Replacing a polymer capacitor (I got lucky)

I broke what I've found to be a polymer capacitor off my motherboard. It looks like this:

More accurately I broke it off a pair of wires which lead to the original spot it was placed. So my situation is really not too desperate.
I'm good enough to solder it onboard if I had to anyway.

I've been able to judge that its 100uf and I think 25v. I'm going by the article the above photo is from which states the pictured capacitors are 560uf and 4v.

My concern is, the pictured capacitors simply have a "4" to denote 4 volts, while mine has "25A".

My other question... I can use one rated for higher voltage, correct? In fact there's a chance it will make my motherboard voltages more stable (when upgrading more than 30 watts beyond specification)?
Maybe I will kill a few birds here, if I can find a 100v 100uf to salvage I don't see the harm in soldering it on.

Thanks for any comments, advice etc.

 

WOW, I am amazed, I just found 3 identical caps on what I think is an HP dv4000. The first number is different, apparently thats the manufacturing date so I guess I should be OK.

 

I don't think its 25V, there are no components inside motherboard uses that much voltage, except for maybe the input from power adapter which is probably around 19-20V. For other areas it should be lower.

Yes you could use higher voltage, and those Sanyo OSCON SEPC is one of the best capacitor for digital circuits but their voltage is low. I suspect the original voltage is 2.5V and its situated on CPU PWM which is only producing whatever your CPU voltage is, usually around 1V. Make sure you know the area where you replace those caps or it might blow up, solid caps blows with more bang than regular electrolytic caps.

Don't buy too high of voltage, the supplied voltage is going to be the same and it won't damage the circuits in any way but the equivalent caps for higher voltage is always bigger and you might not fit in it.

 

I think its for supplying power to the MXM graphics slot.

It already doesn't fit when a MXM 3.0 B card is installed, which is why there are wires running from the original soldered spot to a different location. Maybe why it broke off.

Any way you can confirm its 2.5v? I do have some of those purple SEPC I could salvage although I could not replace all of them.

 

You could always temporarily connect a 100v cap and monitor the voltage with a multimeter.. Then make an informed decision on the most suitable cap from that. good luck tinkering

 

100V is excessive, 16V is good enough and measure voltage across that.

 

Hmm Thanks, this is good information.
I'm still not convinced 25a means 2.5v so I'll use an identical one.

Maybe you can't answer this, but I'm wondering, this laptop was designed for a 90w adapter and 35w graphics cards... if I put a 75w card in it... do you think the voltage over that cap and some VRMs will increase beyond 2.5v? Or would the amperage increase?

 

this is not standard labeling so you can not tell anything besides which end is positive and which is negative, and that the capacitance may be 5.6x10^(?)

if you broke off one like that then you could do couple things:
- if there are still connections then you can measure what capacitance that cap had, with a multimeter;
- you could google what is says on top of it and try to find out what cap it was;
- in a rare case there may be marks on the board itself that would say what cap it was;

other things:
- there's no standard 4V capacitors out there. There are 10V, 16V, 25V, 35V, 50V, 63V, 100V, ... This value specifies up to what voltage the capacitor is good to use, or else it would leak charge between the internal foil and would ultimately burn out/explode/leak out.
- the more voltage the bigger the capacitor is physically, but you can use it as long as it fits. It's actually better to use higher voltage cap as it would have bigger margin of error in case something happens.
- if the capacitor is not used in a resonance circuit (as in with radio transmitting circuits) then its actual capacitance value is most likely not so important - meaning if you put the next bigger in value cap in there then all will most likely work just fine. This is actually how capacitors are chosen - you calculate something, then put the nearest value cap available. Not to say how much electrolyte capacitors vary in capacitance from the label ... lol.

good luck

 

post pics of the actual cap...

 

Basically that is what I did, google it which led me to believe there was somewhat standard labeling. So do you disagree completely with this article?
Capacitor Lab - Types of Capacitors - Polymer Capacitors

Its not electrolyte, its polymer, that article is pretty clear and from a site with capacitor in its name, lol.
Some caps like this on another motherboard clearly say 20v, is that wrong?

BTW how might I be able to tell if it was part of the radio circuit or not?
*stares at motherboard*

In any case I think it will be fine as long as the 1st number is really the date code since the replacement I have is identical except for that.

@Indrek: do you know how to calculate the amperage a cap is rated for at a given voltage?

 

let me just state that this article was made by somebody that is user at some forum ... ummmm ok.

besides, I did read somewhere in there that if you're about to replace a polymer capacitor with an electrolyte one then you should double (or half, cant remember) the value. This is big BS to me. The value of a capacitor sits there just for one reason - to tell you what capacitance that capacitor has, i.e. what charge it can hold.

further a capacitor doesnt have amperage. Well it has maximum safe discharge current rated (in amps) but that's not of main concern as there will always be a resistor (or other resistance) on any circuit that will limit the current being discharged from the capacitor. Or else it will produce spark and could/will burn something.

back in the day when I was a student we found some big capacitor (as big as a beer bottle) and decided to charge it and short it out. So we did that on a metal rail that was next to the stairs of a school building and it pretty much welded itself to the rail, lol. We had to came back later to cut it off as we couldnt disconnect it otherwise. That was some 15 years ago when big capacitors were pretty rare and expensive.

