Sudden pop, then burning smell...

H

hdnut

Gawd
Joined
Jun 10, 2008
Messages
829
I was running "blend" test via Prime95 when I heard this pop followed by a burning smell. Crazy thing was, the system kept on going without missing a beat. Prime95 kept on running, fully stressing all 4 cores, with the system showing no signs of any stability issues. I stopped the program and turned off the system anyway since I didn't want to risk damaging my system any further.

Some system details at the moment of "pop".
- ASUS Rampage III Formula motherboard
- Core i7 920 running at 4.4ghz (210x21) @ 1.35v
- RAM at 1683mhz (8-8-8-24 1T) @ 1.65v
- CPU PLL voltage @ 1.91v
- QPI/DRAM core @ 1.44v
- DRAM Bus @ 1.65v
- IOH (NB) voltage @ 1.26v

- CPU temps maxed out at 78c during "blend" test via Prime95.
- All other temps were good with NB and SB at 48c and 42c, respectively. I monitor all system temps religiously during stress testing and did not see anything out of the ordinary.

I immediately suspected the PSU, but had my doubts since it's pretty robust Thermaltake Toughpower 1200 watt'r. Took a flashlight and looked for possible busted capacitors inside the PSU but I couldn't find any and some were hidden from view. I then started look for possible busted/bulging/popped capacitors on the motherboard and the two GTX 460's. Nothing. My nose was still picking up a faint burning smell coming from the system but I could not find anything. No busted caps that I could see and no signs of any damage.

So, I powered up the system... and it posts. Everything looks good to go and the system boots all the way. I ran 3DMarks Vantage to see if there's anything wrong with the graphic cards. I get the highest Vantage GPU score that I ever scored with perfect graphic rendering on these moderately overclocked GTX 460's - 29,997 with physX off. WTF...

I was still scratching my head at this point. All of the attached peripherals and hard drives were working. Core i7 920 was running perfectly at 4.4ghz before the pop. I considered just leaving the system as it was since it was running fine, but my rational side took over and I decided to completely take the system apart and inspect the components one by one.

First the PSU. I considered opening it up to get better look at the caps, but that was going to void the warranty. I was suspecting the PSU as it seem like the most probable culprit but I could not pick up any burning smell. Gathered all of the necessary RMA information just in case.

With all the components and cables off the motherboard, I started inspecting every capacitor one by one with the brightest LED flashlight I have, my HTC Droid Incredible. The LED's on this phone is literally about 3 times brighter than my actual flashlight. It's freaking awesome. Anyway...

I inspect each cap carefully. Initially I see nothing. Then, a rounded top of a capacitor catches my eye. It's in the worst possible location ever - right below the heatpipe going from the NB to the mosfet heatsink. No wonder I kept on missing it...

Two capacitors, sitting side by side, was blown out. Upon closer inspection, the rubber seal was clearly visible at the bottom of one of the caps because the metal cover had popped off.

I'm still trying to figure out why the caps popped. My temps and voltages were well within limits, imo. I hope ASUS did not resort to less than reliable capacitors due to cost cutting measures.

So... the motherboard is on its way back to Newegg. I am hoping that the replacement Rampage III Formula does not have the case of exploding caps again.
 
damn it, would help but you forgot to post pics of the board....


and/or girlfriend...
 
jiminator said:
damn it, would help but you forgot to post pics of the board....


and/or girlfriend...
Click to expand...

I actually did take some shots of the blown capacitors but the stupid camera flash made everything way too bright at point blank range and made it impossible to tell what you were looking at.

Since I won't share my GF pics, here's a pic of my system naked and currently without a mobo as consolation prize. :D

 
Damn, the magic blue smoke got out did it ?

Hope you get things sorted out soon.
 
EVIL-SCOTSMAN said:
Damn, the magic blue smoke got out did it ?

Hope you get things sorted out soon.
Click to expand...

haha magic blue smoke.

Sucks OP that it happened on a high end board. You spend the so you extra money so you don't run into these kind of problems, but what can you do....
 
xxdragonxx said:
haha magic blue smoke.

