Finagle's Fourth Law: "Once a job is fouled up, anything done to improve it only makes it worse."

In the Trenches with LAROKE

Konsultant's Log, Cyberdate 05.25.1999 (Constructing Grumpy Part II)

SITREP: The last log entry (Cyberdate 03.29.1999 Constructing Grumpy Part I) that dealt with "Grumpy", the new PC for the Boss's daughter, ended with ominous hard drive problems. I knew the new Seagate 3.2GB HDD purchased for the construction of this machine was trashed, but I didn't know what trashed it.

I had ordered and received an equivalent replacement drive, this time a Western Digital model, but I was currently fearful of trashing it too. I wanted to run some tests while I gathered my courage.

TACAMO: 5:16 PM 1/28/99 The new RAM Memory passed all the PC Clinic tests. I replaced it with the memory DIMM that Windows 98 had the problem with and tested it . . . It passed too! After some stumbling around, I figured out how to make PC Clinic test the memory continuously until I press the ESC key, then set it running. We'll see what the results are tomorrow.

7:25 AM 1/29/99 After running memory tests all night, log shows all passes. Aborted main memory tests and started running continuous external cache tests.

9:00 AM 1/29/99 After 1-1/2 hours of external cache memory tests (all passing grades), a continuing series of CPU (main system) tests were started.

Back to Phase III (outfitting the hull)

10:02 AM 1/29/99 After 1 hour of CPU (main system) tests (all passing grades), the testing was aborted. Grumpy was shut down. The PCI NIC was removed as well as the defective HDD. The new Western Digital Caviar HDD was opened and settings checked. The default "Single" jumper was changed to "Master". Drive parameters are 6296 cyl, 16 hds, 63 spt, 3249.3 MB. The new HDD was installed in the drive cage which was in turn installed in Grumpy.

10:54 AM 1/29/99 Grumpy would not recognize the new WD Caviar HDD, no matter what the BIOS settings. After several trys, the 539 MB Maxtor was reinstalled for testing. ~!@#$%^&. Grumpy was started and FDISK run on the Maxtor drive from the Windows 98 Start disk. Grumpy was rebooted and FORMAT C: /S run again.

11:06 AM 1/29/99 Grumpy's cover was replaced and PC Clinic started. A continuous run of HDD tests was started.

11:02 AM 1/30/99 The WD Caviar was installed in 4-Bits to determine if it was DOA (dead on arrival). 4-Bits correctly identified and configured it. I'm beginning to suspect the Biostar M5ALA motherboard is the cause of Grumpy's multitude of maladies. I ordered an ASUS P5A Super Socket 7 motherboard from CDW at this point.

Setback to Phase II (relaying the keel)

12:27 PM 2/1/99 ASUS motherboard arrived a few minutes ago. CDW's packaging leaves a bit to be desired with this shipment - no bubblewrap or anything else, the mobo package was loose in a much larger CDW box - will have to set up soon in case of shipping damage.

11:13 AM 2/3/99 Opened Grumpy and removed HDD cage and all ribbon cables and connectors from Biostar motherboard. Removed motherboard and transferred Intel P55C CPU (200MHz MMX), heatsink and fan to new ASUS motherboard. Set motherboard jumpers on ASUS board to match P55C processor. Moved 8x64e 64MB 168-pin SDRAM DIMM memory module to new mobo.

11:41 AM 2/3/99 Checked the mobo mounting holes for matches with the hex metal standoffs on the Enlight case. One standoff was located where the ASUS board did not have a hole. A nice feature of the Enlight case is that the standoffs are screwed into tapped holes rather than press-fitted, a practice used in cheap cases. I was able to unscrew the standoff and move it to another location where the mobo did have a mounting hole. I also added a standoff for the ASUS board's sixth mounting hole. Using my plastic stirrer alignment trick (See Cyberdate 03.29.1999 Constructing Grumpy Part I), I placed non-conductive washers for all six mounting holes and loosely mounted the motherboard. The ill-fitting PCI Video adapter was inserted before tightening the motherboard mounting screws. It fit a little bit better with the ASUS board than it did with the Biostar.

OHMPH! Brick wall number three!

12:27 PM 2/3/99 Attached all motherboard connectors and turned Grumpy on - She would not start. Checked the connectors and they are OK. The power supply may not have enough juice and the saga continues.

8:11 AM 2/4/99 I ordered a new power supply from Insight yesterday afternoon. With luck it will arrive today. In the mean time, I switched DIMM modules to see if that would let Grumpy boot with the new mobo. No joy there. I rechecked the ASUS jumpers for the third time and all are correct for the P55C CPU. I rechecked the mobo connectors - also Kosher. ~!@#$%^& This isn't fun anymore and I'm ready to rename this machine Dottie (after the asteroid in the movie Armageddon).

2:08 PM 2/4/99 The replacement power supply arrived thirty minutes ago. I installed it in Grumpy, connected the minimum devices (FDD and video adapter), and started the PC. Success! Grumpy started this time. However, I noticed that both the CPU fan and power supply fans started as soon as AC power was applied (before pressing the power switch on Grumpy's case). The fans continued to run after the power switch is turned off - Is this a feature or a defect?

