The melting is caused by excesive heat, the heat is caused by two much curent passing through a weak point in the circuit, adding the mod allows the curent to pass by the conector, and effectavley acts like a larger pipe, annolagy, if you want to push watter through a hose, the amount of watter you can push is limited by the size of the hose, and the presure of the watter. incresing the presure increses heat, due to resistance. adding the mod allows more to pass, at lower presure, and because of the location of the connection, bypass the connector. and as stated above as long as you can return the board to like stock condition, who would know you ever did the mod. you would want to avoid physical damage, ( so I do not recomend the hot glue, note the sauder joints do not have glue on them now. ) and they function. your board has many exposed sauder joints, but they are seperated from grounding thanks to the stand off's.  If you still feel uncomftorable with exposed sauder joints, a good quality electrical tape would be easy to remove,

I have a question for you guys, especially those running E759 boards and BIOS 74.  Did you notice that the E-LEET readout for CPU VTT at load is always about 0.05V-0.07V higher than the value set in the BIOS?  I just want to make sure it is not just my board that is doing this.
 
I'm testing what I can while I have the Classified 3 here in case that I really need to RMA my E759 after this burn out ordeal so I want to check everything.

firerx

dejanh

I have a question for you guys, especially those running E759 boards and BIOS 74.  Did you notice that the E-LEET readout for CPU VTT at load is always about 0.05V-0.07V higher than the value set in the BIOS?  I just want to make sure it is not just my board that is doing this.

I'm testing what I can while I have the Classified 3 here in case that I really need to RMA my E759 after this burn out ordeal so I want to check everything.

hmmm, mine is actually lower in Eleet, Dejanh.

Not sure if I remember that but my ram voltage is like .5 higher than what i set it to, I had to set my ram at 1.6v so it would run at 1.65v.
 
As for the readings yes with that psu things where always lower than what they where set to in bios for the cpu etc, under load my voltages would drop. I asked about this on the forums and was told by other members with the Antec TPQ 1200w this was normal... But not too long afterwards as voltages dropped a bit more is when I started having issues.  Same with my other psu's this happened right as they where about to fail at least they just failed and stopped working.

dejanh

 

recon 7185

antec_jessie

I find it strikingly odd that an Antec rep would post in this thread as Antec has been a very strong proponent of the multi rail psu design.

My posting here has nothing to do with the design of our power supplies. My presence here is to simply provide assistance where possible and to gather information that might help us help you (insert Jerry Maguire here). 

I do have a question concerning the antec TPQ 1200. My northbridge on the x58 LE shorted out and the tpq 1200 kept going and did no shut itself off. Donta said that his other 2 PSU's died and the TPQ 1200 kept going and thats when I assume the pins melted. Is there not a specific safety feature on the TPQ that would protect hardware? Every other case where I have had a short or something out of the ordinary my PSU would shut off. Not bashing at all, in fact I love antec's stuff, just want to check before I get a new SLI approved PSU prefferably from antec

From what you are describing your board shorted out, not your PSU.  In that case PSU may or may not keep running depending on how bad the damage to your board is actually.  donta1979 case, my case, and others, had no issue with the board shorting out first.  It was the PSU that melted then took other stuff with it.  This aside, and not to be rude or anything, but questions unrelated to the problem at hand here should probably be put in its own thread.

My point in asking was to speculate whether or not that particular PSU shuts off under the correct conditions, such as a short in my case or melting the 24pin connector thus taking out hardware. I offered my experience as a way to try to see if the PSU in question does not protect other hardware under certain conditions like a melting 24pin. Anyways, sorry you thought i was trying to get off topic but i am trying to understand why the psu would not shut off, but then again i am not an expert

dejanh

PSU shuts off when it senses that there is too much current flowing through it.  For example a short circuit, or a power surge, will cause the Over Current Protection (OCP) to trip.  If a component on your board goes that does not necessarily mean that the PSU will shut down.  It is possible, but not generally the case.

Got it. All i know is that this problem is making me hold off on another TPQ 1200 and 480 until this is figured out. I was advised not to do the molex mod, and since I just got the SLI3 I am not really in the market for another mobo unless i could step up to the classified 3 and that is not an option

donta1979

I would get the Silverstone Zeus 1200w personally not sure about the corsair 1200ax if voltages drop or the pc power and cooling 1200w.. better for cable management, voltages dont drop like they do on the antec tpq 1200w like your 3.3v 3v, 12v 5v and so on.

when you say they drop do you mean under load or over time?

donta1979

recon 7185

donta1979

I would get the Silverstone Zeus 1200w personally not sure about the corsair 1200ax if voltages drop or the pc power and cooling 1200w.. better for cable management, voltages dont drop like they do on the antec tpq 1200w like your 3.3v 3v, 12v 5v and so on.

when you say they drop do you mean under load or over time?

under load I was worried about it and asked on the forums and other members said it was fine and normal with the antec tpq 1200, but tbh voltages in my opinion should not drop like that.