Sucks OP that it happened on a high end board. You spend the so you extra money so you don't run into these kind of problems, but what can you do....
Click to expand...

Yeah, that's exactly what crossed my mind when it happened. Hopefully it was a fluke and it doesn't happen again with the new board.
 
If you want to save some coin on your return I would suggest going with the Sabertooth x58? The [H] reviews compare it to the RIIIE but for much less.
 
Either poor quality capacitors and/or too much ripple allowed by your PSU. Capacitors are used to smooth out ripple but too much ripple can kill them.
 
xxdragonxx said:
If you want to save some coin on your return I would suggest going with the Sabertooth x58? The [H] reviews compare it to the RIIIE but for much less.
Click to expand...

I got the mobo brand new for less than $250. Plus, I like the onboard sound that Rampage III Formula comes equipped with. No more add-in sound cards for this guy.

larrymoencurly said:
What's the maximum CPU power rating for that mobo?
Click to expand...

What are you asking exactly? Max CPU frequency limit or voltage?

dandragonrage said:
Either poor quality capacitors and/or too much ripple allowed by your PSU. Capacitors are used to smooth out ripple but too much ripple can kill them.
Click to expand...

That is interesting. The Thermaltake Toughpower 1200w received some of the highest accolades for 1200w PSU when it first appeared on the scene. So I'm reluctant to think that it's the PSU. But, since the PSU is almost 2 years old now, it may have degraded a bit and the ripple might be greater than what it used be. If the PSU is indeed faulty, I hope it doesn't kill off other components on the next go around when I get my new board. I will be pushing the i7 920 to 4.4ghz (21x210) at 1.33-1.35v again so we shall see.

[H]ard|OCP review of Thermaltake Toughpower 1200w, specifically ripple and regulation testing here:

http://www.hardocp.com/article/2007/04/25/thermaltake_toughpower_1200w/7
 
larrymoencurly said:
What's the maximum CPU power rating for that mobo?
Click to expand...
hdnut said:
What are you asking exactly? Max CPU frequency limit or voltage?
Click to expand...
I mean the maximum amount of power the motherboard's CPU voltage regulator (all those MOSFETs, capacitors, and coils around the CPU socket) is designed to put out continuously. Intel says the rating should be at least 130W to support the 920 i7 I don't know what a 920 i7 overclocked by by 63% consumes, but because it's a CMOS CPU it wouldn't surprise me if it needed 163% of 130W, or 212W. Does anybody know if it's hard for an 8-10 phase voltage regulator to put out that much? Because motherboards made for old Pentium 4s rated for 80-90W had just 3-phase regulators and regular electrolytic caps that couldn't stand as much ripple current or heat as the solid polymer caps used in the Rampage III Formula. Can reliability for overclocked systems be improved by switching to better polymer caps or by adding more polymers in parallel to the existing ones?
 
larrymoencurly said:
I mean the maximum amount of power the motherboard's CPU voltage regulator (all those MOSFETs, capacitors, and coils around the CPU socket) is designed to put out continuously. Intel says the rating should be at least 130W to support the 920 i7 I don't know what a 920 i7 overclocked by by 63% consumes, but because it's a CMOS CPU it wouldn't surprise me if it needed 163% of 130W, or 212W. Does anybody know if it's hard for an 8-10 phase voltage regulator to put out that much? Because motherboards made for old Pentium 4s rated for 80-90W had just 3-phase regulators and regular electrolytic caps that couldn't stand as much ripple current or heat as the solid polymer caps used in the Rampage III Formula. Can reliability for overclocked systems be improved by switching to better polymer caps or by adding more polymers in parallel to the existing ones?
Click to expand...

Considering that Rampage III Formula is an enthusiast level motherboard, I would be very surprised if ASUS did not factor in how much power overclocked i7 would consume. The reviews I've seen clearly shows that this board can handle 63% CPU overclock and then some. This is the reason why I still believe what has happened to me as an aberation.

