Any owners of PCChips M811LU V3.1 or Syntax SV266AD?

tcsenter

Lifer
Sep 7, 2001
18,893
544
126
Calling anyone and everyone who may have one or more systems built using either the PC Chips M811LU V3.1 or Syntax SV266AD.

I've used several of these boards for budget systems, but a few days ago I discovered a potential bug, flaw, defect, or whatever you want to call it, affecting both the PC Chips branded and Syntax branded motherboards (they are the same board).

The VIA VT8235 South Bridge supports up to ATA/133 (Ultra DMA Mode 6). In every one of these systems, the IDE/ATA transfer mode is limited to Ultra DMA Mode 4 (ATA/66) when an ATA/100 or ATA/133 hard drive is attached to the system. When an ATAPI optical drive supporting ATA/33 or higher is attached, whether CD or DVD drive, the IDE/ATA transfer mode is limited to Ultra DMA Mode 1 (ATA/25). The transfer mode is reported in Device Properties through Device Manager > IDE ATA/ATAPI Controllers > Primary or Secondary IDE Channel > Advanced Settings > Current Transfer Mode.

This is with Windows XP Home, either with SP1 or SP2 integrated, using the latest BIOS from PC Chips and Syntax, and all the latest device drivers including the latest Hyperion. The hard drive and optical drives are on different channels, each being the master with no other drive as slave.

I have tested multiple hard drives; Seagate and WDC 80GB ATA/100, WDC 120GB ATA/100, Maxtor 80GB ATA/133. I have tested different optical drives; Optorite 52x CDRW, BTC 8x DVD writer, and BenQ CDRW + DVD-ROM combo drive. I have tested different ATA/100 compliant 80-conductor IDE cables, all of which are not longer than 18". I have tried shielded round ATA/100 IDE cables. All hardware including the cables work A-OK with other motherboards at the highest supported modes.

All drives are reporting the correct mode and capacity @ POST. PIO and Ultra DMA is set to "Auto" for all drives in BIOS. HDTach 3 tests confirm the HDDs are operating in ATA/66 instead of ATA/100 or ATA/133. I've even swapped-in a couple brand new ATX12V power supplies with output specs far exceeding total system demands, just to leave no stone unturned.

Installing the latest VIA IDE BM Accelerator Drivers results in all drives operating in the correct modes, but using this driver I experience numerous "Delayed Write Fail" errors and "error detected on device \Device\Harddisk0 during a paging operation" related to the hard drive (in all systems with all hard drives). What's more, effective hard disk performance actually DECREASES when the drives are operating at the higher modes, which is consistent with the numerous IDE and disk errors I'm getting.

If you have one of these boards, I would appreciate if you could take a peak at what transfer mode the system is reporting through Device Manager > IDE ATA/ATAPI Controllers > Primary or Secondary IDE Channel > Advanced Settings > Current Transfer Mode.

I have opened a support ticket with PC Chips and Syntax detailing this but have not received a reply. Its only been two days, I'll give them until Wednesday or Thursday before I start making angry phone calls. :p
 

Peter

Elite Member
Oct 15, 1999
9,640
1
0
The thing with ATAPI drives as slaves is that very few of them support UDMA modes past 1 or 2, and many many hard disk firmwares limit themselves to the same UDMA mode when they detect one as slave. I'm a BIOS writer, I've had the complaint ;)

The cure is not to do this.

Round cables are the pure evil, btw. Ribbon cables have significantly less crosstalk between signals - and this is what kills UDMA-100 typically. Windows gradually steps back mode by mode whenever it sees transfer problems, so that might explain why you see that degradation. Your "delayed write fail" problems with the alternate driver point in the same direction.
 

tcsenter

Lifer
Sep 7, 2001
18,893
544
126
"The hard drive and optical drives are on different channels, each being the master with no other drive as slave."

"I have tested different ATA/100 compliant 80-conductor IDE cables, all of which are not longer than 18. I have tried shielded round ATA/100 IDE cables. All hardware including the cables work A-OK with other motherboards at the highest supported modes."

Two days and over 12 hours of eliminating every conceivable source. Its down to the BIOS, a design flaw of the motherboard, or perhaps a batch of bad South Bridges, and I'm leaning away from the BIOS.

I found a discussion on VIA Arena Forums involving the exact same problem with Shuttle AK32VN (3.1), which is also the same board. The latest PC Chips, Syntax, and Shuttle BIOS are all different builds. Given the maturity of the hardware and the cumulative number of hardware revisions this board represents, I expect a fairly high level of refinement.
 

Peter

Elite Member
Oct 15, 1999
9,640
1
0
True, there are so many revisions and other-brand badge jobs of this board out there - PC-Chips, ECS, Shuttle, Syntax, Matsonic, with chipsets from KT266A to KT400A - that I'd second your result that it shouldn't be a BIOS thing ... and most probably, it isn't.

