29 June 2022, 20:38 | #1 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Elbox Mediator 4000Di MKIII PCI lockups
Yo! bromigos!
Any Mediator A4K MKIII Di 3.3v owners out there? Long story short. Sent my A4K off to have a recap with a trusted Bromigo. All cards working at the time which consists of; PCI 1 = Soundblaster Zorro = x-surf 100 + RapidRoad usb Zorro =BigRAM + PCI 4 = Radeon 9250 128MB System has a Blizzard Cyberstorm 060/PPC + CVision card. Amiga OS 4.1FE Update 1 With any PCI card fitted Amiga 4.1 system hangs on Purple Screen. If you use any Zorro-III card then Boots fine Have tried reseating the Elbox and played around with the jumpers on the CPLPD card. Jumpers before; Master = on/closed Swap = off/open Winsize = off/open Have tried; Master = off/open Swap = on/closed Winsize = on/closed Also tried 3 different PCI RTG cards. 1 x 128MB Radeon, 1 x 256MB Radeon and Voodoo 4. Any ideas? Thanks Mangs! |
30 June 2022, 14:02 | #2 |
Thalion Webshrine
Join Date: Jan 2004
Location: Oxford
Posts: 14,337
|
There are several reports of "purple screen" being down to issues with the ROM.
Did you change your kickstart ROM at the same time? |
30 June 2022, 16:22 | #3 | |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Quote:
Even when trying to do a fresh install with 4.1 boot floppy and CD it sticks at the purple screen, so know it’s not a corrupted SD/HDD issue either. Going to try and build a quick 3.1 install with mediator setup and see what PCIINFO shows just to make sure I’ve not got a CPLD issue on the PCI side. |
|
01 July 2022, 02:13 | #4 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Well the plot thickens….Just tried booting off a good known working 3.1 install disk and the system goes into a reboot cycle… Although the ROM screen shows on power up.
Looks like I need to strip the system down and recheck connections.. |
01 July 2022, 04:53 | #5 |
Registered User
Join Date: Aug 2008
Location: London / Canada
Posts: 781
|
Ok, so before you sent it off it was working fine?
Are you sure maprom is disabled in the CSPPC menu (hold ECS down while booting). |
01 July 2022, 05:07 | #6 |
Registered User
Join Date: Mar 2009
Location: texas,usa
Posts: 235
|
Are the roms seated well? maybe a bent pin under it?
If you have another set of roms or old 3.0 roms try them and see if a basic 3.1 floppy boots? as for mediator, reseat the cpld chips in the bridge,could be bad socket contact. i've seen some bridge cards finicky between the zorro and pci slot,you might clean the contacts on the bridge card. |
01 July 2022, 17:50 | #7 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Thanks Mangs!
So I’ve made some progress. System stripped down, cleaned contacts. Tried replacement ROMS. Still the same issue. Changed the ROM speed jumper from 200ns to 160ns. ROMs are the EPROM Cloanto types. Then had a look into the 3.1 issue. The Commodore 68040 Library in Libs: was the culprit for the Guru Software 80000004 error. As soon as I renamed it to .old system came to life. I also reflashed the Cyberstorm to 44.71 and defaulted the SCSI2SD rev 6 with 32GB to default and re-programmed that back with ID and settings. Now It appears I have a fresh canvas with a fresh basic 3.1 workbench. Next I plan on installing MMCDV2 Mediator files with the Mediator Daughter board fitted. Zorro-III side appears to be working as the BIGRAM+ 256MB is showing up as FastRAM. Fingers Crossed the PCI side co-operates. Will report back. |
01 July 2022, 19:42 | #8 |
Registered User
Join Date: Sep 2018
Location: California
Posts: 322
|
Hey there! I have a similar build I haven’t yet assembled and configured. I plan to use 3.2 ROMs. I’m happy to try to reproduce the issue you’re seeing if you’d like, just let me know how to proceed.
|
01 July 2022, 22:40 | #9 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
So. I managed to install Picasso96, MMCDV2 and the MMUP2.6 on 3.1.