I lean in the direction of defect, but it turns out to be a feature. New motherboards that have "wake up when the modem rings" or "wake up when signaled by the network" features must be under power all the time. Those of you working on your own PCs - Unplug the machine from the wall. It is no longer safe to rely on the on/off switch (if it ever was). You can ZAP one of these new mobos real quick, or worse yet, ZAP yourself.

I unplugged Grumpy and set her up with the original DIMM memory. Grumpy was turned back on and BIOS Setup entered. Minimum BIOS Defaults were selected and saved. I started PC Clinic running a series of tests (CPU Main System tests first, Memory tests second).

Back to Phase III again (outfitting the hull)

3:00 PM 2/4/99 Installed the new WD Caviar HDD but Grumpy's BIOS Auto-detect feature had trouble finding it every time until I changed the jumper setting on the drive to "single" (or neutral) setting from "master" and changed the ribbon cable connector from the end connector to the middle connector. Since both these changes were made at the same time, I don't know if it was the jumper (unlikely) or the cable (possibly a short between end and middle connector?) that made the difference. After this change Grumpy's BIOS correctly found and identified the Caviar drive.

After using FDISK to make one large Windows 98 FAT32 partition again, Grumpy was rebooted and the HDD was formatted with the system parameter. Then PC Clinic continuous HDD testing was started.

11:08 AM 2/5/99 Reconnected Memorex CD-ROM drive. During connection, HDD drive cage (with drive in it), was sitting loosely on top of the power supply. It slipped off and fell on the mobo. ~!@#$%^ my fault! There is no visual damage, but it would be just my luck for this incident to cause problems with this machine. Grumpy booted up OK for the second day in a row and I started PC Clinic again and ran one pass of all available tests. If everything tests OK, we will try installing Windows 98 again.

2:56 PM 2/5/99 Once again, the Windows 98 installer rejected the original DIMM memory module by bringing up the "NTKERN" error message. The other DIMM memory module was reinstalled in its place and setup was continued from where it broke down (after first starting in safe mode and shutting down again).

3:39 PM 2/5/99 Everything seems OK, but by now, just looking at Grumpy makes me skittish. Next, a minimal system boot disk with FDISK, FORMAT, MSCDEX, CONFIG.SYS, AUTOEXEC.BAT and CD-ROM DOS drivers was prepared. Then the AFLASH.EXE and AFLASH.TXT flash memory writer files were copied to it from the ASUS CD that came with the mobo. Grumpy was restarted with the boot diskette and AFLASH was run to copy a backup of the current flash BIOS to the diskette (filename: CURRENT). Went to ASUS site to find most current flash BIOS for Grumpy. Grumpy's current BIOS number is Revision 1004 and the newest one at the site is Revision 1005. The new features added did not apply to Grumpy at this time, so I did not download and upgrade her BIOS.

4:50 PM 2/5/99 Installed Diamond Multimedia Sonic Impact S90 Sound Card, Creative Blaster Speakers and drivers. Tested Diamond's CD Player and 3D sound samples. Checked Device Manager and everything is in order.

5:20 PM 2/5/99 Installed EtherPCI LAN adapter and cabled Grumpy to the network. Started Grumpy and installed network components for testing - will finetune later. Everything works and there don't seem to be any conflicts at this time.

2:59 PM 2/6/99 Installed Logitech wireless scroll mouse.

4:17 PM 2/9/99 Finetuned default desktop and enabled multiple users (user profiles). Ran first through SCANDISK and DEFRAG.

2:08 PM 2/10/99 Started to finetune BIOS settings. First, the optimized "Setup Defaults" were loaded. Then, in the "Standard CMOS Setup", "Primary Slave" and "Secondary Slave" Type settings were changed to "None". In the "BIOS Features Setup" section, "Boot Sequence" setting was changed to "C,A".

AARRUUUUGGHH! Brick wall number four!

Installed Diamond Multimedia Stealth II G640 AGP video adapter - mouse cursor began to act erratically. Mouse buttons were malfunctioning also. The video drivers were set back to Standard VGA Display (via keyboard commands ~!@#$%). Grumpy was shut down and the AGP adapter was replaced with the PCI test adapter. When Grumpy was restarted, new drivers were installed (again via keyboard because the mouse still wasn't working properly). The Diamond Display software and the mouse driver software was uninstalled. The mouse was removed from the Device Manager and Grumpy was restarted to reinstall the mouse drivers. Good, the mouse is working again.

5:05 PM 2/10/99 Ordered an ASUS v2740 AGP video adapter (same manufacturer as motherboard for compatibility).

4:39 PM 2/12/99 ASUS AGP video card arrived. Installed it even though the manual stated a Pentium II CPU was required. The readme file stated this mobo is supported even though it is not a Pentium II mobo. Everything seems to be working.

MISREP: At this juncture, the worst of the hardware problems with Grumpy are behind us (I think). We've still got lots of software installation and configuration ahead as well as physically shoehorning Grumpy into her cramped final location in the company reception area. Those tales are for another time.

LAROKE Microcomputer Consultants
155 East Boca Raton Road
Boca Raton, Florida 33432
(561)368-0659 (Tel & Fax)

Issued Tuesday May 25, 1999

copyright © 1996-1999 LAROKE Microcomputer Consultants all rights reserved

Search www.laroke.com the Web for