Welcome guest. Before posting on our computer help forum, you must register. Click here it's easy and free.

Author Topic: Cap replacement advice  (Read 4581 times)

0 Members and 1 Guest are viewing this topic.

comda

    Topic Starter


    Adviser
  • Thanked: 6
    • Yes
  • Experience: Experienced
  • OS: Windows XP
Cap replacement advice
« on: October 07, 2017, 09:59:50 AM »
Greetings,

Got an Old P3 board that needs some new caps. All the board does it flash its keyboard lights once every few seconds (i think its power cycling)

Anyways. the board has 6.3V 2700uf caps and my replacement caps are 6.3V 3300uf. Was wondering if this is acceptable. I have a buddy telling me yes, but wouldn't mind a second opinion. Thanks

Salmon Trout

  • Guest
Re: Cap replacement advice
« Reply #1 on: October 07, 2017, 05:14:47 PM »
How do you know it's only the "caps" that need replacing?


DaveLembke



    Sage
  • Thanked: 662
  • Certifications: List
  • Computer: Specs
  • Experience: Expert
  • OS: Windows 10
Re: Cap replacement advice
« Reply #2 on: October 07, 2017, 05:19:03 PM »
3300uF caps should work, but they will be a little slower at reaching 100% charge capacity and it will be a slightly heavier amperage draw to onboard voltage regulators ( aka VRM's ) and power supply until they level out at the voltage constant. If there is a ripple problem or a demand on this cap, it will hold up stronger than a 2700uF cap in that it carries 600uF more charge capacity as the easiest way to explain it to non Electronics Technicians.

I personally would go with exact same value components vs going with larger storage capacitors. * If you have a strong power supply the 600uF difference shouldnt be a problem, but if its a weak power supply it could cause an excessive current draw at boot that could lead to problems if the voltage rail isnt able to hold steady at a strong voltage with the extra current draw at boot. To charge the additional 600uF per capacitor the unknown is if your VRM's or power supply will be able to carry the heavier initial current draw at boot. These caps should get to full capacity within about a second of boot and voltage should hold strong enough to not affect the motherboard boot. But there is a small chance that you stress a VRM or the power supply and cause more damage.

I have performed the capacitor replacements before and easiest method is to tug at the can and remove it off the board leaving the legs in place. Then solder a new capacitor to the legs of the old. You might need to add additional flux to get it to flow to the legs of the old cap as for the electrolyte oil can make for a nasty soldering. The flux helps get the tin to flow. I have even made a capacitor daughter board with wires between motherboard and daughter board where replacement capacitors would conflict with a socket 478 heatsink for example. I poured hot glue on the bottom side of the daughter board PCB to seat and insulate all the capacitors so that they would never short to ground and I attached the daughter board to the inside of the case in a location that wasnt conflicting with limited space in the SFF ( small Form factor ) case. Also I never changed values for uF ( aka Microfarads ). Whatever value came out went back in.

Additionally Capacitors have Polarity and if you get the polarity backwards they EXPLODE!

I never bothered saving a capacitor plague Pentium III board but have saved a few Pentium 4's back when the capacitor plague was really bad and systems were dropping like stones and HP and Dell didnt have a recall to cover the poisoned electrolyte recipe capacitor problem so people were stuck buying new computers or having someone skilled replace the capacitors. Or migrate the CPU and RAM to a new motherboard that hopefully doesnt have the bad capacitors in it.

Did any of the capactors on that board short and explode blowing off the capactor body or did they just swell and vent out the electrolyte?

I have gotten boards that are bricked by the capacitor plague to boot by removing all the damaged capacitors around the CPU which is where I am guessing your capacitors are located. If it boots with them removed then its worthy of adding new capacitors. If it still acts the same then its a bigger problem then the capacitors if symptoms dont change. Running the system without these capacitors will cause the system to run in a unfiltered power state powered directly unfiltered from the VRM's and power supply and while it may boot now, do not run it for any length of time as for it might boot and look healthy but the minute you go to run something in which the CPU demands a heavier current draw and the capacitors aren't there to assist the system is likely to crash with an error or freeze up on you.

