How to switch an Alpha Personal Workstation 500 stuck in AlphaBIOS back to SRM console,no VGA output

[Booting an Alpha Person]

I need some advice from anyone in the community on getting an Alpha Personal Workstation 500a that had run OpenVMS 7.3 (only, no WinNT, no UNIX) booted up again after about a year and a half of storage.

It seems to be stuck in AlphaBIOS, not the SRM console, and we've lost the VGA output in the midst of our troubleshooting efforts. We observed that the system lost the date and time, and so slid out the main system board to replace the battery, but after sliding back in, no VGA output comes through the adapter any longer, and we don't seem to get anything on COM1 or COM2.

So we need to know a few things in order to try to get this box going again-

-Is it possible to access the system using the COM1 or COM2 ports while it's set for AlphaBIOS?

-If it is set for AlphaBIOS and we don't have a working VGA output, how can we get it back to the SRM console?

[Steven Schweda]

> [...] but after sliding back in, no VGA
> output comes through the adapter any
> longer, and we don't seem to get anything
> on COM1 or COM2.

Do you have any evidence that the thing isn't
fatally fried? (Tried repeated dis- and re-
assembly?)

Do you have a copy of the (HTML) "System
Reference and Maintenance Guide", typically
distributed as (a Zip-compressed)
"MIATASG.EXE"? That includes info on the LED
Status Codes, which might be informative.

The only way I can remember ever getting from
AlphaBIOS to SRM is with the usual menu
sequence, so that would have been using the
workstation keyboard with the graphics
display. I know of no other way.

I can't remember if there was any action on
the serial ports in AlphaBIOS mode. My PWS
has been disused for a while, but I could
probably (excavate and) fire it up, if you
get desperate for details and no one else
volunteers.

[Booting an Alpha Person]

Steve Schweda: Thanks, I found the Installation and Set Up guide, and says in the trouble shooting section that if a UNIX or OpenVMS system SRM console goes back to AlphaBIOS, it would be due to a dead battery, so that mystery is solved.

Still trying to find the DIGITAL Personal
Workstation System Reference and
Maintenance Guide. If anyone has a link or a copy available, or knows where I could find one, would like to decode the LEDs on the system board and see what they indicate.

[Steven Schweda]

> [...] If anyone has a link [...]

Google should have one. Did you try looking
for "MIATASG.EXE" (or "MIATASG.ZIP")?

[Booting an Alpha Person]

Finally found the Digital Personal Workstation System Reference and Maintenance guide (Oddly it's only available in Japanese on HP's own page..LOL)

I found a link to the HTML version-

http://www.turquoisewitch.com/~bradh/miata/Service/

And hopefully can get the VGA adapter working again as well.

[Booting an Alpha Person]

From the AlphaBIOS initialization section of the Digital Personal Workstation System and Reference Guide-

".....The video and keyboard devices are initialized first. If either the keyboard or the video device is not present or not functioning, AlphaBIOS directs its output to the COM1 serial port. This feature exists primarily to support the manufacturing process for the computer, but you may find it useful during the troubleshooting process. If AlphaBIOS does not start on your computer, first check that the keyboard is plugged in and operational. Also ensure that the video card and monitor are properly connected and working.... "

So there's a chance we can still salvage the system, by getting AlphaBIOS to communicate through COM1.

-Allan Beller

[Gene Hirschel]

Steven,
Alan and I are working together on this.

My guess was that when the video card was removed the system would default to Com1. The problem is, we tried every speed we could think of and no response. The cable and connectors as well as hyperterm are doing well as we can get to com1 on another alpha.

Here's the other strange thing: when we noticed the first behavior, that the AlphaBIOS would keep coming back like a bad penny, we ran to radio shack to get a new battery. We came back with 3 new batteries. When we replaced the battery, after popping the motherboard back in and rebooting the Alpha, nothing visible happened. We couldn't get anything at all on the VGA screen, nor from the Com1 port. Before we at least were getting the incessant AlphaBIOs that after F10ing several times would just default back.