I would also like to believe that they are already using "better" caps on these boards since it is marketed to enthusiasts. $300 mobo should laugh at a mere 63% overclock, imo.
 
hdnut said:
Considering that Rampage III Formula is an enthusiast level motherboard, I would be very surprised if ASUS did not factor in how much power overclocked i7 would consume. The reviews I've seen clearly shows that this board can handle 63% CPU overclock and then some. This is the reason why I still believe what has happened to me as an aberation.

I would also like to believe that they are already using "better" caps on these boards since it is marketed to enthusiasts. $300 mobo should laugh at a mere 63% overclock, imo.
Click to expand...
Considering that the Rampage III Formula is an enthusast level motherboard, I would be very surprised if Asus cared more about engineering than about fashion and hype because not every enthusiast is concerned about maximum performance for important reasons (i.e., games, scientific/engineering modeling, graphic arts), as opposed to benchmark bragging rights and how pretty the mobo looks under the lights shining inside the improperly made case (window = improperly made, of course :D). I'd think an industrial mobo built for continuous operation in a 50C room would be better than an enthusiast's mobo at reliably cranking out tons of power to the CPU.

Uh, can't you simply check the markings on the caps to verify their quality? I don't understand how belief is a factor here. :) I can't tell what the Rampage III Formula has, but a few years ago another expensive Asus mobo used a Taiwanese brand called Apaq.
 
larrymoencurly said:
Considering that the Rampage III Formula is an enthusast level motherboard, I would be very surprised if Asus cared more about engineering than about fashion and hype because not every enthusiast is concerned about maximum performance for important reasons (i.e., games, scientific/engineering modeling, graphic arts), as opposed to benchmark bragging rights and how pretty the mobo looks under the lights shining inside the improperly made case (window = improperly made, of course :D). I'd think an industrial mobo built for continuous operation in a 50C room would be better than an enthusiast's mobo at reliably cranking out tons of power to the CPU.

Uh, can't you simply check the markings on the caps to verify their quality? I don't understand how belief is a factor here. :) I can't tell what the Rampage III Formula has, but a few years ago another expensive Asus mobo used a Taiwanese brand called Apaq.
Click to expand...

Are you saying that the hardware giant that is Asus do not understand the overclocking community, also referred to as "enthusiasts"? And that the motherboard is purely for benchmarking e-peen and looking pretty inside your case? Wow, when was the last time you played a very demanding game where squeezing out the last bit of perfomance mattered? Have you seen the BIOS for any of the ROG motherboards? (I'm asking rhetorically but you can answer if you want to) If all the tweakable settings in BIOS don't scream, "Overclock Me and Rejoice!", then I have purchased the wrong motherboard. Also, a modern day motherboard is by no means an epitome of engineering marvel, but still pretty damn good example of solid hardware engineering.

There is a simple reason why it is called an "enthusiast", aka "overclocker", level motherboard. An enthusiast in our hobby, do not (and, imo, should not!) run anything stock, meaning that once all the components are in place, tuning will surely follow. Whether it is for e-peen or for actually squeezing out every bit of mhz for actual gaming performance is not a big factor. The important thing is that folks purchase top of the line enthusiast level components such as the Asus Republic of Gamers Rampage III Formula motherboard because they intend on pushing their hardware beyond its stock specs. Companies know this, review sites knows this, and end users expect this. If you're one those folks that buy top of the line enthusiast level components only to run everything stock for "stability" reasons then you're really not doing your research properly and you're wasting your money. These boards are meant to run at relatively high overclocked settings AND provide stability while doing so. Read any reviews on any of the ROG motherboards and see reviewers gush over the the overclocking performance of these boards.

I also don't have any problem with a hardware looking pretty either. This particular mobo is quite handsome, I must say, but it is one good performing and good looking motherboard. Most of us like our hardware to perform well and look good at the same time. This is normal and I see nothing wrong about it. It's just like how most of us like cars. I don't know about you, but I wouldn't spend $50,000 plus on a BMW M3 if it looked like a Chysler PT Cruiser. Performance matter and so does cosmetic appeal. Companies such as Asus knows this. Not everyone buys a computer simply because it does computations. Just look at all the Apple lovers... (I want to wash my mouth everytime I say the word Apple).