Salmon Trout

  • Guest
Re: Cap replacement advice
« Reply #3 on: October 07, 2017, 05:24:53 PM »
It is is a common misconception by people that don't know zilch about electronics that anything wrong with a device can be fixed by changing a couple of capacitors or better still replacing all of them. I have seen this with vintage hi-fi, plasma and LCD TVs. Their pal told them "it's the caps" and, mostly, if they manage not to butcher the board with the iron, it still doesn't work.

patio

  • Moderator


  • Genius
  • Maud' Dib
  • Thanked: 1769
    • Yes
  • Experience: Beginner
  • OS: Windows 7
Re: Cap replacement advice
« Reply #4 on: October 07, 2017, 05:59:55 PM »
+ 1   ;)
" Anyone who goes to a psychiatrist should have his head examined. "

Mark.



    Adviser
  • Forum Regular
  • Thanked: 67
    • Yes
  • Certifications: List
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 10
Re: Cap replacement advice
« Reply #5 on: October 07, 2017, 06:52:57 PM »
+1 with Salmon, you should bet on it not just being the mobo caps.
+1 with Dave, don't try to desolder the old caps from the underside of the board, yank them out as described and solder the new cap to the old legs. (please double check polarity)

but above and beyond all that, how old is this P3 board?
and what's the likelihood that even if you do get it working, that another failed component won't be on the near horizon?
any chance you can get another 'new' board?

and consider the cost in time, money and resources compared to taking the oppurtunity to finally upgrade the mobo (hence other components) to a more modern architecture.  (budget allowing of course)

Geek-9pm


    Mastermind
  • Geek After Dark
  • Thanked: 1026
    • Gekk9pm bnlog
  • Certifications: List
  • Computer: Specs
  • Experience: Expert
  • OS: Windows 10
Re: Cap replacement advice
« Reply #6 on: October 07, 2017, 07:34:30 PM »
Quote
Got an Old P3 board
All old P3 boards  should be replaced.
The 'replace the capacitors' myth started when a maker used inferior material and gave the capacitor industry a black eye.

BC_Programmer


    Mastermind
  • Typing is no substitute for thinking.
  • Thanked: 1140
    • Yes
    • Yes
    • BC-Programming.com
  • Certifications: List
  • Computer: Specs
  • Experience: Beginner
  • OS: Windows 11
Re: Cap replacement advice
« Reply #7 on: October 07, 2017, 07:56:08 PM »
A lot of devices have known issues with capacitors which tends to result in the "replace the caps" logic being pretty much a go-to suggestion. It sort of makes sense in the sense for older devices, since at that age they can very much be responsible and most people don't have a way to test the ESR of a cap outright- and if they do, well, you have to remove the capacitor anyway- may as well put another one in anyway- Even if the ESR on a 20 year old capacitor is fine, if you've already got it removed- may as well put a new one in.

For example older Macintosh Systems (128K, 512K, SE, etc.) are at the age where the capacitors used are unlikely to be in good shape and if they are they are a few decades old. In those cases, if the system board survived the battery it's a good idea to just replace the capacitors as a matter of course- You have to crack it open to remove/replace the batteries in those systems anyway.

Of course issues could be from any number of things failing, but "replace the caps" tends to be the most accessible route to try for most amateur repair attempts. Actually inspecting the components on a board properly requires things like an oscilliscope which most people not already into electrical engineering have, and which are rather more expensive than a soldering iron. Mind, It's hardly "Babies first soldering project" either, even properly removing through-hole caps can be a pain to do properly without nice things like a desoldering gun, and especially with cheaper irons if one of the legs are connected to a ground plane.
I was trying to dereference Null Pointers before it was cool.

