Adding Two Gigabyte of OCZ Platinum 4-5-4-15 Rev. 1 SDRAM

How Not to Flash Your BIOS

Update August 11, 2007: With memory prices low and before the July 22nd announcement of reduced Intel processor prices, I decided to buy two more gigabytes of memory for my computer. I didn't really need the memory now, but I wanted to add the same OCZ Platinum 4-5-4-15 SDRAM before it was discontinued.

I received the 2 x 1GB matched DIMMs from Newegg and installed them in slots two and four. I left the BIOS VDIMM setting at 1.8 Volts since the replacement DIMMs I had gotten had automatically set the VDIMM setting to 2.03 Volts. I noticed in the POST screen that the 4 GB of memory was recognized but was only running at 533 MHz. Not good.

I decided it was time to flash my BIOS. There were a number of BIOS fixes since the P15 BIOS version I was using including one that affected how memory was reported in the BIOS. I downloaded the file needed for a floppy install and opened my case to install a floppy drive. I have done this before. I just set the floppy drive next to the computer and not properly installed in the case. I tried to put on the four-pin power connector. There was a flash of light! I heard a pop! There was a sickly hint of that bad news electrical smell. The computer immediately died. (Yes, I foolishly had the computer on). I have done this before with the computer on without any problems and had never heard of the possibility of shorting the power pins when putting the power connector on the floppy drive. Definitely not good.

I removed the floppy drive and pulled the plug on the computer. I plugged the main power cord back in and turned the computer on. My mind began to wonder how much damage I had done in addition to the motherboard - the CPU? The memory? The PSU?

My memory is somewhat vague but I believe that the system remained completely dead on the first reboot.

I turned the computer off and back on. The system slowly came to life. The Northbridge fan was not spinning. The front panel 120 MM case fan was barely spinning. The blue LED's were on and then off and then on again. It was all very frightening. For all appearances I had just fried my wonderful Foxconn motherboard. I manually tried to spin the Northbridge fan to get it turning. It tried to spin but just a few revolutions. The front panel 120 mm fan seemed to be spinning faster. If I remember correctly, the PSU, CPU and rear case fans were spinning normally.

I immediately pressed the DEL key to get into the BIOS. The Processor speed and memory speed were a jumbled mix of random characters. Scary not good. It seemed a miracle that the system booted at all considering the garbage I was seeing in the BIOS.

Once again I turned the computer off and left it off for a few minutes. After turning the computer on again all seemed normal. All fans were spinning. A quick look in the BIOS showed normal values. I exited the BIOS and the computer rebooted and Windows started up normally after the typical option screen from an abnormal shutdown.

I exited Windows and shut down the computer. I installed a different floppy drive and restarted the computer. The floppy was not detected. In short, the only apparent damage to the system was the floppy disk controller on the motherboard. Yes, I know I had voided the motherboard warranty and yes I now know that I had made a stupid mistake and yes I consider myself extremely fortunate that no more damage had been done. This Foxconn motherboard must be built very well to take that kind of abuse and still be able to resurrect itself from the dead.

I put in a Kubuntu CD and ran the Memtest 86+ memory test. I was concerned that I had damaged the memory and wanted to test it thoroughly before returning it to OCZ. It made one complete pass without errors.

I still wanted to flash my BIOS, but that obviously wasn't going to happen with a floppy. I had remembered that the Foxconn Live Update application allowed you to flash the BIOS, but it didn't work in the 64 bit version of Vista. I went to the Foxconn website and found a version of Live Update that was Vista compatible. I was making a mental list of what had to happen before I would risk flashing the BIOS from within Windows. Foxconn recommends that the BIOS is flashed from a floppy and not in Windows. Flashing the BIOS from within Windows can void the warranty if it fails. I had already voided the warranty so I wasn't concerned about that. I decided that a clean install of the 32 bit version of Vista with the Vista compatible version of Live Update and no apps installed in Vista was a low risk way to flash the BIOS.

I installed the 32 version of Vista and installed only the modem and Live Update. I connected to the Internet and started the update BIOS procedure in Live Update. I wondered if I was adding more stupidity to an already bad situation. I couldn't watch. I came back several minutes and found that the BIOS flash had been successful.

I shut down Vista and rebooted the computer. I had been running the P15 version of the BIOS. The new version, P30 was now installed and verified during POST. I entered the BIOS and looked for changes.

I was very pleased to see a CPU multiplier setting that would allow for decoupling the memory from the CPU when overclocking. There was also a new setting for HPET that is apparently a better way to handle interrupts.

The more I use the Foxconn 975X7AB-8EKRS2H motherboard the more pleased I am with its performance.

The BIOS update did not change the 533 MHz memory problem. I had a pretty good idea though what the problem was with the memory. I submitted a RMA request with OCZ to return the two new DIMMs to have the SPD settings programmed to match the other two DIMMs. I was somewhat surprised that the new DIMMs behaved exactly like the original two I had ordered in August of 2006. I told OCZ about my floppy drive faux pas concerned that I had voided my warranty. OCZ was kind enough to ignore the issue and wanted me to return all four DIMMs. I told them that it was fine for them to reprogram the new DIMMs and return my original DIMMs, but they wanted to send new DIMMs. That was fine with me and very generous on their part.

I had fortunately timed the purchase before a one week vacation and sent the DIMMs using their RMA process on July 28th 2007. I made a note in the RMA that I would be on vacation until August 7th and to be sure that the new memory modules would not arrive until the 8th. I received the four new DIMMs on August 10th. That was very fast turn-around considering that five calendar days had been spent getting the memory to them. I might have received the memory sooner were it not for my request to have the memory sent after my vacation. As it was, the turn-around was quite acceptable.

I was extremely happy to find that the four new DIMMs were all recognized and running in dual channel mode at 800 MHz. With the new BIOS upgrade, I am now certain that the memory really is running at 800 MHz. Note that I had the VDIMM set at 1.8 Volts - the proper voltage for the reprogrammed SPD settings. The specs for the memory show that the BIOS VDIMM setting should be +.30 Volts to give a final value of 2.10 Volts, but this value should not be changed if you have the DIMMs reprogrammed. I verified in the BIOS that the memory was running at 800 MHz and the PC Health screen reported the VDIMM Voltage at 2.03 Volts. The SPD settings somehow set the proper voltage without manually having to adjust the normal 1.8 Volt setting. I am not concerned that the BIOS reports only 2.03 Volts. The memory is working without errors.

While it is certainly a pain to have to RMA the new memory, OCZ has stellar service and I am very pleased with how it has all worked out. And of course, I am very happy to have four gigabytes of 800 MHz SDRAM in a working Foxconn 975X7AB-8EKRS2H motherboard in spite of my inadvertent abuse.

Feedback Greatly Appreciated. Email To Alan Norton
Back to The Second Intel Build - A Review of the Foxconn 975X7AB-8EKRS2H Motherboard
Back to Pecos SoftWareWorks Home Page

Copyright 2007-2008 Alan Norton Last Updated August 16, 2007