Author Topic: Capacitor failed or not?  (Read 7760 times)

0 Members and 1 Guest are viewing this topic.

Offline madderhatterTopic starter

  • Newbie
  • Posts: 3
  • Country: us
Capacitor failed or not?
« on: December 02, 2015, 05:01:26 am »
I have a solid state capacitor I pulled off a motherboard that will not post or show any signs of booting which is measuring 20% higher than its rating on the side with my esr meter.  All my other caps are showing the proper farads or approximately within 10% usually.  Is this a failed cap?  I never saw a cap have more farads than the rating in any of the vids I've seen.

It has a dent in the top of it btw, which is why I got suspicious of it and pulled it. 

Also anyone want to give me any tips on diagnosing a motherboard with no power would be helpful as well.
 

Offline wblock

  • Regular Contributor
  • *
  • Posts: 239
  • Country: us
Re: Capacitor failed or not?
« Reply #1 on: December 02, 2015, 05:48:25 am »
ESR specs are usually only available in the data sheet, not printed on the capacitor itself.  Capacitance typically has a pretty loose tolerance, and can be higher than the nominal value.  So: what is the measured ESR?  What is the ESR listed in the spec sheet for that model of capacitor?  Note that physical size of the cap affects ESR.

Photos would help.  A "dent" in the top might just be an undamaged vent.
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 457
  • Country: us
Re: Capacitor failed or not?
« Reply #2 on: December 02, 2015, 06:38:04 am »
MB without power  has not much components to test.
1 is it makes beeps sequences when you connect power? It could be RAM, CPU even HDD not proerly connected
If no CPU RAM installed it would not boot and it will signal you that either CPU or RAM is missing or bad.
 

Offline madderhatterTopic starter

  • Newbie
  • Posts: 3
  • Country: us
Re: Capacitor failed or not?
« Reply #3 on: December 02, 2015, 06:55:46 am »
Done all the routine testing.  I'm a pretty experienced system builder.   It's not a simple problem.   Yes tested with multiple powersupplies, memory, ect.

 I'm looking for component level troubleshooting advice.   Lights indicate power to the board, but nothing happens at all when the power button is pressed, no fans, doesn't even try to post, it's acting like there's a short or possibly an open somewhere probably, which is why I thought it probably is caps especially since there's no physical indication of damage anywhere on the board. 

Tested the power button for continuity, now looking for other problems.   The traces from the power button are very hard to track unfortunately.   I guess I need to find the schematics hopefully.


 Even tried booting it without processor or memory completely.


Interesting about the capacitance values... ok guess I'll put it back in.   It's a solid state aluminum cap, so no vents, it just has a big round dent in the top of it.  about a quarter the diameter of the top surface diameter.

esr is .28 ohms,  it reads 1000 uf on a 6.3v 820uf cap.  sorry mixed the numbers up.
 

I guess it's one of these. http://www.ebay.com/itm/321146620216

it's off an x58 system, so no doubt at least 5+ years old now. Pre 2011 obviously.   I bought it as a project mobo off ebay.   I expect it was run that whole time, and probably oc'd as well since that's what these systems were built to do.


« Last Edit: December 02, 2015, 07:19:53 am by madderhatter »
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: Capacitor failed or not?
« Reply #4 on: December 02, 2015, 09:52:12 am »
Caps on motherboards are not nearly as stressed as caps in switching power supplies so a little history would help.
But modern motherboards have an SMPS to power the CPU and the capacitors do work quite hard.
esr is .28 ohms,  it reads 1000 uf on a 6.3v 820uf cap.  sorry mixed the numbers up.
That sounds reasonable to me.
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 4983
  • Country: ro
  • .
Re: Capacitor failed or not?
« Reply #5 on: December 02, 2015, 10:43:01 am »
Yeah, there's a DC-DC converter that takes 12v in and outputs 0.8-1.4v to the CPU.  It's similar to a SMPS but there's no transformers like you see in regular power supplies. There's still mosfets, inductors, capacitors, high ripple currents etc.

As for the capacitors... if you measured the capacitance and/or esr with the capacitor still soldered in the mainboard, you're doing it wrong. Especially in the cpu area where the dc-dc converters are, multiple identical capacitors are wired in parallel, to reduce the esr and gain other benefits as well. Point it, if your capacitor is in parallel with other capacitors, anything you measure would be incorrect.

28mOhm otherwise for an ELECTROLYTIC 820uF 6.3v is very good.  For a polymer, it's kinda bad... but I would expect to see 820uF 4v or even 2.5v rated polymers on the output side of the cpu dc-dc converter.

It would help if you could take a picture of your motherboard and point out which capacitor you suspect of causing problems.
 

Offline fleuroman

  • Contributor
  • Posts: 27
  • Country: au
Re: Capacitor failed or not?
« Reply #6 on: December 02, 2015, 10:48:24 am »
golden rule... check power supplies...
a) get schematics for the mobo
2) check all power rails on the board
iii) check clock signals...
 

