Help!!! CUSL2 Problems

[Fisher999]

Help! I'm putting together my Asus CusL2/PIII system and have been experiencing these problems with JUST the mobo, cpu, RAM, DVD-ROM and 2 hard drives installed thus far
with NONE of the other expansion cards yet installed.

Symptoms:

Lock-ups running McAfee VirusScan v 5.10 (works fine on my other systems)
Lock-ups decompressing some compressed files
Lock-ups during the "drives benchmarks" with SANDRA 2000
Inability to run Mushkin HP rev 1.5 2-2-2 at 2-2-2 (must use 3-3-3 setting or system
will freeze during post)
Other instabilities

System Configuration and other details:

AbleCom 300W P/S in SuperMicro SC760A case
Asus CUSL2 version 1.02 with lastest BIOS update version 1002 [9-13-2000]
Intel PIII 800EB Retail-boxed (133 FSB) cooled with Golden Orb for FC-PGA's
256 MB Mushkin High Performance Rev 1.5 PC-133 SDRAM [2 X 128 MB double-sided DIMMS]
Promise Fasttrak66 controller and 2 identical Maxtor model 51024U2 10GB, 7200 RPM, ATA-66
drives in RAID 0 (stiping) configuration, partitioned and formatted with Partition
Magic 5.0 in 3 primary partitions (one is an extended partition with two "logical
partitions&quot, Drives both set as "Masters" and attached to Fasttrak IDE1 and IDE2
channels
Pioneer DVD-115 DVD-ROM Drive as "Master" on Mobo's Primary IDE Controller
Gigabyte GA-GF2560T (GeForce 256) which worked fine in another system with version
....0522 drivers....CURRENTLY using version ....0168 detonator drivers

Windows 98 First Edition (Full version, "clean install&quot ver ...1998 w/ se updates

TROUBLESHOOTING strategies already tried:
- brought BIOS RAM settings down to 3-3-3
- removed one of the DIMMS and ran system on just ONE 128 MB DIMM
- switched DIMMS and tried again using the OTHER DIMM
- lowered BIOS settings down to 100 MHz bus speed for RAM
- lowered FSB setting to 100MHz so that processor ran at JUST 600MHz
- checked CPU voltage and temperature with BIOS program and "Asus Probe", everything
checked out fine, as did fan speeds and ALL OTHER voltages
- ran scandisk (no disk errors)
- ran Fasttrak "Fastcheck" utility, again, everything checks out fine
- visited Asus and Promise websites and read FAQS and known problems
(I'm aware of the 2 double-sided PC-133 DIMM -4 Row- limitation with the CUSL2)
- verified on Promise's web site that Fasttrak66 supports partitioning and formatting
with Partition Magic (which it does)
- verified all jumpers are correctly set and set all BIOS settings to "setup defaults"
with exceptions mentioned

I should mention that the system DID run for 16 hours running the looping demo of
Dagoth Moor Zoological Gardens with NO system crashes. (System was left unattended
and upon my return the demo was still going).

I'm baffled by the system instabilities mentioned above. It's a pain to decompress
a compressed executable (like when I FIRST tried to install DMZG) to have system either
"freeze" or, as in the case of McAfee, hang and spontaneously reboot. And I haven't even
yet installed the Sigma Design's REALmagic Hollywood Plus Decoder card, ATI TV Wonder, Linksys LNE-100TX NIC or Creative Lab's SoundBlaster Live! MP3+ YET!!!

Any comments or suggestions will be GREATLY APPRECIATED!!!!

Greg

 

[Fisher999]

ttt

 

[Fisher999]

PLEASE HELP!!!!!

 

[Fisher999]

<<<<<<<<<<< begging now!!! OK, I'll check later on today... thanks.. Greg

 

[LXi]

Where is Radboy and Midnight Rambler when you need them? Anyway, try visiting cusl2.com, they might already have the solution to your symptoms.

 

[AndyHui]

I thought Maxtor drives were not supposed to be very co-operative under RAID?

You DID install the Chipset inf drivers right?

Upgrade to BIOS 1003 Final (without audio).

You may want to try setting the SDRAM driving strength to Strong if you think there may be problems with the RAM.

Start off with that, and see if it helps.

 

[Fisher999]

Well HI Lxi and Andy, long time no see! Nice to see familiar faces... well names anyway... Andy has a new avatar.. I like it!!!

Can you believe I've had the Samsung 900NF for almost 2 weeks now Lxi and I still havn't had a chance to hook it up. I WILL today though, even if I DON'T resolve the CUSL2 problems.