Geek-9pm


    Mastermind
  • Geek After Dark
  • Thanked: 1026
    • Gekk9pm bnlog
  • Certifications: List
  • Computer: Specs
  • Experience: Expert
  • OS: Windows 10
Re: Cap replacement advice
« Reply #8 on: October 07, 2017, 09:31:26 PM »
BC, can you document that?
First of all, any PC over ten years old will have issues. It is just too easy to blame the caps. A PC motherboard has a number of devices that can go bad with age. Most often it dues to abuse, not normal use.

Amount other,s the BIOS firmware goes bad with age. A bit her ant there that gets flipped  might not be notices. Or it can impair the whole system. That has been documented, but seldom publicly announced.

Anther issue is water damage. Somebody drops water on the PC hoard and does not clean it off for a day. Some chips are hydrophobic,  they have a minor flaw that makes then subject to rapid  internal corrosion if exposed to water. This is hard to document. Easier to blame the caps.

Devices to test caps 'in place' are part of the radio-TV analog stuff. Trying to find a bad cap in a well populated PC board is near impossible. Unless you  cur one leg of the cap out of the circuit. Too much work to be worthwhile.

If a vendor gives you a five-year warranty on a motherboard, he will use components that he knows will last. Gigabyte does.

Nut if the OP wants to replace the caps, let him. But thee is not easy way to know if the cap is good or bad unless there is a visible abnormality.

Now then, having said that, here are You Tube videos about this topic.
He talks about the ESR meter.
https://www.youtube.com/watch?v=qDABYKoVO4Q

BC_Programmer


    Mastermind
  • Typing is no substitute for thinking.
  • Thanked: 1140
    • Yes
    • Yes
    • BC-Programming.com
  • Certifications: List
  • Computer: Specs
  • Experience: Beginner
  • OS: Windows 11
Re: Cap replacement advice
« Reply #9 on: October 07, 2017, 10:37:46 PM »
BC, can you document that?
The Capacitor problems in the aforementioned Macintosh computers are well documented. Anybody who spends two seconds with a web search can find loads of information on the problem(s). They exhibit themselves as issues with the screen, issues booting, issues  with the sound, etc. That people can replace the capacitors and eliminate these problems seems to conclusively tell us the cause was the capacitors. Obvously if a PC was in a flood or got wet or dropped or something- then replacing the caps isn't going to fix the problem, and assuming that is in any way the sort of scenario to which I referred is dishonest at best- I'm not talking about every single problem, I'm referring to seemingly inexplicable issues or odd behaviour on older systems and motherboards. Of course they be caused by other things- for example the battery leaking onto the motherboard, which I specifically mentioned, and eating away traces and causing it to fail to boot, but lacking any other obvious damage, for a layman replacing the capacitors is as good a place as any to start if they want to try to fix it.  In the OPs case I'd be more likely to check out the power supply and the power switch (and the reset switch) as a first step.

Quote
any PC over ten years old will have issues.
Come on now, you tell me to document something that takes a few moments to determine by consensus of those interested in that particular area, but then go and make a completely unsubstantiated, generalized, and easily falsified claim? I have an iMac G3 and an PowerMac G4 as well as a PowerMac G5 which are all over ten years old. They still work with no issues- so your statement is demonstrably false. Of course if the G3 for example couldn't keep track of the clock, I'm not going to go replacing the capacitors, I'd replace the battery (Which I did preventatively anyway). If it fails to boot and thinks there is no HDD, I'm not going to replace the capacitors then either. I was not suggesting that Capacitor replacement was *always* the best option. It's the best option when somebody really wants a system working again (for some reason) and there is no obvious cause, like a battery leaking all over the mainboard, or obvious corrosion from water exposure. Nobody is going to take a PC that melted in a house fire, was hit with a sledgehammer, or was on a sunken ship for 20 years, replace the caps, and have it working again.