P.S. just replace it with what you found as a good match

 

Haha. And what information is currently being exchanged here that does not involve people on a forum?

Good thing I'm not replacing polymer with electrolytic.

Just one more silly question, although I'm instinctively inclined to be good at these things D) I never took electronics class. Does more uf hold more voltage or is it reverse?

I might replace it with a 25v 220uf cap.
I hope you're not telling me to replace it with whatever I want just to see if I do it and it explodes in my face.

Thanks for your help.

 

uF has nothing to do with voltage, it represents the "amount" of charge the capacitor can hold.

For example if you have a small light bulb that works on say 5Volts, and you have a capacitor that is 1uF/10V, lets say that you charge that capacitor in full from 5V power supply and then use the cap to power your bulb, then the bulb will work for say 1 sec before the capacitor discharges and your bulb dies. If you have 2uF/10V cap then your bulb would work for 2 sec.

if you have a bulb that works on 15V and you try to charge your 1uF/10V from 15V power supply for the same purpose, then your capacitor may blow up while you try to charge it. If you try to do the same over 120V then the cap will most certainly blow up

and I think you said you found almost the same cap ? Easy way would be just to measure it with some multimeter.

 

These are Sanyo Os-Con capacitors, the value is 560uf with a 4 volt rating.

Here is the spec sheet for these guys:

http://edc.sanyo.com/pdf/e_oscon.pdf


Capacitors do have an ripple current rating which is an amperage rating.
This rating is mostly based on the ESR or series resistance of the capacitor and its temperature rating. Since these Os-Cons are some of the best out there that is why they are used in the high current supplies you see in laptops.

Typically I would no lower the amount of capacitance or the type of capacitance as it can upset the performance of the regulator, specially in the high performance supplies we are dealing with. you probably can add capacitance without to much worry, but again you want to stay in the same class of capacitor.


While the explanation of charge is simplistic it gets the point across.

 

I had a similar cap fall off a server PC motherboard. I could never find where it went, but it seems to run fine without it.

I'm not advising that you try running it without, but the reality is that it will most likely work regardless.

 

Really??

I mean I've run graphics cards that had a few missing surface mount components and they worked perfectly. But this isn't part of the power supply.

Its true tho I have been amazed in the past, now 2 high end MXM cards I've seen run perfectly with a missing component - right on the other side of the core.
Makes you wonder how much redundancy is going on and considering the amount of failures - how much it works.

Heck I'll go try it right now without the cap.

 

Interesting,,In a forum I was in read about several of them changing all the capacitors to get better functionality on certain types of mobo's made by certain vendors.

Cheers
3Fees

 

Yea lets get that specific I have an Acer 7738g which will run a q9200 but could also run a 6990m. What should I resolder with what?

 

Well, you were all wrong. Can't get it to work. Tried 3 different but similar caps.

My only choice now is to try and solder to the tiny broken leads of the original.

 

Niff, you are hardcore...that's all i have to say

 

Haha, PLL soldering is really a lot more difficult, those guys are hardcore.

I've gotten a lot better but I'm still not sure if I would have success with a PLL mod.

If I am hardcore though its because I like to replace LCD backlights.

 

no trust me...you are hardcore

As soon as someone even attempts to re-solder caps, they're in a different league

 

Well, someone had already professionally desoldered it from the board and connected it with wires. Maybe you missed that part.

And, attempts are one thing, results are another

All the same, check it:

A bit hard to tell the solder from reflections on the flux from the flash, but the solder is actually a little less shiny.
Didn't realize how much the thick wire was holding me back so I soldered 2 smaller wires on.
Anyway I'm about to try it.

 

See I can solder but am not hardcore, no change at all
Obviously I don't know whats really going on. I can only imagine that I burnt something else out with another capacitor.

The power button LEDs show a very very dim light when I try to power it on.

At least there is one good thing, if I get a new motherboard I wont have to go through all this because none of the components really needed to be moved for MXM type B cards they just work at an angle.

Anyone have ideas of what might have happened?

 

Hey man, a couple of things:

1. The board was designed with a 120W adaptor in mind (q9000 + GT240M + 2 HDDs + decent fan + 17.3" screen)

2. The capacitors do need to be removed otherwise the board would sit on the caps at a much greater angle (would never be able to put the cover back on) and they present a shorting risk.

Are you sure you always had the cap the right way round? It's always safe to over-rate a cap, dangerous to under rate it.

 

well I tend to disagree with that ...

generally you dont want something to hold bigger charge than needed, as it can cause damage at the discharge time, and/or with the longer charging time needed.

Either way, I'm not sure how the whole thing went into this situation, but are you sure the capacitor is your only problem?

 

I meant over rated in terms of voltage.

 

Something else has now become the problem. I might get it repaired.

And.... my m860tu died. I really hope its a coincidence. I'm going to have to put something else together to start testing the RAM I had in it... (talk about a virus, imagine killing 4 motherboards by testing bad RAM...)

Hope I can go through the 3 stages of loss or whatever as easily as I typed those emoticons there. lol