And, uh, I did look at the markings on the caps and found out that these are ESR long-life solid types. Didn't research much beyond that other than reports of people having capacitor issues on these mobos. Came up short on that so that led me to "believe" that what has happened to me as a fluke. Perhaps it is not. We shall found out when I get my replacement.
 
Ouch, I haven't had something like this happen since my NF7-S v2 got surged through my surge protector (go figure)
 
hdnut said:
And, uh, I did look at the markings on the caps and found out that these are ESR long-life solid types. Didn't research much beyond that other than reports of people having capacitor issues on these mobos. Came up short on that so that led me to "believe" that what has happened to me as a fluke. Perhaps it is not. We shall found out when I get my replacement.
Click to expand...
I'm saying this was a problem Asus engineers probably could have prevented easily and cheaply, and low ESR doesn't mean much because caps that differ by 10:1 in ESR are sometimes both called "low ESR". Then there's the ripple current rating of the caps. Maybe Asus decided to divert some money away from the engineering dept. to the styling dept., to make the product more appealing to the superficial. Only ugly people worry what their mobos look like. :D
 
or it may have just been 2 marginal caps placed in a location that they had a little extra heat buildup rather than the ones out in the open, weakest link theory ?
 
larrymoencurly said:
I'm saying this was a problem Asus engineers probably could have prevented easily and cheaply, and low ESR doesn't mean much because caps that differ by 10:1 in ESR are sometimes both called "low ESR". Then there's the ripple current rating of the caps. Maybe Asus decided to divert some money away from the engineering dept. to the styling dept., to make the product more appealing to the superficial. Only ugly people worry what their mobos look like. :D
Click to expand...

or maybe he just had a bad part, shit happens. Personally I would have RMAed the board but good luck with newegg :)
 
bwbob said:
or it may have just been 2 marginal caps placed in a location that they had a little extra heat buildup rather than the ones out in the open, weakest link theory ?
Click to expand...

I was thinking that the caps might have overheated as well.

Dudenell said:
or maybe he just had a bad part, shit happens. Personally I would have RMAed the board but good luck with newegg :)
Click to expand...

I got an email from Newegg that the new board has shipped today. I have heard horror stories about Asus RMA so I'm not too crazy with the idea of dealing directly with them. With Newegg, it's fairly easy to RMA as long as it is within 30 days of receipt.
 
bwbob said:
or it may have just been 2 marginal caps placed in a location that they had a little extra heat buildup rather than the ones out in the open, weakest link theory ?
Click to expand...
More likely, the final factory test checked only the digital part by running a diagnostic program, rather than poke the mobo with tons of needles that read signal readings everywhere, including ripple at several points of the CPU voltage regulator.
 
Those caps in question are probably part of the smoothing filter for the CPU/NB. They don't do much more than smooth out ripple, and thus with a good PSU the loss of a few won't make the system crash as everything will still be within tolerances. This would explain why the system kept running and booting even after these caps literally exploded. They weren't a critical part of the system :)

Which brand and type caps did Asus use in this location? I haven't heard of any bad caps being used recently and solid caps seem to be all the rage these days, which do not tend to die as violently as this.
 
Elledan said:
Those caps in question are probably part of the smoothing filter for the CPU/NB. They don't do much more than smooth out ripple, and thus with a good PSU the loss of a few won't make the system crash as everything will still be within tolerances. This would explain why the system kept running and booting even after these caps literally exploded. They weren't a critical part of the system :)

Which brand and type caps did Asus use in this location? I haven't heard of any bad caps being used recently and solid caps seem to be all the rage these days, which do not tend to die as violently as this.
Click to expand...

That explains quite a bit. I was thinking that these caps probably weren't part of anything critical when the system kept going.

The caps are solid low ESR types but I don't remember which company makes them. I'll check again when the new board comes in.
 
I got the new board few days ago and I've finally finished getting it all tested out. The new board is holding up very well and I've not noticed any popping or burning smells this time around. I've managed to successfully run and stress test the i7 920 at 4.4Ghz but decided to bring it down to 4.2Ghz for every day use. It was taking too much of a voltage bump for measly 200mhz increase.