Later systems with capacitors with the bad formula (Capacitor plague) also have well-known issues. These can be PC motherboards, add-on cards (though add-on cards tended not to use electrolytics) and of course things like game consoles (A Sega Game Gear with no sound (where the volume is turned up, obviously) is going to 99.9% of the time be capacitor problems, it's such a well-known issue and I've never heard of the replacement failing to resolve that issue. In general, replacing electrolytic Capacitors won't cover 100% of the problems with a typical PC motherboard, but they are the most accessible replacable component for somebody who isn't an electronics engineer.

For example- bulging capacitors near the CPU, and the system freezes, even at the BIOS, after 5 minutes. Do you think somebody should be desoldering SMD firmware EEPROMs and then comparing them to known-good copies online to verify they are correct? Should they be testing all the transistors on the board, one by one, trying to check which one failed? Or maybe they should spend 30 minutes to an hour replacing that set of caps near the CPU and see if that fixes the issue, and then move on from there if doesn't? it's not like replacing the capacitors is going to *cause* problems when done properly, and if somebody cannot replace capacitors properly I don't see what hope they would have for replacing a Surface Mount chip or other component even if they could identify it was the issue.

With older systems failures that don't exhibit any obvious error codes tend to be Capacitors. For vintage 5150,5160, and 5170 AT systems, This shows that most failures were caused by a bad capacitor- EPROMs can go bad as this user discovered therein, but most issues that are inconsistent or don't have an entirely clear problem component end up being capacitors. On the list there are 2 or 3 that were found to be the ROM-related, and in at least one of those cases, the problem was pretty well indicated by the quartz window of the EPROM chip being completely exposed.

Quote
Devices to test caps 'in place' are part of the radio-TV analog stuff. Trying to find a bad cap in a well populated PC board is near impossible. Unless you  cut one leg of the cap out of the circuit. Too much work to be worthwhile.
Which is why you replace them all. It's a lot easier to desolder and replace a capacitor than desolder it to properly test the electrical series resistance. Then  you have a set of known good electrolytics installed and if issues persist you can work forward (or not, depending on your skill level) knowing it is not caused by the capacitors. If somebody is willing to replace the caps on a system board from 2000 then I think the 5 bucks it costs to try it a problem.

On the topic of audio equipment which Salmon mentioned, I have a Denon Tape Deck which has electrolytic capacitors. It makes a clicking noise during playback. I won't be replacing the caps to try and fix it, because I was able to determine it was a gear missing a tooth. If it had weird volume issues, strange problems with the level meter, that sort of thing, I might consider the capacitors the problem but I wouldn't replace them because I'm the sort who tends to "butcher the board" with that sort of attempt anyway- (And because of the problem space of audio equipment, capacitors often are not responsible.

Quote
If a vendor gives you a five-year warranty on a motherboard, he will use components that he knows will last. Gigabyte does.
We aren't talking about systems made in the last 5 years, though. Pentium III boards are stretching back to 1999-2001 or thereabouts. Those 5-year warranties aren't particularly useful if they were provided at the time of purchase. Modern motherboards don't use many electrolytics and tend to use Surface mount tantalums which are more robust anyway, and the boards are now tightly packed enough that any sort of amateur board-level repair is usually off the table altogether.
I was trying to dereference Null Pointers before it was cool.

comda

    Topic Starter


    Adviser
  • Thanked: 6
    • Yes
  • Experience: Experienced
  • OS: Windows XP
Re: Cap replacement advice
« Reply #10 on: October 13, 2017, 06:02:16 PM »
Alright. My apologies for the delayed reply. I will attempt my best at this.

 
How do you know it's only the "caps" that need replacing?

The board refused to post. All it did was flash all 3 lights on the keyboard as if it posted but it didn't. I tried different ram, GPU, etc. I had set the jumpers to what I believed was correct according to the manual I got with it. Yet every 5 seconds, those lights lit up. All 4 capacitors by the CPU where bulged up and leaking, hence the replacement. So I did so.

