Own Custom Build Computer With ECS Motherboard-No POST

[Agent24]

Re: Own Custom Build Computer With ECS Motherboard-No POST

Which voltages did you check? And what were they?
You need to check CPU V_core, CPU Vtt, RAM Vdd (should be 1.8V), Ram Vtt (usually about 1/2 of RAM Vdd), Northbridge Vcc, and Southbridge Vcc.
Let me know if you need help locating those.

Agree. With my recent fun with motherboard troubleshooting I would say start with voltages and power good signals.

Is the PSU sending the power good signal, for a start? If not, it will make the board stay in reset state and do nothing.

Get a schematic of the board if you can. If you can't, get a pinout of the SuperIO chip and find the control signals...

[momaka]

Re: Own Custom Build Computer With ECS Motherboard-No POST

- check the standard voltages on the motherboard with a multimeter

Which voltages did you check? And what were they?
You need to check CPU V_core, CPU Vtt, RAM Vdd (should be 1.8V), Ram Vtt (usually about 1/2 of RAM Vdd), Northbridge Vcc, and Southbridge Vcc.
Let me know if you need help locating those.

- keyboard (num, caps, and scroll lock) all light up initially and then go out (after that, it seems that the keyboard does not work)

Does the Num Lock light work if you repeatedly push the Num Lock key? And does it lock into an either OFF or ON state after a while? If yes, that usually indicates a hard crash (i.e. BIOS didn't finish loading or never even started).

Any suggestions for this motherboard recap as I have 10 NICHICON HM(M) 1000uF 6.3V ? sitting in this motherboard while the rests are Teapo capacitors ?

You can try new Nichicon HM, HN, or HZ caps (available on badcaps.net). Other choices would be Rubycon MBZ and MCZ. 820uF 2.5V polymers would work on spots that have less than 2V on them. Probably even Rubycon ZL, ZLG, and ZLJ would work, even though they don't have as low of an ESR as HM.

As for the Teapo - most appear to be on the 5VSB or USB V lines from what I can tell, or other non-important rails. The only troublesome Teapos are the one in the lower right corner of the board in your picture (by the RAM slots), and the one by the "SYS_FAN" connector, which may be for the high side of the CPU Vtt MOSFET to the left of it.

When desoldered from the motherboard the suspected Nichicon capacitor did not fully charge up when I charged it with only normal 9V battery.

That's because you are using a 9V battery on a 6.3V cap. You must NOT go over the rated voltage of the cap at any time. Try it with a 5V adapter. And the more important part is to see if the cap can make a tiny spark if you short it against a metal surface - a crude indication that the cap is more or less healthy.

Now other things I can think of:
What kind of CPU heatsink did you use? Does the board "flex" under the CPU socket when the heatsink is on? Try putting the CPU heatsink on the CPU without locking it on the board (and with thermal compound, of course).

Also, I can't help but notice that your board has only 4 caps on the CPU V_core - that's very skimpy on ECS's part. I doubt this is the problem, since you likely would have gotten multiple crashes before the board died. But if you want to try it for an exercise, get two more 2.5V 820uF polymer caps and add them to the empty spots next to the CPU. Nichicon LF, R5, R7, E5, and E7 series should be okay (among others). Other choices are United Chemicon PSC and PSA, and Fujitsu (now also owned by Nichicon) FPCAP RE series. I think the Fujitsu's are available on badcaps.net store too.
Just avoid eBay at all costs. Too many counterfeit caps there.

[Wester547]

Re: Own Custom Build Computer With ECS Motherboard-No POST

I don't know exactly what you mean by "suspected to become shorted and overfilled because the voltage converted from 5V > 1.5V"

The article he linked holds claim that the bad HMs and HNs were overfilled. But I think there is a common misconception afloat on the internet at large that the announcement came from Nichicon. It didn't (same goes for the "SAR disinfectant announcement" - well, that did stem from Nichicon, but there's no mention of HM or HN there, and Rubycon gave rise to a similar announcement around the same time). I think it came from someone at Dell who thought they observed "overfilling" because the capacitors were "overflowing" with electrolyte (and - interestingly enough - Dell observed this in January of 2004 in accordance to some sealed documents from the A.I.T. lawsuit, and knowing this would affect at least 11.8 million computers, chose not to say anything about it...). Very observant, Dell... and that article makes of note that Nichicon's US representatives refused to comment and that Nichicon was not answering emails about the matter. Dell does claim that they audited Nichicon's plant but I don't believe them. They were caught using deceptive language (see the link below - not bothering to make the distinction of which series Nichicon use and by way of email, alerting employees not to speak in a fully truthful manner).

Another misconception is that Nichicon pulled the bad capacitors in 2004. It was Dell who suspended the use of capacitors (and maybe Intel and HP - can't find the link at the moment but I remember reading a page where Intel claimed that they were inspecting and changing certain components on their motherboard like capacitors on 875 series boards), not Nichicon. That is why, at least judging from this thread, it's safe to say that even 2006 and 2007 HMs aren't safe (as well as this post). Threads like these makes me wonder if Foxconn was hit with a huge rash of fakes (fakes wouldn't surprise me at all - look at this thread). It would make sense as I noticed the Nichicon HNs and HMs of the time have poorer font than the modern ones, IE the stenciling on the bad HNs is quite awful (look at the second image on this page). That might not be it either, though, as HNs don't seem to hesitate to blow as of 2005 at least even with good stenciling.

