digital SERVER 5000 PSU
Andrew Harvey
I am currently retro-fitting an ATX motherboard into a digital SERVER 5000
case.
I have searched the web and the hp site for information on the power supply
in the case without success. The PSU is labled APL 4030, P/N 30-43120-01.
Does anyone know if the connectors are ATX?
Or, how to use 2 of these power supplies in failover mode?
Or, have the pinouts of the connectors?
Or, where to find documentation for these PSUs?
thanks in advance
Andrew
--
----------------------------------------------------
This mailbox protected from junk email by Matador
from MailFrontier, Inc. http://info.mailfrontier.com
ben_myers_spam_me_not
by or for DEC for its large cube-shaped servers, both Wintel and Alpha.
Unless someone can sneak some old old internal documentation to you, I doubt
that you will find more information about this power supply than is already
buried on the web deep inside the HP web site. The system hardware maintenance
documentation for these servers does not describe pinouts, because the field
replacable unit (FRU in IBM's terms) is the power supply itself. Field
personnel would have no need to know the power supply pinout, only how to remove
and replace it.
I can only discourage you from trying to use this power supply, no matter how
robust and well-designed it is, to hook up to an ATX motherboard. Better to
graft an ATX power supply onto the system chassis and use extension cables to
reach the hard drive bay and other items needing power... Ben Myers
Mirko S Veselinovic
That is the same CASE that powers digital ZX6000 server, and two PSU
in ZX redundant PS sistem have same P/N on them. Unfortunately, these
are not standard ATX! Connector with lot of orange and black wires is
3.43V supply, larger with red and black wires is 5.0V, and smaller
with blue, yellow, white and black wires is for 12V and control (power
good) lines. In our case there is an dual PentiumPro inside that
worked well, until we tried to upgrade UW SCSI (Adaptec 3940)
controller inside to newer U160 (Adaptec 29160). Now we are not able
to boot this machine, as it never initialize video, and after about 30
seconds of checking PSUs it resets! We suspect motherboard is the
problem, but we can not find another (at least not afordable, we ware
offered a $400 for new) to try it out. It is a shame to dump this
monster case, but it seems there is not alternative!
On Tue, 17 Feb 2004 04:27:21 +1100, "Andrew Harvey"
<andrewdotharveyatoptusnetdotcomdotauNOSPAM> wrote:
ben_myers_spam_me_not
Two things:
It is likely that the 29160, which meets the PCI 2.1 spec, is incompatible with
the ZX6000 motherboard, probably PCI 2.0 or even earlier. The 29160 adheres
stringently to the PCI 2.1 spec to produce the required data transfer bandwidth.
What does the system do when you remove the 29160 and re-install the 3940? If
it still boots, you'll have to use it as-is with the 3940. If it does not boot,
I am pretty certain I have a ZX6000 motherboard in working order in my
warehouse. Even with shipping and customs, I'm sure I can get it into your
hands for no more than half of the $400 you have been quoted. Let me know.
... Ben Myers
On Tue, 17 Feb 2004 18:28:14 +0100, Mirko S Veselinovic <mirko...@eunet.yu>
wrote:
Mirko S Veselinovic
I thought PCI versions might be the problem. It seems now it is to
late, as it is not possible to boot neither with nor without PCI
devices. 29160 seems also to be damaged, as it can not recognize any
devices attached to in another PC (it freezes in the process of trying
to). While I might be a litle attached to this beast, it is not worth
investing in it, as it is possible to find second hand complete Dual
PPro sistems for as low as 80 Euros. In our case, we would prefer
something more powerfull.
Cthulhu
> It is likely that the 29160, which meets the PCI 2.1 spec, is incompatible with
> the ZX6000 motherboard, probably PCI 2.0 or even earlier. The 29160 adheres
I'm using an Adaptec 29160N in my DIGITAL Server 500, an elder Pentium
233MMX which I don't think is nothing more than PCI 1.0. It works fine
with a Quantum Atlas 10K on one bus, a Pioneer CD-ROM and an elder-too
Tandberg TDC 4100 on the other. Maybe it can not run as fast as it
could, but it for shure boots - that DEC is my home server. The OS is
Linux, Debian 3.
However I remeber that when I was assemblying it showed similar failure,
that I solved resetting the CMOS memory. I know, that sound illogical,
but I'm not vulcanian. :)
Just my 2eurocent.
testimoniatamente,
Cthulhu
--
Vado dove nessuno giunge
Per strade deserte abitate da spettri http://www.rlyeh.it/
Vado nel Nulla, per cogliere l'Essenza. <cthulhu(at)rlyeh(dot)it>
ben_myers_spam_me_not
I fully understand why you would not want to sink more money into the server.
You ARE better off buying a complete server for 80 Euros.
Maybe the 29160 was not properly seated in its slot when it was installed?
... Ben Myers
On Wed, 18 Feb 2004 21:35:20 +0100, Mirko S Veselinovic <mirko...@eunet.yu>
wrote:
>
>
ben_myers_spam_me_not
cube-box DEC servers, so its PCI bus design is probably more up to date. The
dual-processor DEC Pentium systems from this era needed custom-designed chips to
work. And the DEC Pentium Pro servers may well have had custom chips, too.
These computers were designed very early in the Pentium/Pentium Pro days, before
Intel had achieved a high degree of standardization with its motherboard
chipsets.