Andy, I THINK I installed the inf drivers properly. I followed the directions in the manual and used the inf drivers on the disk and everything &quot;cleared up&quot; in &quot;Device Manager&quot; so I'm assuming I installed them correctly. Thank you for your two links. I will upgrade the BIOS.. I feel it can't do any more damage at this point. I am beginning to suspect that I might just have a &quot;bum&quot; CUSL2, after all, Intel has had one disaster after another the last 18 months and maybe the i-815e chipset is no exception. I STILL feel I was scammed by Mushkin, the High Performance Mosel rev 1.5 DIMMS are advertised as CAS 2-2-2 but the SPD, according to the Asus board, reports them as 3-3-3.

As far as the &quot;Strong&quot; setting for the SDRAM, I had that set to &quot;normal&quot; because I thought that &quot;strong&quot; was a setting that might &quot;stress&quot; the RAM. No information is available about it in the owner's manual because it must NOT been an option in the original BIOS. I WILL give it a try though. I also visited my Brother and &quot;borrowed&quot; some Samsung -tc75 PC-133 RAM (2 64 MB DIMMS) that were purchased from Mushkin earlier in the year. I want to stick them in the system and see if THAT eliminates the problems. That way I can be pretty sure that I have &quot;bum&quot; RAM and the CUSL2 is not the problem. The only other CPU I have on hand (short of pulling ones from the system I built for my brother and the one I built for my folks) is a barely-used retail-boxed
Intel Celeron 400 MHz PGA. I might also try popping that in to see if maybe the board just can't handle a 133MHz FSB processor or that the Intel 800EB is &quot;bum&quot;. I will do these swaps leaving ALL OTHER VARIABLES CONSTANT so I can be confident of the results of my experiments. Andy, your link to the inf drivers is not currently working, I'll try THAT one later.

Lxi, thanks for the link to www.cusl2.com (I was unaware of this site). It seems to hold a wealth of info on the CUSL2 and I thank you very much.

I will repost later on tonight. Thanks Lxi and Andy.
Greg

 

[Akaz1976]

well here is my CUSL2 thread, when i had problem with it. i had to exchange the board and now it runs fine. u can find people who are well versed with this board and such problems on my thread tho.

hope it helps

Akaz

 

[yathinkicare]

heat might be the issue.(no its not. i didn't read your thing about the looping test!!). is your case fan working ok?

Also.. try using standard vga drivers to c if its' a video driver issue...(doubt this too!).

can also try turning off raid and using one plain partition

It sounds like a memory or board issue to me.. I would take one drive out for sure...

 

[Fisher999]

Andy, btw, I hadn't heard anything about Maxtor drives not cooperation under RAID. If that IS the case, I guess I'll be sorry that I purchased them for the RAID 0 setup.

It's interesting that you mentioned that because it has occurred to me that the problems that I have been experiencing have been ones which are DISK-intensive as well as CPU and RAM intensive.

Ponder this!
1) The &quot;lock-ups&quot; occur when uncompressing zipped files (and where are the files stored but on the hard drives of course) and other situations that involve intensive use of the disks, like
with the virus scans (when you can hear the hard drive &quot;thrashing&quot.

2) Althought the &quot;drives&quot; benchmark freezes in SANDRA 2000, the &quot;memory bandwith&quot; test and the CPU tests, which I naturally assumed were going to &quot;freeze&quot;, did in fact complete (albeit a bit slowly) and the &quot;numbers&quot; reported were impressive.

3) The Dagoth Moor Zoological Gardens looping demo &quot;froze&quot; when the installation file was decompressed and again on the first attempt to run the program, BOTH DISK INTENSIVE situations, BUT... upon starting the program the second time and the files were &quot;loaded&quot;, the REST of the 16 hour period that the benchmark ran (without a hitch), with 256 MB of RAM, the &quot;swap file&quot; was NOT being hit on the drives; just p/s, mobo, cpu, ram and videocard were basically
&quot;running the demo&quot; for THAT 16 hour period.

ANDY, you MAY have steered me towards the &quot;weak link&quot; in the sytem chain. Possibly either the Fasttrak66 is &quot;bad&quot; or just not very compatible with the Maxtor drives. (Fasttrak66 has the latest BIOS and driver revisions installed but...????)

Anyway, thanks again guys, I'll get back to you. And if you think of anything else please let me know.. Thanks....... Greg

 

[Fisher999]

AKAZ, THANKS for the link to YOUR CUSL2 thread....

YaThink... The SuperMicro Case has one 80mm inlet fan at the front of the case drawing air in and two exhaust fans (the one in the Ablecom power supply and a 92mm exhaust fan at the TOP REAR of the case). Those three fans AND the Golden Orb's fan (for the CPU) are ALL running just fine.

I tried running the VirusScan in windows &quot;safe mode&quot; (which only loads VGA drivers) and just minutes after the scan was started the system &quot;froze&quot;.

