I was taking a look at an old Multia that's been sitting idle for a long time. It does not boot normally, but I'm able to do a fail-safe boot from the FSL image on floppy. When that boots, it spews a bunch of machine checks on the serial console. Before I write this system off, is anyone able to infer anything from these messages? If there is something I could do to fix it, I'd love to know. There's pretty much no FRUs on the VX40s other than RAM however, and I swapped in multiple pairs of SIMMs in one bank at a time to no avail. TIA.
- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
Multia VX40 troubleshooting
- Thread starter coredump4
- Start date
Wildfire
Experienced Member
I sent this to someone who is active in comp.os.vms
Eventually he is able to help or can at least say if this system is a dead brick now.
He has deep knowledge in reading crashdumps...
Eventually he is able to help or can at least say if this system is a dead brick now.
He has deep knowledge in reading crashdumps...
Thank you, I appreciate it!
Wildfire
Experienced Member
Hello,
first of all i want to ask you about the ram you used - are you 100 percent sure it is the right ram ?
I could take a look in my VX40 which i want to get up and running when i have the time...
The ram must be full parity 36 bit, i have 2 modules 4m x 36 in it.
If yours is also right there's no good news, i got the following info:
literal EMB$C_MCHECK_670 = 2; ! A Machine Check 670 - Processor UCE
UCE = UnCorrectable Error
The infos you posted are no crashdump bc the system was still pre boot.
But he was able with this here:
Unexpected Machine Check through vector 00000067
to take a look into some OpenVMS source code he has access to and got the upper info.
There seems to be a troubleshooting manual at https://archive.org/details/udb-man/page/n41/mode/2up
When you take a look in it there is said for the VX40 you can only try new memory modules as a test.
If this would be the other versions VX41 or 42 you could also remove the cache module and replace it as a test.
If that all fails - replace the system board !
Are you a little bit into electronics repairing ??
If yes you could test to get new cache chips and solder them onto the board.
Eventually there is a small chance that some solder points are bad eventually.
If that also fails the system is just a brick... :-(
I think there may be some interest of people here if you don't wanna keep it.
At least for the case with powersupply.
I would desolder or eventually take a saw to remove the cpu with heatsink to keep it as a display item, eventually this processor
will also fit in the AXPpci33 mainboard...
HTH - at least a little bit...
first of all i want to ask you about the ram you used - are you 100 percent sure it is the right ram ?
I could take a look in my VX40 which i want to get up and running when i have the time...
The ram must be full parity 36 bit, i have 2 modules 4m x 36 in it.
If yours is also right there's no good news, i got the following info:
literal EMB$C_MCHECK_670 = 2; ! A Machine Check 670 - Processor UCE
UCE = UnCorrectable Error
The infos you posted are no crashdump bc the system was still pre boot.
But he was able with this here:
Unexpected Machine Check through vector 00000067
to take a look into some OpenVMS source code he has access to and got the upper info.
There seems to be a troubleshooting manual at https://archive.org/details/udb-man/page/n41/mode/2up
When you take a look in it there is said for the VX40 you can only try new memory modules as a test.
If this would be the other versions VX41 or 42 you could also remove the cache module and replace it as a test.
If that all fails - replace the system board !
Are you a little bit into electronics repairing ??
If yes you could test to get new cache chips and solder them onto the board.
Eventually there is a small chance that some solder points are bad eventually.
If that also fails the system is just a brick... :-(
I think there may be some interest of people here if you don't wanna keep it.
At least for the case with powersupply.
I would desolder or eventually take a saw to remove the cpu with heatsink to keep it as a display item, eventually this processor
will also fit in the AXPpci33 mainboard...
HTH - at least a little bit...
Very interesting. Cache seems as though it would be a likely source of machine check exceptions. I have a working VX40 and a VX41 (with the CPU from a VX42), and @coredump4 and I have been working on these Multias for many years.
- Alex
- Alex
I have a VX40. The battery voltage was low so I removed it. So to boot it I have to go into the BIOS (or whatever setup is required) to make it boot. It has NT 4.0 on it. I really don't use it but can look at things for you if needed.
Wildfire
Experienced Member
@coredump4
What happens when you try to boot it normally without the fsl floppy - is there any output over the serial console you can also post ?
What happens when you try to boot it normally without the fsl floppy - is there any output over the serial console you can also post ?
Re: RAM, I used genuine DEC 8MB that came from other VX40s and tested good last year.
Looking at the machine closer, it had had E215 replaced due to Heat Death, but the repair wasn't clean. I'll re-do the repair and see if the machine checks persist.
Looking at the machine closer, it had had E215 replaced due to Heat Death, but the repair wasn't clean. I'll re-do the repair and see if the machine checks persist.