I also noticed that at 4.4Ghz, it was disrupting my GTX 460 SLI from running smooth. This was especially evident when I ran Crysis Warhead and although I was getting decent FPS, the game stuttered and felt very choppy at certain points. Reducing the overclock to 4.2Ghz solved this.

I am happy to report that the new Rampage III Formula is good to go.
 
hdnut said:
I got the new board few days ago and I've finally finished getting it all tested out. The new board is holding up very well and I've not noticed any popping or burning smells this time around. I've managed to successfully run and stress test the i7 920 at 4.4Ghz but decided to bring it down to 4.2Ghz for every day use. It was taking too much of a voltage bump for measly 200mhz increase.

I also noticed that at 4.4Ghz, it was disrupting my GTX 460 SLI from running smooth. This was especially evident when I ran Crysis Warhead and although I was getting decent FPS, the game stuttered and felt very choppy at certain points. Reducing the overclock to 4.2Ghz solved this.

I am happy to report that the new Rampage III Formula is good to go.
Click to expand...

Man i am happy for you. I wish my G50VT-X1 would have been as good quality as the mobo seems...Damn thing BOILED the gpu 4 times, last time was 315f I almost was able to get a screenshot. But alas i burnt my wrist on the plastic trying to ctrl v and ctrl s
 
hdnut said:
Are you saying that the hardware giant that is Asus do not understand the overclocking community, also referred to as "enthusiasts"? And that the motherboard is purely for benchmarking e-peen and looking pretty inside your case? Wow, when was the last time you played a very demanding game where squeezing out the last bit of perfomance mattered? Have you seen the BIOS for any of the ROG motherboards? (I'm asking rhetorically but you can answer if you want to) If all the tweakable settings in BIOS don't scream, "Overclock Me and Rejoice!", then I have purchased the wrong motherboard. Also, a modern day motherboard is by no means an epitome of engineering marvel, but still pretty damn good example of solid hardware engineering.

There is a simple reason why it is called an "enthusiast", aka "overclocker", level motherboard. An enthusiast in our hobby, do not (and, imo, should not!) run anything stock, meaning that once all the components are in place, tuning will surely follow. Whether it is for e-peen or for actually squeezing out every bit of mhz for actual gaming performance is not a big factor. The important thing is that folks purchase top of the line enthusiast level components such as the Asus Republic of Gamers Rampage III Formula motherboard because they intend on pushing their hardware beyond its stock specs. Companies know this, review sites knows this, and end users expect this. If you're one those folks that buy top of the line enthusiast level components only to run everything stock for "stability" reasons then you're really not doing your research properly and you're wasting your money. These boards are meant to run at relatively high overclocked settings AND provide stability while doing so. Read any reviews on any of the ROG motherboards and see reviewers gush over the the overclocking performance of these boards.

I also don't have any problem with a hardware looking pretty either. This particular mobo is quite handsome, I must say, but it is one good performing and good looking motherboard. Most of us like our hardware to perform well and look good at the same time. This is normal and I see nothing wrong about it. It's just like how most of us like cars. I don't know about you, but I wouldn't spend $50,000 plus on a BMW M3 if it looked like a Chysler PT Cruiser. Performance matter and so does cosmetic appeal. Companies such as Asus knows this. Not everyone buys a computer simply because it does computations. Just look at all the Apple lovers... (I want to wash my mouth everytime I say the word Apple).

And, uh, I did look at the markings on the caps and found out that these are ESR long-life solid types. Didn't research much beyond that other than reports of people having capacitor issues on these mobos. Came up short on that so that led me to "believe" that what has happened to me as a fluke. Perhaps it is not. We shall found out when I get my replacement.
Click to expand...

