Using RAID port as IDE port

[Bovon]

Howdy daverme. Thanks for the reply. We were trying this with the Iwill board, and the Iwill has a different raid controller I believe...AMI?...sounds right, and that in its self could be a problem. I think I first read where using the raid headers as IDE came from some Abit users...so..maybe it will not work with the other raid type controllers.

How about this, if you have time, review the link above to the amdmb forum...look over what was tried, and see if any of those efforts were consistant with what you did. Then secondly, explain exactly how you set this up. It seems that if you load Windows from the IDE headers, then move the drive over to the raid headers, you have corruption problems with the OS.

Thanks for the input...Bov

[Ace-a-Rue]

Bovon, I found you. You couldn't hide for very long. I could track you very easy with those links! LOL!!!!!!!!

[Bovon]

Okedoke all mainboard raid types input needed.

Several of us started a thread over at amdmb forum concerning using the raid feature as extra IDE headers. The tech support guys say yes, but no one seems to have actually worked out the sweet spot yet.

Here is the thread over at amdmb, and It appears to have died a slow death over there. One reason being, the server that hosts amdmb is the pits. So often you can't get there from here...but, if anybody has some thoughts about using the raid feature as extra IDE headers, go to the following thread to see what has been tried so far, and if you have some suggestions, we will certainly appreciate hearing back from you...thanks, I know somebody, somewhere has the technological expertiese to figure this out.
http://www.amdmb.com/vb/showthread.php?threadid=26627

[daverme]

Lemme say this about that: I have an Abit BX-133 RAID mobo (High Point HPT370 RAID controller) installed in my wife's PC and we are doing MORE than what you are asking - the boot drive is connected to the primary RAID port and there is a second drive which we use only for backup cloning on the second RAID port. She has a data drive on the first IDE port. So, for that mobo at least, the answer is emphatically YES.

[daverme]

OK, I'll read that thread tonight at home. Meanwhile, the system I'm referring to was actually set up by a dealer (long story there, I have always built my own in the past). However, all they did was to plug the one HDD into the RAID primary. I checked the BIOS later and found that they didn't even set the boot sequence to boot from RAID. I later added a second HDD on the RAID secondary and a third on the IDE primary, all without touching the BIOS; it just worked.

I'll take another look at the BIOS tonight to see if I overlooked something but it almost seems as though Abit built the board to work this way. Remember it's a BX-133 so, without using the RAID ports, the HDD speed would be limited to ATA-33 and the board's main selling points are ATA-100 support plus the reliability of the BX chipset, both of which seem to have bene realized, BTW.

[Bovon]

Ok, `preciate it.

Just wish the durn Intel chips wern't so pricey...I would get a BX133 myself. I have used AMD processors for a long time and am quite satisfied with them...its this VIA KT133 chipset that is screwing everything up. I have an older version of the VIA inuse now...the FIC VA 503+ mobo with the K6-2 500 AMD, and when this chipset first came out, there were a lot of problems to be worked out. You would think that they would have matured by now. Sometimes I get the feeling there is something going on in the background between Intel and AMD. The new onboard Raid for these VIA chipset boards is not compatible it seems like. The Highpoint raid appears to be the better choice, and Iwill appears to be the better KT133 board...so, maybe the answer (for me at least) would be to go with the KK266 and get a Highpoint controller card to get the extra IDE controllers. Personally, I don't want to do that... I think maybe somebody will have something to suggest that may be the answer. The techs at Iwill say it will work, but didn't offer any setup procedures...so, they may not know, or realize there is a problem with their version of Raid...which I think is AMI...

This thread was created for input from anybody that may have some thoughts. So, I appreciate yours..but, what you have setup on your wifes machine may be unachieveable on the KT133 boards.

Thanks again

[AuraEdge]

Im sorry, I think I missed the question...
If the question is "can you run IDE devices on the RAID ports?" then yes you can, but it is only suggested you use Harddrives or other mass storage devices on it (no CDrom or other ATAPI stuff). Dont know if thats true, but with all those connecters, it never was a problem.
The RAID ports is just a IDE controller capable or RAID integrated into the board.
I used to have 2 harddrives, one as a master on each of the RAID channels. No RAID applied.

