I have a 132-BL-E758-A1 (X58 SLI) Hardware Rev. 1.1 and used the diagram to do the resistor mod. and just placed a Xeon X5680 I got off Ebay for $125 and removed my i7 920 C0 stepping poor overclocker. I think I got really lucky and bought from someone with NEW old stock, as the CPU heat spreader was so pristine and there was no marks on the pads that I don't think it was ever even tested! The seller had sold 53 before I jumped in on the same sale that is now done! (Seller:Keystone Memory Group, LLC)
The R538 resistor is the smallest SMT resistor I have ever seen. I have a Weller soldering iron station and a hot air station and have done some very difficult soldering and have soldered every trace and repaired two layer cracked boards and reballed and replaced small BGA ram chips and think nothing of replacing any connector or cable header even in a small phone. It is also located between the AUX_FAN black plastic header and the edge of the north bridge heat sink. I took some measurements with my meter and the side of R539 closest to the north bridge is tied to ground so this is a pull down to ground resistor. This means that it can be replaced with a dead short if needed. As the clearance is so small, I was afraid to get my hot air in there and grab it with tweezers as I know the resistors that are half under the north bridge could blow away under the heat sink and that would be the end of the board. The limited angle of attack meant that I gained NOTHING by removing the board from the case. I only removed the Video card. That is it! The rest of the case was undisturbed, including the memory.
I set my soldering gun on the lowest setting that would melt the crappy RoHS solder and took my smallest soldering wick and cleaned the crap solder off the two pads for R539 and added a good glob of flux and re-tinned the pads with good lead solder and made two nice new solder beads on the little pads. It looked so pretty at this point! (No Photo taken)
The head of my soldering Iron was big enough to easily touch both sides of the resistor at the same time. I touched it from the side at the only angle it can be attacked between the header and the north bridge heat sink and it instantly came loose and I pushed it over to the freshly prepared pads of R539. The resistor would not grab to the prepared pads when I melted the prepared solder and there was no room to get a pair of tweezers or even a dental pick in there at the same time with my other hand. The resistor simply stuck to the tip of the soldering iron and would not let go! By the time I gave up trying to get the resistor to stick to the new pads and brought it over to my wet sponge, it had melted and dissolved into nothing but a black speck.
At this point, I had nothing to loose by replacing the resistor with a dead short across R539, as a pull down logic resistor is not really needed if the input is simply a logic high or low input and not voltage output. I took my dental pick and scraped off the now dirty resin covering the pads and applied fresh resin coating and ever so lightly added a bit of solder and dabbed between the two pins and got a small short across R539 on the third attempt. It was not pretty, but I was worried I could dissolve the pads all together if I went for a perfect looking soldering job. I had tinned a thin wire wrap wire, as I would normally not bridge something without wire but decided against continuing after I tested the "VIA" point and it was now shorted to ground through the shorted pads of R539. You can see that the side or R539 is tied to a "VIA" Hole in the board that goes through to the other side of the board. If someone were to rip the pads off the board for R539, you could short the VIA point on the bottom of the board to ground and recover the mod that way.
MOD:
BEFORE: (Taken while system running, so not best shot)
AFTER MOD: (Yeah, there is still some excess flux around the aux fan connector. It is not hurting anything and will melt away.)
CPU:
I set the CPU BCLK to 144 with default voltages for my initial successful overnight test. I have speedstep and turbo turned on in the bios and it is stepping from 1.59 GHZ to 3.8 Ghz and the turbo multiplier is going to up to 27 with the highest BIOS setting being 25x since this is a x5680.
MEMORY:
I have 3 sticks of unmatched OCZ ram from when I first built this thing back in 2009:
Memory Type: DDR3
Part Number: OCZ3G1600LV2G 㫨
Serial Number: 00000000
Size = 2048MB(double side)
Size = 2048MB (BRC=3x14x10)
Speed = 533MHz (DDR3_1066)
CL: 5,6,7,8
Taa=13.125ns, Trcd=13.125ns, Trp=13.125ns
Tras=30ns, Twr=15ns, Trrd=7.5ns
Trc=49.5ns, Trfc=110ns, Twtr=7.5ns
Trtp=7.5ns, Tfaw=30ns
533MHz (DDR3_1066): Tck=1.875ns
Tcl=7, Trcd=7, Trp=7, Tras=16
Twr=8, Trrd=4, Trc=27, Trfc=59
Twtr=4, Trtp=4, Tfaw=16
400MHz (DDR3_800): Tck=2.500ns
Tcl=6, Trcd=6, Trp=6, Tras=12
Twr=6, Trrd=3, Trc=20, Trfc=44
Twtr=3, Trtp=3, Tfaw=12
333MHz (DDR3_667): Tck=3.000ns
Tcl=5, Trcd=5, Trp=5, Tras=10
Twr=5, Trrd=3, Trc=17, Trfc=37
Twtr=3, Trtp=3, Tfaw=10
I then added three more unmatched 8 MB sticks with an actual 1333 Mhz SPID rating in the ram a few years back that I got on a black Friday sale for $99 a stick. I was going to pull the old ram, but I got greedy and they played nice together at the OC speed I was getting out of the i7, so I left them in there.
Memory Type: DDR3
Part Number: 9905403-559.A00LF
Serial Number: CD310A53
Size = 8192MB(double side)
Size = 8192MB (BRC=3x16x10)
Speed = 666MHz (DDR3_1333)
CL: 6,7,8,9
Taa=13.125ns, Trcd=13.125ns, Trp=13.125ns
Tras=36ns, Twr=15ns, Trrd=6ns
Trc=49.125ns, Trfc=260ns, Twtr=7.5ns
Trtp=7.5ns, Tfaw=30ns
666MHz (DDR3_1333): Tck=1.500ns
Tcl=9, Trcd=9, Trp=9, Tras=24
Twr=10, Trrd=4, Trc=33, Trfc=174
Twtr=5, Trtp=5, Tfaw=20
533MHz (DDR3_1066): Tck=1.875ns
Tcl=7, Trcd=7, Trp=7, Tras=20
Twr=8, Trrd=4, Trc=27, Trfc=139
Twtr=4, Trtp=4, Tfaw=16
400MHz (DDR3_800): Tck=2.500ns
Tcl=6, Trcd=6, Trp=6, Tras=15
Twr=6, Trrd=3, Trc=20, Trfc=104
Twtr=3, Trtp=3, Tfaw=12
333MHz (DDR3_667): Tck=3.000ns
Tcl=5, Trcd=5, Trp=5, Tras=12
Twr=5, Trrd=2, Trc=17, Trfc=87
Twtr=3, Trtp=3, Tfaw=10
I bumped the CAS and RAS rate from 8 to 9 to keep the old ram happy and set the ram voltage to 1.6 volts and have a total of 30 GB. I placed the 2 GB sticks in slots 1,3,5 and the new 8 MB ram in slots 0,2,4.
I'm not sure how far I'll push this machine, as I use it as a media server and my daily runner for everything and it runs 24/7 since 2009 and has clocked over 7 years of continuous service for me without any problems. I have all 10 drive slots in use and it sits in a super tower with a Zallman copper/heatpipe air cooling block on the CPU and a much larger muffin fan glued to the north bridge after the MB fan gave out a few years ago.
MONITORING:
If you have any suggestions on voltage and other timings to adjust that worked for you with this CPU and memory type please reply.
post edited by scanman1 - 2016/07/06 06:30:08