or skill, you did OK.
If you want to change capacitors on your mobo, at start you should check
what capacitors are used. You want to improve reliability/performance,
not make things worse, right? In general increasing capacitance lowers
ripple voltage, which is a good thing, but only if capacitor’s other
properties are also good.
You see all capacitors are not created equal. If your mobo already has a
set of good low impedance, high ripple current capacitors and you change
them to bigger general purpose capacitors, you might make things worse.
Capacitors on motherboard, Rubycon ZA-series 1000uF 6.3V.
This capacitor has max ripple current at max operation temperature
(105C,100kHz) of 1650mA and impedance of 24mOhm (at 20C, 100kHz)
Some general purpose, cheap-ass-capacitor, Phillips 037-series 2200uF 16V.
This capacitor has max ripple current, at max operation temperature
(85C,100Hz) of 1150mA and impedance of 150mOhm (at 20C, 10kHz)
As you can see from this (a bit extreme) example, good capacitor can have
lower impedance and higher ripple current at higher temperature and
frequency, even though capacitance is lower.
After you have found out what capacitors are used, you should check
manufacturer information about ripple current rating, impedance,
operating life under high temperature (capacitors tend to heat up in
high current and frequency situations), capacitance, dissipation factor
and voltage rating of course. You should be very carefull when
interpreting the information, especially impedance.
There are a lot of
general, or even special purpose capacitors, that manufacturers don’t
provide full information about impedance in high frequency cases.
Impedance and ripple current rating is a good measure of quality. Best
is to look capacitors meant for switched power circuits. Although when
capacitor has higher capacitance other properties are usually better too
but still you should remain vigilant.
Conclusion: Look for capacitors that are at least as good as capacitors
you are about to replace and get a next higher capacitance grade of
that. There also could be some issues about resonance/ringing if
capacitance is raised too high, so don’t over do it.
FYI: Take a look in Deja in the Soyo newsgroup. Look under the “dead
6BA+IV” thread. There are quite a few posts about the 1000, 6.3 caps
blowing on some of the 6ba+III’s and IV’s. Folks have replaced them with
higher rated caps and raised them from the dead. There is also some technical reasons for blowing (oc puts more current through) and replacing with larger.
I have a Soyo BA+IV motherboard, running a
PIII 700 @ 1085MHz (155FSB). After running this
system a long time, it suddenly died.
I checked around in the newsgroups since I had heard
other people with dead BA+IV.
Most of the people RMA’d their boards to Soyo
and got new ones, but I had no warranty left…
After checking every component in the system, I found 2
“pregnant” capacitors (close to the voltage regulator),
one of them leaking electrolyte. I thought about replacing
them, and so I did: Old ones: 6.3V/1000uF; New ones: 10V/1000uF.
And the system just booted up fine, everything was OK
I posted my experiences in the newsgroup and more and
more people begun fixing their dead boards. At least 10
boards have been fixed.
I recently went buying a Abit BE6II 2.0 and checked out Abit’s
newsgroup. I found post here too with dead motherboards.
Many of them had also managed to replace their caps.
My conclusion is that these older BX boards, officially supporting
only 800 MHz, don’t handle that much current when the speed
is around 1GHz, with or without higher Vcore.
Most of the dead boards have been overclocked, but some of them
have been running at default speed.
To make a difference, you would have had to choose say a
2200uF 6.3V or higher cap. If you had went
with a higher uF rated cap the answer is maybe it would have helped. As a rule of thumb, more capacitance will equal less voltage ripple on the output of the power supply. It can also change the effective filtering frequency, but USUALLY that is not a concern….although it can screw
up some switching supplies.
The only concerns in this particular case are that the capacitor may have different loss characteristics than the one that it replaces, which can cause problems in regulator feedback loops or its high frequency characteristics may not be as good as the one it replaces. This leads to excessive ripple and noise in switching regulator circuits. These are both very critical for the capacitors used in the low voltage CPU voltage regulators.
This is why it is common to see 10 capacitors in parallel for the CPU power instead of just 1 capacitor with 10 times the size. The important characteristic in this case is not capacitance in microfarads, but ESR (equivalent series resistance) and ESL
(equivalent series inductance) which have a direct effect on the critical
voltage ripple that the cpu sees.
The picture you have shows 1 input and 1 output capacitor for that particular regulator, so ESR and ESL probably were not important in that particular circuit and your replacement will work fine.
As far as making a motherboard more stable: It is possible that replacing the existing capacitors with better ones will make the board more stable, but if the board was designed and built correctly in the first place, the improvement would be marginal. The replacement capacitors should be the same size or larger and should have better ESR and ESL characteristics.
The most likely use of a cap in a mobo (especially a large one like your
1000uf cap)is to stabilize the voltage. The cap will store voltage when the supply voltage increases over the cap’s stored voltage, or if there is
constant significant variation, when the supply voltage is over the average voltage. This pulls the surplus electricity away from the load.
the supply voltage is lower than the cap voltage, etc., the cap will
discharge voltage into the load. This adds the surplus electricity when the supply voltage is insufficient. The end result is for the cap to make a “noisy” or inconsistent power supply and make it “cleaner” or more
consistent. This effect is also useful when the load varies its power demand like a computer.