[Bovon]

Hello AuraEdge and thanx for the opinion here. I am just kinda batting this around some here at this forum. If you could go look at the other forum (link in my first post) you will see what all has been tried. It is possible to use the raid headers as ide, we have determined that. The problem is with the Iwill board it seems. Iwill uses AMI raid I think...(I gotta go look that up and stop guessing). Abit uses Highpoint, and that is where I got the idea to do this. I don't actually need or want raid, but I could use the extra ide headers. Some of the Abit guys have it configured and working fine, but for some reason, the Iwill dosn't want to do it without errors. I am leaning towards buying the Iwill board, and if the raid can be setup as ide I will do that, if the raid on the Iwill won't setup for ide...and I still get an Iwill board, I'll get the non raid instead.

[daverme]

OK, here's the deal: The High Point RAID controller (HPT370) has an option that let's you specify the boot device. I re-booted and set the boot device to HDD-0. The Award BIOS also has an option that lets you say you want to boot from the RAID controller, "ATA100" in the boot sequence. On the re-boot after setting the boot device in the HPT controller, I went into the Award BIOS and set the boot sequence to say "ATA100" after the floppy. I then let the machine boot to Windows and all went well. The BIOS still did not see any drives on the IDE ports but Windows was fine anyway.

I then re-booted a third time and went into the Award BIOS and set all IDE devices to auto-detect. On the subsequent re-boot, both the HPT controller and the Award BIOS showed all the devices connected to their respective ports and Windows came up OK, and it also sees everything.

My read on Jason's comments is that the jumper on the Iwill board disables the entire RAID controller so you want to leave it off. Qcheck's comments suggest that the rest goes like I just described.

LOL, Dave

[Bovon]

daverme,...Man, I do appreciate your contribution. Sence I do not have the Iwill board, (or any other) I cannot follow thru your setup notes and theory. I do believe the HPT controller may be better, or have the better options to do this, than the AMI controller found on the Iwill board. There have been a couple or responces yesterday and today to the thread at amdmb forum that indicate a sucessful end to this, but so far, no one has done this as a utility to keep. I am going to post a link at the amdmb forum to here and see if that gets some input. I like what I read about the Iwill board, and very few have had any problems, other than user. I really do not want to buy the KK266 and then add a HPT controller card...but, if HPT has the clear cut option to do what I want, and AMI does not configure properly, that is what I will do. From the input here and at amdmb, it appears tha HPT is the better raid-type controller.

Thanks again my friend...

[daverme]

No sweat, Bovon. I don't always have solid info to contribute and it's a pleasure when I can.

[daverme]

Bovon, I read the first few posts in the thread you referenced and I think Qcheck has it right ... just plug 'em in! You probably have to go into the std BIOS and tell it to boot from the RAID controller and you may have to config the RAID controller to tell it what device to boot from, but that should be all you need. I think the reason that Jason seemed to have no help is because he can't understand why anyone is confused; it's dirt simple!

I've checked my wife's BIOS and read the manual and, quite frankly, we have it set up wrong ... but it still works! I guess the High Point controller and Award BIOS are pretty forgiving. I will tell you what we have now and later I will try to set it up correctly and let you know the outcome.

The std BIOS activates first then the HPT BIOS activates. In the std BIOS, we have ALL four IDE devices set to NONE. Yep, NONE. In the HPT BIOS, we did absolutely nothing, which means the HPT controller is active but no RAID is defined. We connected a HDD to each HPT header and we have one HDD and two CD-ROMs on the IDE headers. When the machine boots, it says there are no IDE drives but the HPT BIOS shows the two on it's ports. Somehow, Windows knows about ALL of them.

Now, the way it SHOULD be config'ed is as I described in the first paragraph. Also, the Abit doc DOES say the RAID headers are to be used for HDD's only.

[Nocell]