When running “PCIINFO” I’m getting no joy Tried 1 card after another and PCIINFO shows 3 PCI slots as “UNKNOWN”. Put jumpers as all closed. I even got a defective card issue when adding in the BigRAM plus The system was working before I sent this off. All CPLDs are securely in place. Looks like it’s a new mediator |
01 July 2022, 22:41 | #10 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Are you putting together a 1200TX system? If so the A4000 Mediator is different as it has both Zorro-III and PCI slots
|
01 July 2022, 22:51 | #11 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
|
02 July 2022, 03:56 | #12 |
Registered User
Join Date: Sep 2018
Location: California
Posts: 322
|
|
02 July 2022, 09:20 | #13 | |
0ld0r Git
Join Date: Mar 2009
Location: Cornwall, UK
Posts: 1,570
|
Quote:
|
|
02 July 2022, 13:37 | #14 | |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Quote:
The only worrying thing is, that pciinfo only shows PCI 1 to 3. Yet I’ve placed RTG cards in PCI 4 |
|
02 July 2022, 13:41 | #15 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Feel free Bromigo!
I’m using the following; PCI 1 = Soundblaster Zorro 2 = X-surf100 + RapidRoad USB Zorro 3 = BigRAM+ PCI 4 = Radeon 9200 or 9250 or Voodoo 4 RTG Jumper setting on mediator control board; Master = closed Winsize = closed Swap config = closed Have also tried; Master = Open Winsize = open Swap config = open Also tried; Master = open Winsize = closed Swap Config = Open |
02 July 2022, 13:59 | #16 |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
So Vendor.txt is in S:
PCIINFO only shows; PCI 1 and PCI 2 Vendor ID= $FF02 vendor = Unknown Although I only have 1 card in PCI 4 |
03 July 2022, 02:30 | #17 |
Registered User
Join Date: Aug 2008
Location: London / Canada
Posts: 781
|
I was involved in beta testing OS 4.1 Classic. I can state that we were not able to test all combinations of cards in the Mediator in conjunction with Zorro cards. That means you could have address space issues or other problems.
But if you say it was working before, then it must be something else. You could try the recommended card order (see OS4.1 Classic FAQ - lots of good info in there.) Try eliminating cards. We did not test the BigRAM as it wasn't out at the time, but we did test ZorRAM cards (very similar I believe). Darren |
03 July 2022, 02:31 | #18 |
Registered User
Join Date: Aug 2008
Location: London / Canada
Posts: 781
|
One more thing, you could turn on serial debug and use the debug kernel. Save the output to another machine connected via serial cable. That would tell us what's going on on the OS4 side.
|
03 July 2022, 06:29 | #19 |
Registered User
Join Date: Aug 2007
Location: USA
Posts: 359
|
I have an Amiga 4000D with a CS MK II 060 and a Mediator 4000D - the newer 3 volt version. I am using a Radeon 9250-256meg, Spider USB, and a Lan card. I have 2 versions of the Bigram 256 (2016 and 2019 I think) and neither one of them works with the Mediator daughter board. Both versions of the Bigram work fine in my other 4000D with the original C= daughter board. Both of my 4000s have a rev 11 buster.
|
03 July 2022, 11:42 | #20 | |
Banned
Join Date: Feb 2022
Location: Anywhere and everywhere I have a contract
Posts: 822
|
Quote:
So know these all worked before |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Mediator 4000Di 3V MKIII PCI memory issues | amigapt | support.Hardware | 11 | 09 February 2020 13:34 |
Mediator PCI 4000Di 3V | ElectroBlaster | support.Hardware | 4 | 27 October 2017 05:23 |
Mediator 4000di MkIII card order | cmsj | support.Hardware | 1 | 21 July 2016 15:43 |
WTS/Trade Mediator PCI 4000Di Busboard | soulboy | MarketPlace | 10 | 05 December 2009 07:13 |
Mediator PCI 4000Di | lopos2000 | support.Hardware | 11 | 01 December 2006 18:03 |
|
|