 
3300uF caps should work, but they will be a little slower at reaching 100% charge capacity and it will be a slightly heavier amperage draw to onboard voltage regulators ( aka VRM's ) and power supply until they level out at the voltage constant. If there is a ripple problem or a demand on this cap, it will hold up stronger than a 2700uF cap in that it carries 600uF more charge capacity as the easiest way to explain it to non Electronics Technicians.

I personally would go with exact same value components vs going with larger storage capacitors. * If you have a strong power supply the 600uF difference shouldnt be a problem, but if its a weak power supply it could cause an excessive current draw at boot that could lead to problems if the voltage rail isnt able to hold steady at a strong voltage with the extra current draw at boot. To charge the additional 600uF per capacitor the unknown is if your VRM's or power supply will be able to carry the heavier initial current draw at boot. These caps should get to full capacity within about a second of boot and voltage should hold strong enough to not affect the motherboard boot. But there is a small chance that you stress a VRM or the power supply and cause more damage.

I have performed the capacitor replacements before and easiest method is to tug at the can and remove it off the board leaving the legs in place. Then solder a new capacitor to the legs of the old. You might need to add additional flux to get it to flow to the legs of the old cap as for the electrolyte oil can make for a nasty soldering. The flux helps get the tin to flow. I have even made a capacitor daughter board with wires between motherboard and daughter board where replacement capacitors would conflict with a socket 478 heatsink for example. I poured hot glue on the bottom side of the daughter board PCB to seat and insulate all the capacitors so that they would never short to ground and I attached the daughter board to the inside of the case in a location that wasnt conflicting with limited space in the SFF ( small Form factor ) case. Also I never changed values for uF ( aka Microfarads ). Whatever value came out went back in.

Additionally Capacitors have Polarity and if you get the polarity backwards they EXPLODE!

I never bothered saving a capacitor plague Pentium III board but have saved a few Pentium 4's back when the capacitor plague was really bad and systems were dropping like stones and HP and Dell didnt have a recall to cover the poisoned electrolyte recipe capacitor problem so people were stuck buying new computers or having someone skilled replace the capacitors. Or migrate the CPU and RAM to a new motherboard that hopefully doesnt have the bad capacitors in it.

Did any of the capactors on that board short and explode blowing off the capactor body or did they just swell and vent out the electrolyte?

I have gotten boards that are bricked by the capacitor plague to boot by removing all the damaged capacitors around the CPU which is where I am guessing your capacitors are located. If it boots with them removed then its worthy of adding new capacitors. If it still acts the same then its a bigger problem then the capacitors if symptoms dont change. Running the system without these capacitors will cause the system to run in a unfiltered power state powered directly unfiltered from the VRM's and power supply and while it may boot now, do not run it for any length of time as for it might boot and look healthy but the minute you go to run something in which the CPU demands a heavier current draw and the capacitors aren't there to assist the system is likely to crash with an error or freeze up on you.

Im not at home, so I don't recall the board model/number. Its a high end board for its time that has a 867Mhz P3, AGP, PCI and even ISA on the board. Im trying to save it for a retro gaming PC. I did end up using the caps I asked about and now the board flashes its lights on the keyboard once, as if its posting, but not displaying anything on the screen. I will be trying different multiplier on the CPU, and a different PSU just in case. However im doing this project very slowly as my days are full.

Thank you to everyone who commented, I will continue to update, thank you for all the help.