Not trying to hijack the thread or scare anyone. I'm sure HMs and HNs, at least those sourced from Nichicon or an authorized distributor, are at least better than KZGs and KZJs. I may be entirely wrong and Nichicon might have taken care of business (I noticed modern HNs and HZs are rated from -25C to +105C according to the datasheet, whereas the original datasheets only rated them from -40C to +105C, suggesting that maybe the electrolyte was reformulated).

The HMs are 1.5V capacitors now that they have been on the VRM output for so long. As for the fact that they won't charge with a 9V battery, I believe momaka's method is to do that, remove the battery, then short the leads against some sort of metal. No sparks means they've gone leaky. I also highly recommend you replace the small Teapos as they have great notoriety for failing without any signs and may even be your issue.

[LENOVO-A880]

Re: Own Custom Build Computer With ECS Motherboard-No POST

I don't know exactly what you mean by "suspected to become shorted and overfilled because the voltage converted from 5V > 1.5V"

I'm actually using a digital multimeter to check the voltages for the capacitors around the memory slots which also connected to one of the MOSFET chips or voltage regulators and it has a 1.5V on one of the capacitors which is also a Nichicon capacitors. What I did to check the voltage of the capacitors was I removed every single piece of the hardwares and powered the motherboard up with only the PSU and battery. It seems it would never make sense that when I tested across for almost all the Nichicon capacitors would be reading 1.5V for each of the capacitors I tested by just poking my meter on the leads. I really don't know what I'm checking is correct or wrong so I just assume it should be a capacitor problem since the voltage is not up to tolerance that it should have which is 5V. When desoldered from the motherboard the suspected Nichicon capacitor did not fully charge up when I charged it with only normal 9V battery. One of the reason I suspected this Nichicon capacitor is because some of this capacitor don't even show any sign of failure or pop up when some of the manufacturer claimed it will still can cause trouble if we don't remove this plagued capacitor.

(http://news.cnet.com/PCs-plagued-by-...3-5942647.html).

Anyway does this show that the capacitor that is unable to be fully charged will has some leakages ? Please tell me what else should I check to actually narrow down the situation for this motherboard or anything that can prevent it from booting up properly? And I must say to you sorry that I don't have the exact picture showing the test I've done to show it on my post.

[c_hegge]

Re: Own Custom Build Computer With ECS Motherboard-No POST

I think your capacitors are fine. A Nichicon HM that new is unlikely to have a problem. I don't know exactly what you mean by "suspected to become shorted and overfilled because the voltage converted from 5V > 1.5V" Do you mean that a voltage regulator which should have been giving 1.5V failed and outputted 5V? If so, the cap will survive that, being rated for 6.3V maximum, but whatever it was powering would not have.

[LENOVO-A880]

Background

I have my first custom build of computer running fine for almost 3 years before it finally unable to boot up properly recently. I have my computer installed with Intel Celeron 430 (1.80GHZ), 75W processor with fan; 1GB Kingston Module DDR2 667MHZ ram; ATI RADEON graphic card and 450W/500W Power Logic PSU. All the fans are working (graphic card, cooler and PSU) when I turn the computer casing power button ON. The power LED won't turn ON and seems to have been shorted. I have plugged in the motherboard buzzer or speaker, graphic card, LCD monitor and VGA cable into place but whenever I turn it on I have no signal from my motherboard and no POST beeps as well....only DARK SCREEN!

Motherboard

This is an ECS G31T-M7 REV:7.0 motherboard made by Elitegroup. It will not POST. Specifications link:

http://www.ecsusa.com/ECSWebsite/Pro...uID=16&LanID=9

Service Manual at:



Symptoms

- HSF fan never ending spinning
- No LED light on case F/PANEL
- No beeps
- No displays
- No POST

Tests Done So Far

Here is what I have tested so far and found. All resulted in no beeps and no POST:

- all tests done below with the motherboard "breadboard" (that is out of the case sitting on a cardboard box)
- reseated the existing 3V battery (measures 3.16V DC on multimeter)
- cleared CMOS battery by removing it for a day and tried moving the clear CMOS jumper
- all voltages on PSU are within tolerances
- check the standard voltages on the motherboard with a multimeter
- check the capacitors voltages are within tolerances with a multimeter
- desoldered the NICHICON HM(M) 1000uF 6.3V capacitor suspected to become shorted and overfilled because the voltage converted from 5V > 1.5V
- keyboard (num, caps, and scroll lock) all light up initially and then go out (after that, it seems that the keyboard does not work)
- tried one DIMM in slot 1 and then slot 2 and get no beep
- tried removing all DDR memory and still get no beeps
- turn the board on and off by shorting the power SW-PINS (southbridge is partially working)

Questions

What can I do or anything to try to make this board work again ? And what should be the best repalcement capacitors for NICHICON HM(M) 1000uF 6.3V ?
Any suggestions for this motherboard recap as I have 10 NICHICON HM(M) 1000uF 6.3V ? sitting in this motherboard while the rests are Teapo capacitors ?
PLEASE HELP!!!