Hello Bov, I have just post a reply on the amdmb but the server seems to return an error. So dun no if that post works. Well just for your info if that post doesn't work, I have actually tried the setup dave suggested and well I just have to add as the Qcheck said, for the Iwill board just leave the jumper to enable, plug the HDD to AMI0 and plug ur CDROM/CDwriter/DVDROM on the IDE0 and IDE1. Then power up. You'll be fine. ALthough the main BIOS will never detect your HDD but the AMI BIOS will and that's for sure (at least for my case IBM 30.7 GXP detected as 29.8 Gb UDMA5).
FOr the boot device in the BIOS, set to watever u want, the KK266R will be able to boot from ur HDD (if of coz it's bootable)
So the point is if you really like the KK266R and afraid if you can use the RAID port as IDE port (for HDD ONLY), then you will have no worry because it will definitely work. Only problem is to verify whether connecting HDD to AMI will have the same or not better performance as if connected to IDE port. And the problem about windows registry error on win98Se/winME. I dun find that on win2k. If the win registry error is not because HDD connected to AMI0 then 'GO FOR THE BEST..... KK266R'
Just for ur info about HDD performance, I partitioned my HDD into 3 drives, C http://www.sysopt.com/forum/frown.gif17Gb/winME) D http://www.sysopt.com/forum/frown.gif8 Gb/win2k) E http://www.sysopt.com/forum/frown.gif5GB)
And using Sandra Professional 2000 Pro, I get:
Under WinMe:
C: => 21239
D: => 17166
E: => 15152
(Win2k obtained 'slightly' higher result)
3dMark2000 : 5531 (winMe)
3dMark2000 : 6355 (win2kSp1)
Hope this is useful for you Bov, Thank you for your help on trying to resolve that 'No arrays defined' message during bootup from AMI BIOS.

[RGone]

Hey, Bovon, are you still checking this thread. I have been trying "still" to determine what will work for a hard drive on the AMI raid controller as set as master(hdd). I am having registry corruption in WiN 98 and don't see this as a good thing. I know that the AMI controller is slower than HPT or Promise but if I could get around the Registry corruption, I could get by.

I must have the plague as none will answer on the amdmb Iwill forum. Maybe you will see his over here and be able to give answer. Thanks RGone...

[Bovon]

Hello, RGone,

Yep...I still check this thread if it makes it back to the front again...no use trying to keep up with it if it gets back on page 20 somewhere.

I can't understand no one not answering your posts over at amdmb forum. Those folks usually jump right on a new post, if they have any thoughts. The trouble with that problem is, nobody has a clue how to fix it. Its not a configuration thing it seems...so, I suppose its up to Iwill to figure out whats going on, and get a fix out. Hopefully, it won't take a mobo upgrade (as in a new southbridge).

There is still a problem with the chipset on all of these KT133a boards. Seems like it may be confined to the 686B southbridge, but I'm not that much aquanted with the problem myself. I have yet to order a board. I will order the KK266, and leave the raid version alone for now...maybe the next upgrade in a few years.

From what I read, it appears the corruption comes from the chipset and/or the southbridge. Iwill and others are trying to find a fix thru a bios upgrade...but, some ppl believe the 686B southbridge may be the problem, and needs to be re designed.
I read at all of the forums, Abit, Asus, Iwill, Azza...the lot...and I see the same problem everywhere to some degree. Even the SB Live! sound system problems are not limited to the KT133a boards...I read where one guy at least has the same problem with an Intel setup on a BX board.

Maybe the cpu's are out running the technology of the mainboards?...easy to do.

As far as the IDE hard drives on the raid channels, it appears that the boards with the HPT controllers are having better success at it than Iwill and its AMI controller. So, I'll just get the KK266 (I have to have at least one ISA socket, so that limits my choices of mobos).

Hopefully, there will soon be a bios upgrade that will fix the corruption problems.

Hello to you too, bigpoppasmooth...glad you stopped by. I'll read your other post next.

[RGone]

BOVON: They be posting some to the Iwill forum right now about the AMI for EIDE and one thing that one poster said was that his prob seemed to crop up when he win with the 4.29via's and Ace-A-Rue mentioned that he had one of the "few" kk266r's that would use the AMI as EIDE and only then at NON-aggressive mem timings etc.

So I thought on he fact that I had only been using the 4.29's since they worked so well on the 8kta+ that I just set up for someone else.

For the 26th time foramtted and reloaded win98se with Via4.28's this time no Via Bussmaster. Had one reg error called in SFC and rplaced setupx.dll and all was smooth "UNTIL" I ran the defragger from Norton called "SpeedDisk" that sucker flies and after running it, the thing was hosed. No boot no chance to recover nothing. Pity that it was @135 FSB and O'cd 350mhz. I like "speeddisk" as it is SO fast and can put the swap file at the edge of the platter. But i do believe its' speed is what caused the slow AMI to hose windows. Got email in to swap board for some different route. I need to use 5 EIDE devices and planned to go raid eventually.

Going to give it a shot without using "speeddisk" and see if I can make it last. Will just leave the swap file move alone. Speed is what it looks like the AMI can't handle. RGone...

