Welcome, Guest. Please login or register.

Author Topic: Memory woes - part 2  (Read 1179 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline IznougoudTopic starter

  • Newbie
  • *
  • Join Date: Jan 2007
  • Posts: 21
    • Show only replies by Iznougoud
    • http://www.virtual.st
Memory woes - part 2
« on: August 03, 2008, 04:21:10 PM »
I've been at this before, in another post, but have finally managed to find some kind of structure in this mess by consequently running MemTest:

No errors when AOS3.9 is loaded

Errors reported in FastMEM and ChipMEM outside AOS3.9 (in DOS and AOS3.1)

Errors are also reported in ChipMEM when the 2040-card is deactivated.

This is odd. I find it hard to believe that a malfunction of the Blizzard-board would cause errors in ChipMEM - even when the Blizzard-board is deactivated. I find it equally hard to  believe that a malfunctioning motherboard/ChipMEM, would cause errors in the FastMEM of the Blizzard-board.

And, as noted in another post, no other software, like CheckMEM and others, report any errors whatsoever. Something is wrong though, as indicated by instability.

I've also received the message "2060-removable driver failed" on a couple of occasions lately. And, I might add, the program "WhichAmiga" hangs the entire system, unless I deactivate the 2040-card.
 

Offline Zac67

  • Hero Member
  • *****
  • Join Date: Nov 2004
  • Posts: 2890
    • Show only replies by Zac67
Re: Memory woes - part 2
« Reply #1 on: August 03, 2008, 06:02:31 PM »
Well - in theory - this could be caused by flaky bus buffers (74245) and/or their _OE signal. Provided you've double checked the PSU(!) and ruled out heat issues, you could use a logic analyser to check the _OE pins (19) on the 74FCT2245 chips for any spikes when the board is deactivated.

Do you have any chance to test the board in another 2000?

Are the error bits always the same or is there any pattern in the errors?

PS: make that a logic tester, no analyser  ;-)
 

Offline IznougoudTopic starter

  • Newbie
  • *
  • Join Date: Jan 2007
  • Posts: 21
    • Show only replies by Iznougoud
    • http://www.virtual.st
Re: Memory woes - part 2
« Reply #2 on: August 03, 2008, 06:56:57 PM »
I've tried using another PSU - no change. Then again, all PSUs for the A2000 are more or less the same age (<> 20 years) and could possibly display the same symptoms of fatigue.

I'm looking for an opportunity to test the Blizzard-board in another A2000 - but they are not easy to come by these days.. ;)

As for the error bits, I honestly don't know. The listing is quite overwhelming, so I've never even tried taking notes.
 

Offline orange

  • Hero Member
  • *****
  • Join Date: Dec 2003
  • Posts: 2796
    • Show only replies by orange
Re: Memory woes - part 2
« Reply #3 on: August 03, 2008, 08:12:35 PM »
have you tried another testing program?
IIRC, memtest-20 would always display errors here. (try memtest-00)
Better sorry than worry.
 

Offline IznougoudTopic starter

  • Newbie
  • *
  • Join Date: Jan 2007
  • Posts: 21
    • Show only replies by Iznougoud
    • http://www.virtual.st
Re: Memory woes - part 2
« Reply #4 on: August 03, 2008, 09:04:25 PM »
Unfortunately, MemTest-00 doesn't give any output at all. It just causes a crash.