Thanks for the advice. What about the silverstone strider 1500w? I ask because it is in stock at my local store

I am running on a 875w Thermaltake right now. Can't complain. I guess I will have to read some reviews... Can I PM you if I have more questions to keep the thread on topic ? I was thinking and maybe this was mentioned before but would an adapter that would include a molex going between the male and female parts of the 24pin work? Or does the problem persist with overloading those points on the 24 pin?

i read this thread lastnite from page one was realy good reading minus a few bits that spoilt it.can i call this dontas inferno.?thanx donta for the heads up on what may happin to my rig i think i will lay off oc for a while maybe put a probe on my 24 min keep an eye on it cheers

I find it rather ironic that DFI, which is out of the enthusiast mobo market, had the foresight to add additional PCI-E power connectors (FDD style) to its boards for years...

Well i'm worried about this problem since i'll have Tri SLI 480GTX cards on my 760 MB.  Still in the process of buying the final parts for the water cooling and now i read this.
 
So, 8 days left on my step up and dont see the new board as an option for step up.. Do i just mod the board and see what happends? Or.. buy a new Class 3 for $365 or grab a 4 way 762 and a new EK full block?? The last option would cost me $380 for the board before rebate and another $140 for the new WB for a total of around $576 after overnight shipping.
I'd go for it but ofcourse i cant find another Galaxy 480GTX and DD all nickle MB for the 4th card to match the other 3. :( There now selling a SC version thats all blue! :(
I guess freakin mod the thing.. but boy its gona be fun to find a spot to plug it in.

Yes that really sucks.
In the same boat too and just got sunk.

Just do the mod or get a 770 or a 4 way. It takes some kind of setup to make it happen. If you have a 759 I would do the mod if you have a  760 and can stepup do that. If you cannot afford to get a 770 with a 760 do the mod.  It wont always happen "have to have the right setup for it" but when it does happen it sucks. But yeah i have to agree doing the mod yourself you should not have to do that kind of thing, it should had been fine out of the box....

Welp, i guess that last talk with the wife settles the matter for me. Mod the board as blowing that much more to fix the problem with buying a new MB and another full MB water block would be nuts. I guess another $2-300 in the end after selling this one and after the rebate is worth more then all the money i got in this new system and risking the lifes of my family and everyone else in this condo complex.
 

Well, I have an update from my end, post Molex mod on the E759.
 
I have soldered on two +12V Molex connectors to the back of the board, one per each +12V pin of the 24-pin ATX connector.  I have tested the system for a few hours (approx. 4-6 hours total now) under LinX, Multi-GPU FurMark 1.8.2 Xtreme Burning Mode with Displacement mapping enabled, and also with both running simultaneously while overclocked (CPU at 4.21GHz, HD5850s at 1000/1250), and the voltage on the +12V rail seems more stable than before. 
 
Under LinX with all DRAM tested the drop is 0% (voltage stays steady at 12.04V).  When loading the CPU and GPUs fully voltage drops to 11.81V which is only 1.91% drop, perfectly acceptable.  Before the mod I was getting 1.25% drop on the +12V rail when under LinX all DRAM load only, and 2.90% drop when under LinX and FurMark load.  There definitely seems to be a positive impact of this mod on the stability.
 
I also noticed that the peak amperage reported by GPU-Z at 1.2875V for both HD5850 GPUs overclocked to 1000/1250 does not exceed approx. 60A.  Prior to the mod it would hit 75A+, even as high as 85A per GPU when running FurMark at 1920x1200 resolution with settings specified above.  The temperatures also dropped by about 3C-5C at load on the GPUs (my ambient temperatures are consistent to within 2C).
 
I am quite confident that this was a low voltage induced burnout because the pins simply could not supply enough to meet the demand of the CPU and GPUs at the same time.

The mod was dirt simple, and it is very clean.  You cannot see anything at all from the front.  Extra Molex connectors are nicely tucked away at the back of the board and hook up like fans and drives behind the motherboard tray of the case (between it and the back panel).
 
Edit: Just to illustrate, here is a quick run I started in the last few minutes.  Look at the load and the voltage on the +12V rail