Offline wblock

  • Regular Contributor
  • *
  • Posts: 239
  • Country: us
Re: Capacitor failed or not?
« Reply #7 on: December 02, 2015, 01:54:21 pm »
I'm looking for component level troubleshooting advice.   Lights indicate power to the board, but nothing happens at all when the power button is pressed, no fans, doesn't even try to post, it's acting like there's a short or possibly an open somewhere probably, which is why I thought it probably is caps especially since there's no physical indication of damage anywhere on the board.

An open cap usually just makes the system flaky, because it is no longer smoothing out power supply ripple.  It's a matter of degree, because there are usually a bunch of caps in parallel.  So when one fails, the ripple goes up and the system becomes a bit shakier.  How much shakier depends on how close to the edge it was before that cap failed.

Quote
Interesting about the capacitance values... ok guess I'll put it back in.   It's a solid state aluminum cap, so no vents, it just has a big round dent in the top of it.  about a quarter the diameter of the top surface diameter.

esr is .28 ohms,  it reads 1000 uf on a 6.3v 820uf cap.  sorry mixed the numbers up.
 

I guess it's one of these. http://www.ebay.com/itm/321146620216

If that one came with the motherboard, it is probably real, but stories of fake capacitors from ebay are common.  A quick search did not find any "Matsuki" spec sheets that showed ESR, but that might not mean anything.

The symptoms do not make it sound like a capacitor problem, though.
 

Offline senso

  • Frequent Contributor
  • **
  • Posts: 951
  • Country: pt
    • My AVR tutorials
Re: Capacitor failed or not?
« Reply #8 on: December 02, 2015, 03:09:51 pm »
You can buy packs of caps from badcaps to replace them, but those caps around the CPU that smooth the CPU power rail are a bit hard to remove, about 2 months ago I replaced over 40 caps in a Pentium 4 motherboard from a surveillance system, worked like a charm, for a week, the chipset died(always on for years in a row, full of dust, hard rock thermal paste, it was even dark under it because the chipset was passive cooled, warned then that it was lost time...), a "new" board from ebay costed something like 40€, works like a charm.
 

Offline madderhatterTopic starter

  • Newbie
  • Posts: 3
  • Country: us
Re: Capacitor failed or not?
« Reply #9 on: December 02, 2015, 04:17:02 pm »
golden rule... check power supplies...
a) get schematics for the mobo
2) check all power rails on the board
iii) check clock signals...

ok, how do you check the rails/clock signals then?  You mean the phases right? just with a multimeter or do you need a logic analyzer?


Btw yes as I said I tried multiple know working components already, cpu et all.   I have 5 x58 systems and tons of older ones that I've repaired in various states already, like broken pins, bad caps ect, but this power issue is a new learning thing for me, so trying to see what I can do. I'm fixing socket pins/reballing/replacing obviously fried smds and all sorts of board level stuff, but trying to learn board level diagnostics a lot better, and get a better idea of the architecture beyond just replacing obviously physically damaged stuff.

Yes, I've taken one of the cpu caps out to analyze it as well, they are tough, but doable.  It was also physically damaged, it looks like from someone trying to put too big of a cooler on it or something, it also showed a similar overcapacitance, so I guess they're probably ok.   Thanks for the help in that.


There are usually caps by the atx power supply connector that I've found bad caps there before that were completely bad in other boards that definitely stopped the system from booting.  I agree though probably caps near the socket, the system might still power because it would still be getting power to the chips/architecture that boots the system right?

From videos I've watched there seems to be a main controller chip that checks voltage and connections that when you press the power button it decides if it wants to try to run the power supply but that was an older board, so maybe that is all in the cpu now I guess?  I need to find schematics for this board.  It's an xfx 58i..any idea where I would find those?  A quick google search yielded nothing.   Maybe there are general x58 chipset schematics somewhere that would prove useful?


Found a really good vid... so this is for a 2011 board, but I guess the main chip I'm looking to measure is the pwm.



I wonder if I solder wire onto the caps from the last point before they go to the cpu and then measure from ground if I can monitor voltages going to the cpu like that one vid that Dave Jones had with like 8 multimeters going to monitor all the different phases. That would be cool. lol


The pwm on this board is an it8268r. so maybe I can figure it out if I look up the pinouts on it.

http://images.people.overclockers.ru/195575.gif

https://translate.google.com/translate?hl=en&sl=ru&u=http://www.rom.by/forum/Mat_GIGABYTE_GA-EP35-DS4_zapuskaetsja_no_net_pitanija_na&prev=search

It's so weird that all the board level repair stuff is either in Russian, hindi, or chinese.  Wth everyone in the western world just throws their boards out I guess.

Only one pin on the pwm seems to have power...sooo now where do I go? lol   Also which pins tell it to power on..hmm.


Anyone know of a video/explination like this in english and newer? lol




Turned it on using this trick.



Still no post though.   Nothing blew up or burned out after replacing all the caps, so I count myself lucky for that at least.   If something had blown up though I might have known what to replace. lol



oh the pwm is the isl6336, not the other one...researching that now.

ok well at least I can see some circuits now.

 http://www.datasheetdir.com/ISL6336+PWM-Controllers
« Last Edit: December 04, 2015, 05:53:47 pm by madderhatter »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf