| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 7272.1 | is it parity memory? | 34544::WALLACE | David Wallace, SBU Sales, @WKO | Wed Dec 18 1996 21:24 | 7 |
| 7272.2 | chip check? | ODIXIE::SIMPSONT | PC = World's biggest conjob? | Thu Dec 19 1996 10:28 | 12 |
| 7272.3 | | 30461::LIN | Bill Lin | Thu Dec 19 1996 11:43 | 12 |
| 7272.4 | no markings | ODIXIE::SIMPSONT | PC = World's biggest conjob? | Thu Dec 19 1996 16:06 | 24 |
| 7272.5 | for digital PARAGON = non-qualified vendor! | JGODCL::BLOEMENDAL | Wim, JGO B-1/06, 889-9364 | Mon Dec 23 1996 06:07 | 15 |
| 7272.6 | who qualifies? | ODIXIE::SIMPSONT | PC = World's biggest conjob? | Mon Dec 30 1996 23:41 | 13 |
| 7272.7 | Only Digital Memory is supported ie:warranty | PCBUOA::WHITEC | Parrot_Trooper | Tue Dec 31 1996 12:03 | 14 |
| 7272.8 | | QUARK::LIONEL | Free advice is worth every cent | Tue Dec 31 1996 12:13 | 6 |
| 7272.9 | That boards' a real MOTHER! | ODIXIE::SIMPSONT | PC = world's biggest con job! | Mon Feb 24 1997 19:36 | 25 |
| Update...
I still have the problem with #$%@#$ parity errors.
Installed the BIOS that supports the 83mHz overdrive chip and installed
an overdrive chip (found one at a computer show for $100). No real
change in the problem... Upgrade works fine when the memory
cooperates.
I checked the power supply voltage with a DVM and also a o'scope.
Looked solid measured at the power supply pin of a chip at 5.09v.
All three banks of memory give me parity errors (2xmb4 DEC, 2x4mb and
1x16mb Paragon), but have worked ok when installed in an LPv system at
the office.
I'm guessing it's motherboard time... Any other ideas? Might anyone
know what the cost will be? I really would like to keep the system for
my daughter's use (so I can buy the new one for me :) ), rather than
scrapping it.
I still hate PCs...
Regards,
Tom
|
| 7272.10 | | BBQ::WOODWARDC | ...but words can break my heart | Tue Feb 25 1997 17:11 | 5 |
| >I still hate PCs...
that's ok - from past experience, they hate us too ;')
H
|
| 7272.11 | still down! | ODIXIE::SIMPSONT | PC = world's biggest con job! | Wed Mar 05 1997 17:33 | 24 |
| Well, I'm still without a functional desktop PC... Thank goodness my
HiNote hasn't gone on strike!
I took the system to the DEC Service Center. They kept it for a few
days and said they could not find a problem. No problem, I expected
that since it is an intermittent problem. However, when I plugged it
and checked it out, I could get it to fail in less than 5 minutes. It
won't complete a WNT boot without failing...
When I took it back to the Service Center and explained the situation to
a different guy at the counter, he said it could be a software problem.
I told him that I have problems with 3 different operating systems on
the same system. He said that's probably the problem...these systems
weren't made to boot multiple OS's... I'm sorry, but I just don't buy
that. He also said that memory parity errors could be a software
problem. I'm not sure that I buy that either. I'm begining to lose
faith in our MCS PC repair contractors.
Can memory parity errors be caused by software problems?? I always
thought that a parity error was triggered by a hardware parity check
failure. I've never had a failure running QA+ diagnostics.
Regards,
Tom
|
| 7272.12 | not yet...! | ODIXIE::SIMPSONT | PC = world's biggest con job! | Mon Mar 17 1997 21:56 | 18 |
| Still no desktop system. They say it's fixed, I check it and it fails
in less than 5 minutes...
This time he said it was a known problem with SCSI speed VS the
internal SCSI ribbon cable. They slowed the bus speed down to 5BM/sec
and said that it ran fine. It fails for me, no problem.
The service center has been great so far, in not charging me for the
repairs, since it is out of warrenty, but this is really getting to be
a drag! I wouldn't mind paying if they could just get the damned thing
fixed... It's real close to going under the back wheels of my car...
Anybody have a better suggestion?
rgds,
Tom
|
| 7272.13 | | TARKIN::LIN | Bill Lin | Mon Mar 17 1997 23:56 | 9 |
| re: .12 by ODIXIE::SIMPSONT
Tom,
How about breaking down and using Genuine(TM) DIGITAL memory? Worth a
try? Are you sure the servicenter used your memory and didn't use
their own known good memory for test purposes?
/Bill
|
| 7272.14 | tried originals... | ODIXIE::SIMPSONT | PC = world's biggest con job! | Tue Mar 18 1997 19:23 | 8 |
| Bill,
I have tested with only DIGITAL memory simms. It was painful to go
back to only 8mb of memory, but it too, eventually failed with parity
errors. So all three sets of simms are bad?
Regards,
Tom
|
| 7272.15 | | TARKIN::LIN | Bill Lin | Tue Mar 18 1997 19:36 | 10 |
| re: .14 by ODIXIE::SIMPSONT
Hi Tom,
There's still the cache... I don't recall if the cache is parity
protected or not. If there is parity, then failure may be detected.
If not, then errors could occur and escape detection for some time.
Or it could be a bad SIMM slot causing intermittent errors.
/Bill
|
| 7272.16 | I don't think so | ODIXIE::SIMPSONT | PC = world's biggest con job! | Wed Mar 19 1997 14:56 | 11 |
| Bill,
Been there, done that... Replaced the CPU w/83mhz Pentium Overdrive,
so that cache has been replaced. The errors still occur with the 256kb
l2 cache disabled via the BIOS setup. Is there anything else?
I tend to lean toward a bad SIMM slot or flakey memory buss... I'm
returning it to the DEC Repair Center again, today.
Thanks,
Tom
|
| 7272.17 | looks like Adaptec problem | ODIXIE::SIMPSONT | PC = world's biggest con job! | Mon Apr 21 1997 18:44 | 19 |
| To (hopefully) finish this thread...
Well, it looks like the problem is in the Adaptec SCSI controller. I
pulled it out and replaced it with a new IDE controller and disk, rebuilt
W95 and the problem has not occurred since. After I rebuilt the system
disk, I put the SCSI controller back in, booted off the new system
disk and tried to use the user disk and CD-ROM. Still gets parity errors,
apparently disk parity errors, not memory. I guess that's why it didn't
call out segment:offset memory addresses when the errors occured?
I tried both 5 and 10mb transfer rates and different ribbon cables.
Still got errors. Now I just need to verify that the disk and CD-ROM are
not to blame before I send the controller back to Adaptec.
This stuff is not rocket science, it just takes a lot of persistence
(and money).
Regards,
Tom
|