Knowing BIOSes from the inside, I know that you'd never end up with an odd DMA mode like UDMA-1 by BIOS decision. That's 99.9 percent Windows having stepped down from UDMA-2 due to transfer errors. Those transfer errors you observe with the alternate driver - and they're coming from _somewhere_ on the IDE cable or the IDE traces on the mainboard, from the plug to the southbridge. Else they'd be PCI-ish bus errors on V-Link, not IDE transfer errors.
 

tcsenter

Lifer
Sep 7, 2001
18,893
544
126
There have been a couple driver glitches with XP SP1 and VIA South Bridges. There was a flaw in viaide.sys bundled with SP1 that caused the VIA 686B South Bridge to falsely be detected as supporting UDMA-6.

And there is this KB Article, where the problem seems to be the opposite; UDMA-6 support isn't enabled on chipsets that support it:
SYMPTOMS
After you install Windows XP Service Pack 1 (SP1), IDE ATA-133 (Ultra DMA Mode 6) devices on motherboards that use VIA Technologies chipsets are not enabled. This occurs although Windows XP SP1 supports Ultra DMA Mode 6.

CAUSE
When you install Windows XP SP1, devices on your computer are not reinstalled in the process. The Ultra DMA Mode 6 devices were not supported before you installed Windows XP SP1.
I've done clean installs of XP Home with SP1 integrated, and also with SP2 integrated. Just for kicks, I installed XP with SP1 integrated then patched to SP2. Before or after SP2 and the VIA Hyperion, nothing changes.

There could be yet another flaw in the driver causing this problem, but with the VIA IDE Accelerator Driver resulting in Delayed Write Fail and Paging Operation Errors when the system is operating in UDMA-5, I tend to agree there is an unstable condition relating to the hardware causing the driver to downgrade to UDMA-4.

I would be more forgiving if there were an unstable condition at the highest supported mode of UDMA-6 (since the hard drives I want to use in these systems are UDMA-5), but down-grading two entire modes? Someone screwed up and selected resistors or capacitors with the wrong values or something.
 

Peter

Elite Member
Oct 15, 1999
9,640
1
0
You'd be surprised to see that there typically is exactly NOTHING inbetween the southbridge chip and the IDE connectors, just traces. That's all integrated into the chip, pretty hard to fsck up in board layout.

 

JHawk56

Junior Member
May 21, 2005
4
0
0
Tim, I just set up an M811LU with a clean install of WinXP Pro, no service packs as yet, using the XP native VIA IDE driver. WDC 80 GB ATA-100 drive on primary is reported as UDMA Mode 4 by XP. CD-ROM on secondary is reported as Multi-Word DMA Mode 2.

John
 

tcsenter

Lifer
Sep 7, 2001
18,893
544
126
Originally posted by: JHawk56
Tim, I just set up an M811LU with a clean install of WinXP Pro, no service packs as yet, using the XP native VIA IDE driver. WDC 80 GB ATA-100 drive on primary is reported as UDMA Mode 4 by XP. CD-ROM on secondary is reported as Multi-Word DMA Mode 2.
Hey, I didn't see your reply until now. Thanks for responding! Most likely your CD drive only supports up to MWDMA-2.

I've almost completed the entire collection. I purchased two Shuttle AK32VN boards and tested those using the latest Shuttle BIOS - exact same problem without exception.

I also purchased a Syntax SV400 based on KT400/VT8235. All drives are operating in the correct mode using the SV400 with the same IDE/ATA cabling (in addition to other motherboards I've tested this hardware on).

Still haven't received a reply from PC Chips, I guess its time for attempt #2. Maybe my supplier will agree to swap them for the SV400.

Grrr...
You'd be surprised to see that there typically is exactly NOTHING inbetween the southbridge chip and the IDE connectors, just traces. That's all integrated into the chip, pretty hard to fsck up in board layout.
Actually, there are four SMC resistors along the traces between the IDE connectors and the VT8235. I can't be certain those are IDE lines, but I presume so since they are routed directly under the connector and not around it.
 

JHawk56

Junior Member
May 21, 2005
4
0
0
OK, as you know you can complete the first set by getting an ECS K7VTA2 Pro v3.1.

But then you'll need an ECS L7VTA2 v1.0 and a Kobian/Mercury KOB KT400 FDSX to match your SV400. Could become an expensive hobby!

John
 

amp06

Junior Member
Jun 18, 2005
1
0
0
1st of all, Sorry for bumping this old thread.

I, too have an PC Chips M811LU V3.1 board. I am limited to Ultra DMA Mode 4 as well. Aside from that, the board has been worked without issue since I bought it ~2years ago.