The uf is simply a measure of how much power the cap can store, and
therefore, how much it will clean the power signal. If the cap is used for this purpose, increasing the uf will make the board more stable.
However, the cap’s responses to variable voltage is used for other purposes, most notably, having to do with frequencies. For example, in the previous example, the cap is in parallel with the power supply and the load, but if the cap is put in series or “in line”, it will have the opposite effect and block the DC voltage while passing the AC voltage, which is the opposite of the previous example and what makes sound amplifiers work. In most cases, increasing the uf will also make the cap better at this function.
Another example of a frequency purpose is that, many timing circuits use
caps as a timing devise. Given an specific load resistance, the larger the uf, the longer it takes to charge the cap. (The internal capacitance of CPU’s is one of the limiting factors for how fast it can be clocked; and in fact, if one has a CPU that will not post when cold, it is probably because of this effect. The higher the charged area, the higher the capacitance; therefore, shrinking the size of the transistors on the CPU will lower capacitance and raise attainable clock rates.)
If one changes the uf rating, then one is going to change the speed of the timing device. For example, a 133 MHz timer would operate at a lower MHz. However, caps are not the only way to time a chip and I doubt many timer caps are soldered to a mobo. Anyway, the upshot is that changing this type of cap will change the system timer and may even disable the board.
Therefore, yes, increasing the uf may make the board more stable. However, given the possibility of ruining the mobo with one’s solder iron or changing important functions, I have never attempted it. In addition, since I try to buy high quality mobos, I doubt I would get much in the way of stability in any event.
My recommendation: Replacing the original capacitor with higher rating is good, but up to a point only. I usually have a rule of thumb, where I would never replace with more than 30% higher rating than the original value. The voltage rating on the caps also determines somewhat the life of the capacitors. All capacitors lose capacitance overtime due to drying
out of the internal chemical.
So when I find a bad cap, I usually replace with 30% more on both values. Keep in mind that this theory only work with filtering supply capacitors only. For other purpose caps, like high or low frequency shunt, capacitor of the original value is recommended.
As a rule, the guys who design these boards are far smarter than the average bear, so don’t go screwing with their designs. If you replaced the 1000uF capacitor with a 2000uF capacitor, you could easily trash the signal characteristics of something.
If you are thinking about changing the capacitors on the motherboard (around the CPU socket, near the DIMMS, and so on) to increase stability, it may very well be possible. The main reason that those caps are there is to prevent voltage dips caused by large surges in current (very similar to the usage of large caps in car audio applications –
ever see people’s headlights dim when their bass hits?).
We would get a large current surge whenever the PC goes from doing nothing to full load (ie idle in WinBlows then firing up Prime95). Without those caps, the CPU or whatever would increase its current draw, and that sharp increase would cause a dip in the available voltage, likely causing your processor or RAM to freak and the PC to lock up solid.
Your original cap there looks to be part of a voltage regulator circuit for the DIMM slots. Increasing the uf value (capacitance)
might add some stability to the system, a lot like the fact that the majority of motherboards that are the most stable usually have a few really large caps (or a lot of smaller caps) around the CPU / DIMM slots than boards that aren’t as stable.
Chris Swires – Electronics Technician for our most wonderful United States Navy.
Replacing a 1000uF, 6.3V cap with a 1000uF, 35V cap will have little or no effect. The 35 Volt capacitor is much larger because of the beefier insulation required to achieve the higher voltage rating. Generally, manufacturers will choose a voltage rating of 150% of the maximum DC voltage expected to be on the line that they wish to filter as a safety margin. Of course, a volume manufacturer could save a ton of bucks by disregarding that rule, and it’s often seen in practice.
A 6 Volt, 2000uf capacitor would be, say 1.5 times the size of the 1000uF unit (the plates are wound in a spiral, so the pi X diameter rule comes into effect here) and would offer better filtering locally, possibly increasing stability.
HOWEVER… replacing all of the electrolytics on an MB with ones say, twice the capacitance, would result in a smoother DC voltage at the nodes in question, but the extra surge current (double) required to initially charge the capacitors upon power up could cook your power supply, or even fry circuit board traces.
WHEW! If you read all of this, I am impressed! Terrific contributions from some of our great readers – Thanks All!
SUMMARY: Sometimes a stupid error opens your eyes…
OK – you’re probably wondering why this picture. A little background:
This is a shot of my ABIT BX6-2 (arguably the last great ABIT mobo), upper right hand corner. The interesting thing about the picture is the large capacitor – it’s a replacement I soldered onto the board; the original one was the same size as the smaller one in the pic. I knocked the original one off when I was testing Kingwin’s Mobile Hard Drive Rack (ALWAYS look before you leap).
Now, the original capacitor was rated at 1000 uf, 6.3 volts. After I knocked it off, I hot-footed over to Radio Shack and, after looking at all the capacitors, found a 35 volt, 1000 uf one. Figuring it was close, I bought it and proceeded to solder it in place.
With some trepidation, I booted it up and it worked fine (it DOES NOT work without it – all these parts are there for a reason). Now having done this and not being a EE, I asked around about if replacing a capacitor with a larger one was a good thing or not; apparently, it COULD BE a good thing, as it may add more stability to the board.
So, I am asking a question here:
If some of you folks who have circuit or motherboard design smarts could email me, I’d really appreciate it.