[Bovon]

Yeah...may very well be speed. I ran across the following 'test', if you can call it that today..its interesting, but does not have any answers. I know Don, and correspond with him by email a lot (Ace-A-Rue)..

catch you later...gotta hit the sack...
http://www.realworldtech.com/page.cfm?ArticleID=RWT042201222938

[bigpoppasmooth]

I have been able to use the RAID controllers as IDE on the ABIT KT7A-RAID board but that has given me other problems.
If you want to read the thread. http://www.sysopt.com/forum/Forum5/HTML/006240.html

[RGone]

Yeah, I have read his test/reports before and Ilike his style. I am still waiting on reply to email about board swap. It is very disheartening to hear that a few "can" but most including self can't when you have spent to buy. I don't from what I can see on other sites slated more at the kk266r think that I can make it work. And when I posted what I did yesterday about the other posts and what I had managed to see happen myself, they were appearing to have "trouble" with getting an add-in ABIT controller card to work and it uses HPT370 which "has" worked. It is the pits. I have always wondered how much trouble crossing the magic "1000 mhz barrier" was going to cause with long traces. RGone...

[Kuasimodem]

I'm running the Abit KT7-RAID with 2 Western Digital 30Gb ATA100 drives. I played with RAID 0 for a while (SANDRA benchmark 4127), but got bored with that.

Right now I have mine set up as a dual-boot system, with Win98se on the first drive, and Windows ME on the second. To select which OS to boot up, I just hit "control H" when the prompt comes up, and then select which drive to boot from.

I love my Abit RAID board, and when I build my new DDR system, it will have RAID!

[Bovon]

Kuasimodem, sure glad to hear that you have had such great success. Not everybody does however, especially with the AMI raid controller on the Iwill board. It does seem like the Abit users with the HPT controllers are a little more successfull, but even those ppl arn't successfull every time. I have done much reading and study, asking questions everywhere. It seems to me that, about 3 in 10 are able to achieve running hard drives in an IDE configuration on the raid headers. So, till the mobo manufacturers can whip this puppy, I believe I'll just go with a non raid board, and if needed...get a controller card. This really sucks. Here are several good boards, and none of them are 100% certain to run HDDs in IDE mode from raid controllers, yet you are almost assured to be able to do that with a HTP (for instance) controller card. Whats up with that?...why will a controller card work every (?) time, but the same chip integrated onto a motherboard, and it goes to hell in a hand basket.

[Bovon]

Yep...I follow you right along there. I have 'played' with RF for over 40 years in Ham Radio, so I am well familiar with RF cross talk. The higher the frequency, the shorter the wavelength. This is one reason, I can't understand how a mainboard actually works, with such frequencies as 1 ghz...and above. Dam...we are almost talking radar frequencies. But...does the traces on a mainboard actually carry RF?...or is that confined to the processor primarially.

Crosstalk dosn't have to be associated with wavelength completely, for it to cause a problem.

I was just over at the Hard forum, and some guy has posted a new? approach to one of the problems...a fix?...I doubt it, but maybe worth a try anyway.

http://hardforum.com/showthread.php?threadid=133897

[Kuasimodem]

There are many possible variables that can affect systems. Some chips may have originally designed for addon RAID controller cards, and create problems when integrated into a motherboard.

Also, we must look at the manufacturing tollerances that various board producers have. There can be differences in size and length of "runs" on the motherboard, different materials used for pins (brass, gold plated, tin plated...), and variances in the quality of the interconnect cables that different board makers provide.

You also have to consider what motherboard components are near that controller chip. Unless each run or wire inside the computer has it's own shield (like coaxial cable), there is bound to be some "crosstalk" or exchange of signals, it's just a fact of electronics (ever notice that the tuner portion of any RF receiver is in a shielded box?).

An ATA66/100 IDE cable has 80 wires, but only 40 pins, every other wire in that cable is a ground. The reason is, that grounded wire between the signal wires acts as a shield to prevent crosstalk.

If you stop to think about it, our computers are already operating in the radio frequency ranges (900Mhz processor, 200Mhz FSB, 33Mhz PCI bus,133Mhz memory bus and 66Mhz AGP bus). Each component will emit a minor amount of RF radiation (no, not nuclear, LOL) that could, in theory, interfere with other components.

It's a wonder that these things work at all!!

[kandela]

follow the link below to see possible answers to your questions.
http://www.fullon3d.com/kk266faq/kk266faq.shtml#onedriveraid

Hope it helps.