Adaptec has always been one of the brand names with designs more tolerant of
older systems, but the 29160 represents a generation of SCSI host adapter beyond
the older 29xx-series and 3940 PCI cards. Adaptec may well have decided not to
support the older buses with the 29160.
Con i migliori saluti,
Ben Myers
Mirko S Veselinovic
I have tried with CMOS reseting, and I have tried with aditional video
card in PCI slot 1 (as sugested by manuals), but no use. I remember
similar problems (no video) with some P1 mobos with PCI video cards,
some of which I was able to resolve by inserting ISA video card, or in
one case even copying EPROM (in EPROM burner) from a working PC with
the same mobo. Unfortunatelly in ZX6000, there is no ISA slot to try,
and I suppose EISA slots are not adequate (I have tried with them
also). As a last chance, I am thinking of reflashing EPROM, but I am
not shure which of the two chips I should work on, and where I should
find "working" contents for them...
On Thu, 19 Feb 2004 00:06:50 +0100, Cthulhu <spa...@rlyeh.it> wrote:
ben_myers_spam_me_not
You can install an ISA card in an EISA slot... Ben Myers
On Fri, 20 Feb 2004 00:59:09 +0100, Mirko S Veselinovic <mirko...@eunet.yu>
wrote:
>
Cthulhu
> I'm fairly sure that the DIGITAL Server 500 is a newer design than the older
> cube-box DEC servers, so its PCI bus design is probably more up to date. The
Sorry, I was thniking that bigger the number, more recent the server!
At least with PDPs and VAXen, it was (amost) that way...
> Con i migliori saluti,
:)
ricambiatamente,
Cthulhu
> the same mobo. Unfortunatelly in ZX6000, there is no ISA slot to try,
> and I suppose EISA slots are not adequate (I have tried with them
> also).
Has Ben said, an ISA card should work in an EISA slot. But we are
talking about computers, not about deterministic events. ;)
> not shure which of the two chips I should work on, and where I should
> find "working" contents for them...
Try a look here:
http://h18000.www1.hp.com/legacysupport/digital/index.html
also, if I have rightly understood what kind of machine you have, look
here:
http://h18000.www1.hp.com/legacysupport/digital/epid62.html
indirizzatamente,
Cthulhu
P.S.: one point to HP for still keeping that old infos.
Mirko S Veselinovic
As I said, I have tried that also, but no use. ZX6000 manuals suggests
only PCI video card replacement in PCI slot 1. I remember having
problems moving EISA cards without first reconfiguring with EISA
configuration utility, which now I am not able to access!
Mirko S Veselinovic
Yes, that is the machine! I am a little bit puzzled with a number of
BIOS versions. On an sticker attached to orginal BIOS, Version 1.04 is
printed, and there seems to be two versions of 1.04 BIOSes! Some
versions of BIOSes are refered to as MP (multi proc.), but as far as I
know, all these machines are MP (max. 4 proccessors), although it is
possible to configure and attach only 1 (one) proccessor. Another
puzzle is two BIOS chips on motherboard. On the other one, a printed
sticker states Version 1.05! I can pay a visit to a friend that has
EPROM burner, but I would not like to be it to many times...
Cthulhu
> Yes, that is the machine! I am a little bit puzzled with a number of
> BIOS versions. On an sticker attached to orginal BIOS, Version 1.04 is
> printed, and there seems to be two versions of 1.04 BIOSes! Some
It seems that the second release fixed some kind of bug. But why bother
with 1.04, there is 3.01!
> versions of BIOSes are refered to as MP (multi proc.), but as far as I
> know, all these machines are MP (max. 4 proccessors), although it is
Oh, I suspect that this is only some misnaming standard. :)
> possible to configure and attach only 1 (one) proccessor. Another
> puzzle is two BIOS chips on motherboard. On the other one, a printed
> sticker states Version 1.05! I can pay a visit to a friend that has
> EPROM burner, but I would not like to be it to many times...
Maybe the BIOS of that machine is splitted on two chips, and when the
motherboard was firm-upgraded they put the new label on the other chip -
exploding one of that .EXE avaliable for download should give some hints
here (I can't do that myself, I'm under Linux and that files are not
autoexpanding ZIP). Maybe one of the two chip is a simple ROM, not an
EPROM, and is used with that funny named "Crisis Recovery Diskette", as
a fail-safe BIOS. Maybe strange things happened inside your machine and
a new kind of self conscious server is growing up. Who knows? :)
misteriosamente,
Cthulhu
--
Vado dove nessuno giunge
Per strade deserte abitate da spettri http://www.rlyeh.it/
Vado nel Nulla, per cogliere l'Essenza. <cthulhu(da)rlyeh(in)it>
Mirko S Veselinovic
>Mirko S Veselinovic <mirko...@eunet.yu> wrote:
>
>It seems that the second release fixed some kind of bug. But why bother
>with 1.04, there is 3.01!
Well, I think if I can recover previous working state (I presume it
worked with BIOS 1.04), than I can try to upgrade to 3.01!
>exploding one of that .EXE avaliable for download should give some hints
>here (I can't do that myself, I'm under Linux and that files are not
>autoexpanding ZIP). Maybe one of the two chip is a simple ROM, not an
>EPROM, and is used with that funny named "Crisis Recovery Diskette", as
>a fail-safe BIOS.
OK! I will inspect those EXEes first thing in the morning...
Thank you for your time.
Cthulhu
> Thank you for your time.
I have aeons to spend. :)
Good luck!
auguratamente,