Weird CyberVision PPC issue (duplicate entries in the EarlyBootMenu/Boards)

[tinner]

Hello all,
This is my first post so just wanted to say hello to everyone and say that I am really glad to see so many people still interested in Amiga world.

I have a bunch of stability issues with an A4000 i bought a while ago and finally had a chance to boot it up.
The config is as follows:
- A4000D mobo in a tower, (default 150W PSU)
- CS MK3 060/50MHz
- CVPPC
- Micronik PCI board (PCI slots unused - the old style one) daughter board
Standard A4000D daughterboard (bought a new one for testing a while ago as well).
- Multiface, Toccata, VLab, Octagon (board rev working with 060 - don't remember the exact number but definitely higher than the one that causes 060 to freeze)
Ariadne II.
Scan Doubler, some SCSI drives/streamer/CDROM
ATA boot drive (340MB from the good old days that I kept from my previous A4000D back in a day)

Stability was the issue before I have managed to:
1. remove some RAM (as 4x16MB on MK3 caused errors - in one of the mem tests I've found on aminet) - so currently I have 2MB chip on the mobo + 32MB on the CS (2x16MB)
2. remove the leaking battery - done (first thing)
3. format the system and check the SCSI drives for bad sectors - done
4. install 3.1 on the first partition and update setpatch, 68040 library and... what was the first thing I did and don't remember now?
5. Removed all the boards apart from the scan doubler (as I had to bin C= 1942!!) and use the new daughter board that is definitely good.
6. System works and is stable! - everything works as expected until....
Until I plugin the CVPPC - this causes a strange error in the Early Boot Menu - when I go to "Boards" - I can see the CS 060 identified properly and then all other entries are duplicated with the ID of the CVPPC - don't have to say that booting such system causes WB 3.1 (no CGX installed at all) to crash shortly after loading...

Strange thing is that I remember booting this machine up shortly after purchase (Feb 07) and there was nothing like this going on...

My question therefore is if someone has experienced similar issues and if possible did anyone find a fix for it?

I would greatly appreciate your help - got the beast in the server room ready for testing so I can try any bit of advice shortly after work (usually around 5pm weekdays).

Well - thanks in advance.
tinner

[tinner]

Thanks for your reply.
I am not 100% sure if it should appear or not, but surely it should not appear more than once? Currently it appears about 7 times... I will post some pictures when I get back home.

Yes the installation was a fresh install - nothing extra installed on top of 3.1

[tinner]

Thanks for your interest.

Pictures below









The power supply is the standard A4000D PSU 150W.
Which actually is a good point considering that I have done all the tests running off the SCSI drives (2 SCSI drives, SCSI streamer attached, SCSI CDROM attached)... I will disconnecting everything in a minute leaving just CV PPC and ScanDoubler in (cannot do anything without the scandoubler as I don't have any other monitors I could use).

Well - keep the ideas coming - beers on me if that works!

tinner.

[tinner]

Right - disconnecting all SCSI devices from both the controller and the PSU did not help - same issue.

The PSU is actually 145W.

bugger...

[tinner]

Some more information:
ScanDoubler is made by arXon

Amiga Hardware

CyberVision PPC is the Rev 1.1 made by Phase 5 (1998) etching on the board states "Carda/Keller"

SuperBuster v11,

Daughterboard: Rev B, 1992 (GX B-V0) 9318 (Fish, Nines, Berlin)

Mobo: A4000 Desktop rev B, 1992

Cyberstorm: "Cyberstorm PPC Carda/Keller/Knabel/Zhang" Phase 5, 1996,1997

CPU: XC68060RC50 Malaysia (possibly never overclocked - as the quatrz seems intact and never soldered out)

Memory: 2x16MB 60ns (set to 70ns in CS' EarlyBootMenu), 2MB Chip onboard (standard 2MB module).

Is there anyone living close to Watford, UK? If I could just grab the CPU+GPU and test it in the other machine - that would really help...

thanks
tinner

[tinner]

Unfortunately - bought it with the 060/50 straight away... don't have any spare CPU boards...

Voltimeter is a good idea - will give it a go... but how do I check if the amperage is right? I cannot measure it while the computer is running as I would need to cut through the cables - right? Maybe I will try a PC PSU first... (after buying the ATX adapter) (don't feel too confident about playing with live PSU)...

thanks anyway. Any other ideas though?

[tinner]

Thanks Piru - this is great news. I remember flashing the firmware in 2007... However - could you confirm the source of this info? Or perhaps could someone with the latest flash confirm the duplicates in the boot menu? What worries me is that shortly after I plug cvppc in the machine it becomes unstable... hence why I've started to poke around and found the issue in the first place.
Well I will take it apart once, clean all the sockets, press-in all the chips and install the CGX. Then I will try to use the other socket on the cvppc without the scandoubler. but that might take a few days... Anyway thanks for yout help everyone - greatly appreciated.

tinner

[tinner]

I do apologise Piru - I've just found out that you are more than trustworthy source of information so no confirmation needed. However - do you happen to know if I could just flash back to older version? Reason being I remember the Amiga being more stable 2 years ago (before I've flashed it).

Kind regards
tinner

[tinner]

Right, Somehow I have managed to make it work more or less stable... I am still getting the multiple entries in the boot menu but I believe Piri is right - it could be normal.

After rearranging the sims on the CyberStorm I have managed to find a positioning where they actually work together (2 pairs of 16MB each, totaling 64MB on MK3) then have also managed to find 4x4MB that worked together and put them into the mobo. I am not sure what could cause it but after about 20-30 reboots and mem tests the whole system ran stable for an hour or so. (DOPUS filling the memory, session of SlamTilt, filling the Ramdisk with files, and copying them back to the HDD. System booted from SCSI drives as previously with clean OS 3.1 installed. Eventually I have put the Micronik daughter board in, filled all the slots with the cards I had and basically continued to test. Again after a painful process of putting one card at a time, booting the whole system up playing with dopus filling the Ramdisk and then restarting without startup-sequence to test the memory the whole thing seems to be stable! Then I have measured the voltage.
+12V measured to be 11.5V and the +5V measured to be 4.9V (after booting with all the cards and the CV PPC in). SCSI drives seemed to drain approximately 0.2V on the +12V rail as without them the total voltage raised to 11.7V and 5V. -12V seemed to be always the same = -11V
Taking all the cards out (except for the Scandoubler and CV PPC) resulted in (sic!) raising the voltage to +11.6V, +4.9V. Then I have connected another ATA drive (old power hungry Maxtor) and that seemed to overload the PSU as the Amiga did not boot at all (although the voltage only dropped to +11.4V +4.85V) but when I pressed the reset button the meter went nuts - measuring tops of +3V on the +12V rail - without the maxtor - Amiga still seems to be stable. Although I think missing 0.5V on +12V rail and 0.1V on +5 might be the edge for this configuration. I will look into a new PSU soon but for now it seems to work? Although I still haven't installed any software, did not play with it for more than an hour. So rearranging the cards, assembling the whole thing properly (all the screws, etc), pressing all the chips across the mobo and the expansion cards seemed to help for now. I will get a new PSU, new battery and post more results later on when I find a free day to play with my "new" toy

Well - thank you all for your help - it looks like we've nailed it!

kind regards
tinner.