May I ask you this? Is there a way to disable one of the drives in the RAID 0 array WITHOUT having to &quot;delete&quot; (Promises's terminology) the array and repartition and reformat the drive?
I don't think I could try your suggestion without deleting the array and repartitioning and reformatting, then trying just ONE drive hooked the MOBO's IDE channel. That would involve a LOT of TIME because I already have the entire OS, the second edition updates, and several other programs installed. I WILL keep this in mind though and try this as a last resort BEFORE calling Mushkin or ASUS about product replacements.

Thanks ALL...

 

[Fisher999]

bump

 

[mcbiff]

It definitely sounds like your hdd setup is to blame. Just put one of the drives on the mobo IDE and try that. I don't think you can use a drive from an array without re-partitioning it though, so get ready for some reinstalling.

 

[AndyHui]

Perhaps you could try and borrow a hard drive from someone. Setup and install your system on that, independent of your RAID array and see if that gives any problems. If it does not, then it appears that we have narrowed it down to the RAID array.

Traditionally, setting SDRAM driving to Strong stabilises a CUSL2 system with 3 PC133 SDRAM DIMMs.....you know that having 3 PC133 SDRAM DIMMs in other i815E systems results in instability. It also helps with DIMMs that are PC100 and you are trying to push to 133MHz.

You should install the inf drivers from the link that I gave you....they are more up to date.

Intel's ATA drivers are normally not necessary for operation, so I will ignore them for the moment as being a possibility.

IBM's GXP75 drives are known to be the best under the RAID setup. No problems, and they perform well.

 

[Fisher999]

Andy: Thanks, that's my last resort and I'm about to start it now at 10PM EST. Check back if you have time 'cause I'll post with this Dell (sure am glad I still have this sweetheart) and let you everyone know if the FastTrak66\Maxtor RAID setup is the culprit.

Greg

 

[JaiKnight]

I've got a setup similar to yours, running 75GXPs instead of the Maxtors though with FastTrak66 RAID0. I'm just wondering, how in the world did you get Partition Magic 5 to work with it? I can't start that program up for some reason, it doesn't load properly! I thought it might be that it doesn't understand the RAID array?

 

[MushkinTechs]

<< I STILL feel I was scammed by Mushkin, the High Performance Mosel rev 1.5 DIMMS are advertised as CAS 2-2-2 but the SPD, according to the Asus board, reports them as 3-3-3. >>



The reason your motherboard is reporting the REV 1.5 module as 3-3-3 is because we program the SPD's for all of our CAS 2 modules to run at CAS 3 by default when the memory timings in the BIOS are set to automatic. We do this because there are a lot of other BIOS settings that can interfere with SDRAM running successfully at CAS 2.

If we didn't set the SPD to default to CAS 3 we could be inundated with calls because it is harder for our less technical customers to get there motherboards to run as stable at CAS 2 than it is when memory is run at CAS 3. This could result in increased prices because of the cost of adding staff to help with the increased amount phone calls, and possibly a temporary decrease in customer service standards. If the customer doesn't know how to manually set the BIOS timings, then we think it's best for them to temporarily run there computer at more conservative settings until they become familiar with there motherboards BIOS and there computers stability. This is because the amount of frustration caused by the possibility of having an unstable computer far outweighs any performance difference.

For customers who aren't familiar with there motherboards BIOS, I would suggest the following links:
(WinDrivers BIOS Center)
(Adrian Rojak Pot's BIOS Optimization Guide)
(BIOS Survival Guide)
(BIOS Tweak Guide)

I just want to make it clear that we are not selling modules advertised as CAS 2 that only work at CAS 3. Our CAS 2 modules can be run at CAS 2 but they have to be setup to do so manually through the BIOS. In some motherboards setting the CAS timing to 2 when the modules SPD is programmed for CAS 3 will result in a warning message at bootup that says &quot;SPD settings recommend CAS 3&quot;. This will not affect stability or performance and can be safely ignored. And having to set the memory timings manually won't detrimentally affect performance either.



<< Inability to run Mushkin HP rev 1.5 2-2-2 at 2-2-2 (must use 3-3-3 setting or system will freeze during post)

As far as the &quot;Strong&quot; setting for the SDRAM, I had that set to &quot;normal&quot; because I thought that &quot;strong&quot; was a setting that might &quot;stress&quot; the RAM. >>



The REV 1.5 modules have been tested to be compatible with the ASUS CUSL2. Unofficially, I have gotten three REV 1.5 at a time to work at 2-2-2 on the ASUS CUSL2. But for anyone looking for more than 256MB, I would suggest our 256MB REV2 modules as Intel doesn't support running all three DIMM slots at 133MHz. There are a few reasons that the REV 1.5 might not be working but lets start by double checking your BIOS settings.