LMAO did you check transistors/mosfets under the heatsinks around the CPU ? If you heard a pop then a smelled burning you most likely can track it down.

bigdaveberg said:
Man i am happy for you. I wish my G50VT-X1 would have been as good quality as the mobo seems...Damn thing BOILED the gpu 4 times, last time was 315f I almost was able to get a screenshot. But alas i burnt my wrist on the plastic trying to ctrl v and ctrl s
Click to expand...
I have a feeling that NOTHING this guys owns can possibly be "bad". Especially if its labeled Asus... :p
 
TT2eNT said:
LMAO did you check transistors/mosfets under the heatsinks around the CPU ? If you heard a pop then a smelled burning you most likely can track it down.
Click to expand...

Holy jebus, do people not read anymore? If you have read my initial post, you've known that I did find the location of the "popped" capacitor.

TT2eNT said:
I have a feeling that NOTHING this guys owns can possibly be "bad". Especially if its labeled Asus... :p
Click to expand...

Stop being a troll and move along.


Quality of some of our new members never fails to amaze me.
 
Im noob @ this forum.. Your noob @ OCing ....
hdnut said:
I also noticed that at 4.4Ghz, it was disrupting my GTX 460 SLI from running smooth. This was especially evident when I ran Crysis Warhead and although I was getting decent FPS, the game stuttered and felt very choppy at certain points. Reducing the overclock to 4.2Ghz solved this.
Click to expand...
o_O
 
TT2eNT said:
Im noob @ this forum.. Your noob @ OCing ....
o_O
Click to expand...

You are one hell of a TROLL. And how old are you, 12? Go troll somewhere else because [H] obviously is not for you. 4Chan got a nice spot reserved for trolls like you.
 
TT2eNT said:
LMAO did you check transistors/mosfets under the heatsinks around the CPU ? If you heard a pop then a smelled burning you most likely can track it down.


I have a feeling that NOTHING this guys owns can possibly be "bad". Especially if its labeled Asus... :p
Click to expand...

I never said that. It yerks me thought when the company wont do anything to help the customer out. I had sent the laptop in 2 times to asus. Both time i was without PC for 2 months. The second time they lost my laptop for 4 days. The gentlemen that signed for the package from UPS lost it. Then Bestbuy sent it out 2 times. Until the finally time when i brought it in and showed the guy again. It is frustrating being at college, and work with my main device gone countless times. And for them to say its not a lemon. Bullshit,
 
hdnut said:
I got the new board few days ago and I've finally finished getting it all tested out. The new board is holding up very well and I've not noticed any popping or burning smells this time around. I've managed to successfully run and stress test the i7 920 at 4.4Ghz but decided to bring it down to 4.2Ghz for every day use. It was taking too much of a voltage bump for measly 200mhz increase.

I also noticed that at 4.4Ghz, it was disrupting my GTX 460 SLI from running smooth. This was especially evident when I ran Crysis Warhead and although I was getting decent FPS, the game stuttered and felt very choppy at certain points. Reducing the overclock to 4.2Ghz solved this.

I am happy to report that the new Rampage III Formula is good to go.
Click to expand...

Gald to hear the new board is working out.

That being said.. since 4.4Ghz was being jerky while 4.2Ghz was not, it is pretty evident that the system was not quite stable.

Most likely the ECC error checking on the CPUs cache was catching a ton of errors and slowing the system to a crawl once the CPU was loaded up a certain amount.

It should have also shown up as slower test speeds in Prime95, IntelBurnTest, etc.

IntelBurnTest is especially easy to spot because the times for each loop of the test should be really close to each other. Once they start jumping all over the place, you know you have a problem.
 
cyclone3d said:
Gald to hear the new board is working out.

That being said.. since 4.4Ghz was being jerky while 4.2Ghz was not, it is pretty evident that the system was not quite stable.

Most likely the ECC error checking on the CPUs cache was catching a ton of errors and slowing the system to a crawl once the CPU was loaded up a certain amount.

It should have also shown up as slower test speeds in Prime95, IntelBurnTest, etc.

IntelBurnTest is especially easy to spot because the times for each loop of the test should be really close to each other. Once they start jumping all over the place, you know you have a problem.
Click to expand...

Now that you mention it, I did notice a difference in how fast Prime95 was running during its Blend test mode at 4.4ghz compared to 4.2ghz. All the cores were running without failing but it was noticeably slower.
 
You must log in or register to reply here.
Back
Top