Hi Tony!
On Mon, 2023-06-26 15:45:14 +0100, Tony Duell <[email protected]> wrote:
> On Mon, Jun 26, 2023 at 3:38 PM Jan-Benedict Glaw via cctalk
> <[email protected]> wrote:
> > As I'm preparing to setup my old hardware, I fetched two VAXstations
> > (4000/90 and /96) from storage and cleaned one of them throughoutly.
> >
> > Then I gave power (to both of them), but both won't really start:
> > all 8 diag LEDs are on (--> power available but CPU didn't start
> > executing instructions.)
> >
> > I took the PSU (from the cleaned /90), a DEC H7819-AA, and measured
> > it. Unfortunately I didn't find pinouts or schematics at a first
> > search. The plate states that there should be 3.3V, 5V, 12V, -12V and
> > -9V. I found most of that:
> >
> >
> > +------------------------------------------------------------------+
> > | o |
> > | DEC H7819-AA PSU 10 +---+ 1 |
> > | (view at the bottom side) | | |
> > | | | |
> > | | | |
> > | 18 +---+ 9 |
> > | o |
> > | +-------+ +-------+ |
> > | | Fan | | Fan | |
> > +----------------+-------+----+-------+----------------------------+
> >
> > With above pin numbering, this is what I could find / measure / deduce:
> >
> > 3V3 brown 10 1 blue 12V
> > 3V3 brown 11 2 black GND
> > GND black 12 3 red 5V
> > GND black 13 4 red 5V
> > GND black 14 5 black GND
> > 5V red 15 6 black GND
> > 5V red 16 7 white -12V
> > 5V red 17 8 blue (0.78V)
> > (4.91V) lilac 18 9 brown (-1.65V)
> >
> > Most values look plausible, except those three in parentheses. At
> > least one of them should probably be -9V wrt. GND I guess, but that's
> > totally absent. And what's the other two? (If I got the colors wrong:
> > Please forgive, I'm red-green blind.) That could be some "power-okay"
> > indicator, or external switch-off?
>
> I don't know this machine at all (too modern :-)) but that -9V sounds
> at thought it might be for an internal ethernet transceiver. Now an
> ethernet transceiver is supposed to be isolated from ground and one
> way of helping with that is to have a totally isolated 9V output on
> the power supply with its own 'ground', not connected to the ground
> for the rest of the supplies.
>
> Using a high impedence voltmeter like most DMMs, you'd see random
> noise voltages on those 2 lines wrt the normal ground.
>
> What voltage do you measure between the suspect blue and brown wires?
> Connect your meter between them, not to the black ground and something
> else.
Did that any you're totally right here: Using pin 8 (blue) as GND, pin
9 (brown) becomes -9V. Just the lila wire remaining as a mystery.
HOWEVER! A Polish friend suggested that even with all LEDs lit (which
is as severe as it gets), it might just be a dead RTC chip. I dremel'd
it open and soldered some wires, added a battery (not yet with a
proper battery clip) and ... it started up.
So I learned something about the PSU (--> most of its external
pinout), the usage of -9V for ethernet (would never guessed that!) and
that the stupid battery-backed RAM may completely kill a machine.
Just added a PiSCSI, uploaded a NetBSD ISO image, configured a 2 GB
HDD and fired up again:
>>> sh conf
KA49-A V1.0-006-V4.0
08-00-2B-35-5D-DF
128MB
DEVNBR DEVNAM INFO
------ -------- --------------------------
1 NVR ?? 001 0016
3 DZ OK
4 CACHE OK
5 MEM OK
128MB 0A,0B,0C,0D=16MB, 1E,1F,1G,1H=16MB
6 FPU OK
7 IT OK
8 SYS OK
9 NI OK
10 SCSI OK
1-RZ28M 3-SCSI 6-INITR
11 AUD OK
>>> sh dev
VMS/VMB ADDR DEVTYPE NUMBYTES RM/FX WP DEVNAM REV
------- ---- ------- -------- ----- -- ------ ---
EZA0 08-00-2B-35-5D-DF
DKA100 A/1/0 DISK 2.10GB FX RZ28M 0568
DKA300 A/3/0 RODISK 370.80MB RM WP SCSI 2304
..HostID.. A/6 INITR
Thanks for the help!
MfG, JBG
--