Under the 'ADVANCED MENU' choose 'CHIP CONFIGURATION' and set:
* SDRAM Timing: [User Define]
* SDRAM CAS Latency: [2T]
* SDRAM RAS to CAS Delay: [2T]
* SDRAM RAS Precharge Time: [2T]
* SDRAM Cycle time: [6T,8T] or [7T,9T] - Depending on which BIOS version you have. After you verify that your system is stable try [5,7], but that may or may not be stable.
* SDRAM Page Closing Policy: [All Banks] - After you verify your computer as stable, move to [One Bank]. This setting will only slightly affect real world video performance in 3D games, although you can ignore this setting if you are only using an external video card.
* Command Per Cycle: Disabled - After you verify your computer as stable, move to [Enabled]. This setting will only slightly affect real world video performance in 3D games, although you can ignore this setting if you are only using an external video card.
* SDRAM Data Driving Mode: [Normal] - Change to [Strong] if you have instability problems, especially if you have multiple memory modules installed. If it is set to [Strong], and you are having problems with one or multiple memory modules, change this setting to [Normal]

Also, make sure that the I/O voltage jumper on your motherboard is set at its default of 3.4V. If none of this helps, you may want to try a lower or slightly higher voltage.

 

[Fisher999]

Mr MushkinMan: THANK you sir for your response to my &quot;suspicions&quot; about the Mushkin High Performance Rev 1.5 memory modules. I am WELL aware of the memory timing settings you mention. As for the SDRAM CYCLE TIMES setting in BIOS (I am using BIOS revision 1003), the system will NOT pass POST at the setting of 5T/7T.

I ALSO appreciate your suggestions for the SDRAM PAGE CLOSING POLICY setting and the COMMAND PER CYCLE setting and will follow YOUR suggestions.

I am signing off now to talk to mushkin representatives about trying to &quot;resolve&quot; their POSSIBLY &quot;deceptive&quot; advertising of the rev 1.5 modules.

Although I can appreciate why you program your SPD's on the DIMMs in question to &quot;default&quot; to 3-3-3, I still feel very strongly that if a mobo's &quot;auto&quot; or &quot;SPD&quot; setting detects RAM as CAS 3-3-3,then as far as I'm concerned, what I was sold IS CAS 3-3-3 RAM when in fact both your website and my receipt describe it as 2-2-2.

 

[Fisher999]

Mushkin Techs: BTW, though I forgot to mention it in the original post, changing the I/O voltage jumper to the 3.3 Volt setting (from the default of 3.4) was also one of the troubleshooting solutions I tried early on. This had no perceptible effect on system performance, either positive or negative.

 

[Fisher999]

I FORGOT to post that I did in fact determine the Promise FastTrak66/Maxtor RAID 0 combo to be the &quot;culprit&quot; of my system problems. After pulling out the fasttrak, repartitioning and reformating the Maxtors, setting them as Master/Slave on the primary IDE channel, peforming a &quot;clean&quot; reinstall of Win98 and some apps..... BAMMMM... NO MORE HANGS, FREEZES, LOCKS....

I think I'm going to LIKE the CUSL2/Mushkin combo. (Still wish I wouldn't have allowed myself to be swayed from purchasing Mushkin rev 2 though).

I want to thank EVERYONE for your help in my getting this DISASTER resolved....

Greg

 

[Radboy]

Thx for posting the findings. Love my CUSL2. You're not the first to have probs with a FT. I'd like to set one up, just to say I did. Is it a Maxtor thing? .. that has FT probs?

You have your OS/apps on a stripe?

 

[bmg]

A short time ago I posted my cusl2/Promise Fasttrak100 problem on the Motherboard forum. Basically I get great read performance but horrible write performance. It's looking like there's an basic problem with the cusl2/Fasttrak. I'm using two IBM 45Gb drives with the Fasttrak. It works great with my KT7 system.

The write performance is so bad that it appears that the system is hung. Given enough time the write eventually finishes.

 

[MushkinTechs]

<< what I was sold IS CAS 3-3-3 RAM >>



The REV 1.5 module isn't a 3-3-3 module, it is a 2-2-2 module. The SPD on the module is programmed so that it will default to 3-3-3. Programs like SiSoft Sandra don't really determine what the module is actually capable of as far as memory timings, they just read the SPD on the memory module and report it back in a readable fashion.

 

[Radboy]

Now that's what I call tech support.

 

[Midnight Rambler]

Note that Mushkin is not the only company that sets their SPD like this. Micron/Crucial sets up their CAS 2 modules the same way.