Profile

unixronin: Galen the technomage, from Babylon 5: Crusade (Default)
Unixronin

December 2012

S M T W T F S
      1
2345678
9101112131415
16171819202122
23242526272829
3031     

Most Popular Tags

Expand Cut Tags

No cut tags
Tuesday, November 17th, 2009 02:56 pm

Despite taking it apart and lubricating every coarse mechanical track, linkage and pivot, I have been completely unable to cure the intermittent (roughly 50% of the time) tray jam on the LiteOn SATA DVD burner [livejournal.com profile] databeast gave me the last time he was up here. There is nothing visibly mechanically wrong, and nothing sticks when cycled by hand; nevertheless, at least half the time, when actually connected to a machine, the drive either fails to open, fails to lock closed, or fails to unlock.

Nevertheless, despite it's mechanical problems, its logic is working fine, and I was able to use it to confirm that the MSI6566E P4 board that [livejournal.com profile] lwj2 sent to us has not the slightest hesitation about attempting to boot from a SATA device connected to the SiI3124-based PCI SATA controller I bought when I first started trying to do a ground-up OS refresh on babylon5. This only deepens the mystery about why babylon5's ASUS A7V333 AthlonXP board will talk to that controller, and SATA devices connected to it, perfectly happily once it's booted, but utterly refuses to detect any SATA devices at boot time. I even tried installing a 16MB CF card on a PATA adapter with grub on it to boot from a SATA disk, and that didn't work, either. I can only conclude that, as the A7V333 motherboard (introduced in 2002) predates SATA (introduces in 2003), the BIOS contains no SATA support whatsoever, and for some reason none of the subsequent BIOS updates (the most recent, version 1017, dates from July 2006) has ever added any SATA support.

There's one post on the ASUS support forums about inability to get an A7V333 to boot from an add-in PCI SATA card, but it's never received any replies. Further research online finds occasional other posts of the same problem, none of which ends in a successful resolution. This appears to suggest that SATA on the A7V333 is a problem to which there is no known solution.

I guess that means I'm squarely behind the 8-ball on converting babylon5 to SATA¹, unless I also replace the motherboard. And therefore the CPU, and the RAM, and probably the graphics card...

Well, it has been seven years since I last did it. So I suppose that's not so bad.

[1] "Why not just use PATA?" Because PATA is brain-damaged, and because the board only has two PATA channels, and I like to run mirrored boot disks on all of my *nix boxen, which means two disks and an optical drive, which means some two of those three devices are going to have to share a PATA channel. And that's when PATA starts to REALLY suck, because bus contention is going to cut throughput on that channel by a factor of at least two. That's one of the major reasons why SATA was invented.

Tags:
Tuesday, November 17th, 2009 09:56 pm (UTC)
If the mechanics work fine when the drive is standing alone, but fail when it's mounted in a chassis, the next thing to check is whether or not mounting the drive changes its shape, causing a binding.
Wednesday, November 18th, 2009 04:07 pm (UTC)
Sounds like a failing motor, or some part of the power feed to the motor. (Then again, I studied EE, Ten year old, Hammer....)

A SATA DVD burner goes for about $35 at MicroCenter. Right now that is a show stopper for me, but it really isn't much to replace it.

I am having intermittent success with SATA drives on older motherboards. (I only have older motherboards.) I have some PATA<>SATA converters that work reasonably well, but they still have the throughput issues you mention.

I am rapidly coming to the conclusion that using SATA does require a new motherboard and the rest. Meaning I am toast for upgrades for a while. That is a problem for me since zeus is having problems. I would rather not do a reinstall on a dual PII 400MHz system, it just takes too long. I recently acquired a 4U ATX case, I was hoping something existing could go in there. 1TB drives only come in SATA. I also like to run dual DVD's. (I won't even mention the change to PCIe...)

I am trying to migrate my main system to a SATA drive from the two PATA drives in currently runs. The progress is sporadic. I do not yet have a reliable boot on that system (939 Motherboard, it is only SATA 1.0, but it should work better than it is.)
Saturday, November 21st, 2009 03:14 pm (UTC)
Odd - what third-party SATA controller are you using? The SATA controller card should have its own built-in BIOS that will let you boot from it, regardless of BIOS support in the motherboard. It may have its own secondary setup program with hoops you need to jump through, and you may also need to fiddle settings in the motherboard BIOS to get it to boot the add-in BIOS first.

All that said, I freely admit I could be talking through my hat - my last serious BIOS development work took place about the time USB started becoming ubiquitous, and PCI was just starting to gain traction. I have only the vaguest idea how all the moving parts fit together.
Saturday, November 21st, 2009 05:45 pm (UTC)
Well, I have two. One's a 3Ware that doesn't seem to work any more. The other's some no-name-brand SiI3124-based card. I might have better luck with, say, an LSI Logic card, but I don't have one on hand to try right now. I see that PCI LSI MegaRAID 150s can be had for about $75 refurbed, though that's only a 2-port controller, and there's an LSI 523B2 6-port SATA controller (battery-backed, even) on eBay for $60 right now. That's PCI64, though, and I don't know whether it'll step down properly to PCI32. It's probably worth a shot, but I'm going to see if I can confirm bus compatibility first.