Now we were getting a blank screen but plenty of fans whirring and drive clicking happily like our other Alphastation. It seems like the system is getting somewhere, but no clue as to what to do next. Yanking out the video card and hoping to get to talk to the Com1 port was our "hail mary" pass. Multiple CRs did nothing, and an LED serial port monitor was lit up but not flickering, making me guess there we were *not* in fact at the wrong baudrate, but nothing in fact was happening at all.

I did read about a bad bios which could have resulted during the replacement of the battery, perhaps a spike or some noise during the replacement. Any hint as to how we can "reset" the firmware or perhaps run a fail-safe firmware?

Thanks!

Gene and Alan.

[Steven Schweda]

> [...] fail-safe firmware?

Nothing I ever had to do, so I know nothing.

Did you ever try the other serial port?

[Hoff]

The FSL on these uses beep codes and LED codes, and wants to have a FSL floppy loaded in the drive.

Any beep codes? Any LED codes?

I'd swap in a different VGA board, as there might well be an issue with the existing VGA board, and given AlphaBIOS really wants to see a VGA board. See if that board works elsewhere, etc.

Also ensure the main board is properly seated and locked, as I've had one or two of these come unclipped, and a sufficiently canted main board can act entirely weird.

While in the box, I'd probably also re-seat memory, etc.

If you want to try the serial path (and I'm not at all certain you'll get anywhere with it), make sure you're connected in on COM1 and not on COM2. Make sure there's no direct connection to a keyboard or mouse (only via serial), and remove the VGA entirely. When you're connected, tap ^C ^C ^C during the power up sequence. 9600 8N1.

Is this really an -a and not an -au? If it's the former, then it was officially an NT "whitebox" configuration, and at least the older firmware definitely didn't play nice with VMS.

[Steven Schweda]

> Is this really an -a and not an -au?

When I added a Qlogic SCSI card (KZPBA or
similar) in my old-style PWS 500a, it started
calling itself a "500au". Trust no one, I
always say.

> [...] older firmware definitely didn't play
> nice with VMS.

Other than never getting the (unsupported and
known-not-to-work) IDE hardware to work, I
didn't have any particular VMS-related
problems with mine. But I'd've updated the
firmware if that was possible, so I may never
have tried it with "older firmware".

[Artlace]

Hi,

I am totally new to the forum but I have the same issue with my PWS 433a.

I have exactly the same symptoms: no VGA output, self-test is passed properly, nothing on (either) serial console with or-without null-modem cable (or straight cable) at any baud. I tried 9600-8N1 first.

However, keyboard if connected gets initialized and caps lock, num lock leds work (or is it just the keyboard controller doing  its job?). The machine complains if it doesn't get memory (after removal) with the documented 1-3-3 beeps.

The machine also responds to password reset, it beeps twice (long, short) and flashes all 8 LEDs.

Also all diagnostic LEDs go out at the end of POST but one. Right now the code reads "04" which is the second LED from the left at the bottom row. I found no information about this one either.

I tried removing all devices even the cache memory to no avail. 

Checked the clock switches which are properly configured for a 433a. As a last resort replaced the battery too.

Is this a known problem? Was there any resolution at the end? Should I accept that this one is dead too or is it just stuck in AlphaBIOS?

Thank you in advance

[Artlace]

Hi,

Just for future reference, I got another CPU board for the PWS433a but it has exactly the same problem. While I don't think this was a known issue or a regular fault, I opted in to replace the riser board too.

Since the riser board contains some PCI logic that still be at fault. It's possible that the ARC/SRM simply believes to have VGA output but it never gets that far - hence no serial output either.

[Artlace]

For future reference again:

After replacing the riser board the system came to life. Obviously either the board itself is faulty or one of the FPGA-s or ASIC-s simply gave up.

It did solve the issue for me, I have now VGA/Serial output. I have problem with the CD-ROM now but that's a whole different cup of tea, now I have the chance to debug anything.

Regards,

Artur