I suddenly realized that this controller, of which I have two such cards and
one actually installed in an Intel 915GAVL motherboard, are 64 bit cards and as
best I know without tearing boxes up to look, the PCI slots in these 915GAVL
motherboards are 32 bit 5 volt slots!
The one such box which is working with MCP2 latest everything including the
'final' official release of the Adaptec 29160 drivers, which *ARE* a later date
and different from the driver citations in the eCs report page, has been
working just fine for months now with a Seagate LVM 32 bit hard drive. I have
had no trouble with it at all including compete tape backup work with BA2K
Server Pro to my SCSI DAT3 tape drive even for 160,000 plus files and restore
as needed with the 29160 card.
Duhh .. how did I do this? Without traveling to check this box, is this
because even though I didn't realize it, the 64 bit plug-in pin configuration
for the 29160 Adaptec controller can be inserted into a 32 bit slot, which is
for sure running at 5 volt levels? And just because the 64 bit plug-in on the
29160 controller card has a slot at the right place, I could and did just plug
it into the 5 volt PCI slot?
This particular box has one other PCI slot card in it, an SBLIve 5.1 Model 220
card, which is also definitely a 5 volt card with the slot toward the back of
the PCI plug-in pin assembly.
The reason I am asking for thoughts is that there are another two SMB Single
Board Computer cards headed here for research with Intel P4 CPU, Intel 915 or
945 video level stuff and the companion 14 slot passive backplane for them
definitely has only what look like 32 bit PCI slots with the sole 5 volt rear
slot PCI sockets. I originally asked for a passive bckplane which could
handle 3.2 volt PCI or 5 volt PCI stuff, but got told none of the possible SMB
computer cards were compatible with it. Just the 32 bit 5 volt stuff.
Thoughts? Thank you!
--
--> Sleep well; OS2's still awake! ;)
Mike Luther
> The one such box which is working with MCP2 latest everything including the
> 'final' official release of the Adaptec 29160 drivers, which *ARE* a later date
> and different from the driver citations in the eCs report page, has been
> working just fine for months now with a Seagate LVM 32 bit hard drive. I have
> had no trouble with it at all including compete tape backup work with BA2K
> Server Pro to my SCSI DAT3 tape drive even for 160,000 plus files and restore
> as needed with the 29160 card.
>
> Duhh .. how did I do this? Without traveling to check this box, is this
> because even though I didn't realize it, the 64 bit plug-in pin configuration
> for the 29160 Adaptec controller can be inserted into a 32 bit slot, which is
> for sure running at 5 volt levels? And just because the 64 bit plug-in on the
> 29160 controller card has a slot at the right place, I could and did just plug
> it into the 5 volt PCI slot?
Just so happens I was messing with a 29160LP the other day and noticed
the same. I checked the manul on line as it is a s/h card I got
somewhere and it says it will work in either slots - Don't recall
anything about voltages though. Google should turn up the manual.
HTH
--
Regards
Dave Saville
Dave Saville wrote:
> Just so happens I was messing with a 29160LP the other day and noticed
> the same. I checked the manul on line as it is a s/h card I got
> somewhere and it says it will work in either slots - Don't recall
> anything about voltages though. Google should turn up the manual.
>
> HTH
>
Aha! Thanks Dave. I don't know about the 29160lp version, but to try to help
whomever here, these are the drivers which I am using at present:
10-17-01 10:56a 6822 0 7870PRES.EXE
10-17-01 10:55a 6822 0 78U2PRES.EXE
4-10-02 5:19p 154140 0 aic7870.add
2-22-00 3:16p 666 0 aic7870.ddp
4-10-02 5:19p 58573 0 AIC7870.map
4-10-02 5:19p 19812 0 AIC7870.sym
10-17-01 10:55a 156412 0 aic78u2.add
2-22-00 3:16p 665 0 aic78u2.ddp
10-17-01 10:55a 58445 0 AIC78U2.map
10-17-01 10:55a 19844 0 AIC78U2.sym
10-17-01 10:55a 169548 0 aicu160.add
2-22-00 3:16p 663 0 aicu160.ddp
10-17-01 10:55a 60257 0 AICU160.map
10-17-01 10:55a 20532 0 AICU160.sym
5-01-01 8:26a 79208 0 Reados2.htm
5-01-01 8:27a 20244 0 Reados2.txt
10-17-01 10:55a 6822 0 U160PRES.EXE
They came out of the archive:
9-28-02 7:08p 451388 0 7800FAM.EXE
I also have a file which seems to be an earlier version of the driver which I
think was released by Adaptec itself. If my memory is correct, it may be the
one which produces the driver package for the 29160 which was the focus of the
bad comments in the eCs forum and is different that the files above which came
from the IBM official download:
12-29-00 6:21a 122789 0 u160_os2_drvs_v100.exe
I think I would check your device driver version for your work but that's just
a suggestion. All I'm trying to do is help like you did very much help me!
> the PCI slots in these 915GAVL
> motherboards are 32 bit 5 volt slots!
I'm pretty sure that standard PCI slots have been dual 3.3/5V for a long time.
--
Trevor Hemsley, Brighton, UK
Trevor dot Hemsley at ntlworld dot com
Trevor Hemsley wrote:
> On Thu, 20 Aug 2009 22:03:56 UTC in comp.os.os2.setup.storage, Mike Luther
> <mike....@ziplog.com> wrote:
>
>> the PCI slots in these 915GAVL
>> motherboards are 32 bit 5 volt slots!
>
> I'm pretty sure that standard PCI slots have been dual 3.3/5V for a long time.
>
That is what I thought too. But research in the Wikipedia on this seems to
indicate that there is actually a physical requirement for the pin-plug
configuration which 'controls' this! From what I read, if the gap in the PCI
plug-in wedge is at the rear of the system from the back of the motherboard,
this designates only a 5 volt operation. If the gap in the PCI plug-in wedge
is at the edge closest to the back of the motherboard, it is a 3.2 volt
operation. Only if BOTH the 'front' and the 'rear' slots are in the wedge
plug-in, is this a dual voltage capable operation.
Further, the 29160 cards I have here, which are the long high speed card that
does work in the high speed motherboard PCI slots, do have the dual slot in the
32 bit pin-part for this controller card.
Which all figured into my initial confusion over wanting a passive backplane
for the needed upgrade on mission critical and EMP pulse certifiable relay rack
systems for the future of us and OS/2 with even fiber optic network cards of
chips in what might be coming.
Personally, I have never even had in my hands a motherboard or a passive
backplane which has the 64 bit long slot socket! Nor have I even seen such a
passive backplane for SMB industrial CPU card use shown in the pictures on the
web here.
And was also stunned when even the most 'modern' latest version(s) of the Cyber
Research SMB backplanes didn't have this and even the so-called 5V and 3.2V
passive backplane wasn't even said to be usable with their latest and greatest
CPU SMB that has no End Of Life production and CPU availability that we are
trying to get to work OS/2 all the way back to ISA and forward for at least
100,000 hours into the future.
Thanks!
If a card has two gaps in the connector it will fit into any PCI slot.
If a card has only one gap it requires a 3,3V or a 5V slot depending on
the location.
> Further, the 29160 cards I have here, which are the long high speed card
> that does work in the high speed motherboard PCI slots, do have the dual
> slot in the 32 bit pin-part for this controller card.
=> 5V and 3,3V compatible.
Marcel