Geek-9pm


    Mastermind
  • Geek After Dark
  • Thanked: 1026
    • Gekk9pm bnlog
  • Certifications: List
  • Computer: Specs
  • Experience: Expert
  • OS: Windows 10
Re: Cap replacement advice
« Reply #11 on: October 13, 2017, 07:40:08 PM »
For reference:
Some types of capacitors are will-suited for motherboard design. High capacity should translate to better stability, But more is involved.
Here is a general coverage of capacitors classified as 'electrolytic' types.
https://en.wikipedia.org
https://en.wikipedia.org/wiki/Electrolytic_capacitor
Here is where it talks about the ESR.
/wiki/Electrolytic_capacitor#ESR_and_dissipation_factor_tan_.CE.B4

Because of costs, reliability  and availability issues, the bust choice of capacitors for motherboards has shifted.  Check the references at the end of  link.
Also: https://en.wikipedia.org/wiki/Tantalum_capacitor

Hope the OP can fix his motherboard.   :)
But if it were me, I would give it up.  :-\

Quote
requirements for capacitors increased, especially the demand for lower losses. The equivalent series resistance (ESR) for bypass and decoupling capacitors of standard electrolytic capacitors needed to be decreased.[28]

Although solid tantalum capacitors offered lower ESR and leakage current values than the aluminum electrolytics, in 1980 a price shock for tantalum in the industry dramatically reduced the usability of tantalum capacitors, especially in the entertainment industry.[29] [30] In search of cheaper alternatives, the industry switched back to using aluminum electrolytic capacitors.

The development of conducting polymers by Alan J. Heeger, Alan MacDiarmid and Hideki Shirakawa in 1975 was a break-through in point of lower ESR.[31] The conductivity of conductive polymers such as polypyrrole (PPy) [32] or PEDOT [33] are better by a factor of 1000 than that of manganese dioxide, and are close to the conductivity of metals. In 1993 NEC introduced their SMD polymer tantalum electrolytic capacitors, called "NeoCap". In 1997 Sanyo followed with their "POSCAP" polymer tantalum chips.

DaveLembke



    Sage
  • Thanked: 662
  • Certifications: List
  • Computer: Specs
  • Experience: Expert
  • OS: Windows 10
Re: Cap replacement advice
« Reply #12 on: October 16, 2017, 01:06:40 PM »
Quote
I will be trying different multiplier on the CPU

Are you able to get into the BIOS to adjust the multiplier setting?

Im assuming that this board once worked with that exact CPU in it and then it broke?

comda

    Topic Starter


    Adviser
  • Thanked: 6
    • Yes
  • Experience: Experienced
  • OS: Windows XP
Re: Cap replacement advice
« Reply #13 on: October 23, 2017, 09:42:45 AM »
For reference:
Some types of capacitors are will-suited for motherboard design. High capacity should translate to better stability, But more is involved.
Here is a general coverage of capacitors classified as 'electrolytic' types.
https://en.wikipedia.org
https://en.wikipedia.org/wiki/Electrolytic_capacitor
Here is where it talks about the ESR.
/wiki/Electrolytic_capacitor#ESR_and_dissipation_factor_tan_.CE.B4

Because of costs, reliability  and availability issues, the bust choice of capacitors for motherboards has shifted.  Check the references at the end of  link.
Also: https://en.wikipedia.org/wiki/Tantalum_capacitor

Hope the OP can fix his motherboard.   :)
But if it were me, I would give it up.  :-\

Thanks LOL> Sorry for the slow replies. Life is crazy, so I appreciate you guy helping.

The board is rather a nice one for its era. I forget the exact model number but its a shuttle board, that has a 867Mhz P3 on it, AGP, PCI and yes even ISA. Perfect for my classic gaming machine. Im trying to fix it, as its a nice looking board.

Are you able to get into the BIOS to adjust the multiplier setting?

Im assuming that this board once worked with that exact CPU in it and then it broke?

im not able to get into the BIOS. Before the cap replacement, id get a black screen, and the keyboard lights would flash once every 4 seconds. After the cap replacement they flash once, the minute I turn it on as if its posting, but nothing on screen. The multiplier settings are jumpers on board, and I got the manual with it so I set them (to what I believe is correct) but ill have to double check when im. back at home.

Looking again. The board is a SpaceWalker Mainboard AV11. Last revision if I recall.