help with radeon 9250 in mediator1200 first edition

Ahhhhhhhh yes 'fun' we'll call it then :-)

Ok, I did some more testing, wondering if the newest libraries from Elbox might be causing lockups.

System Details:

- A1200 Motherboard 2MB CHIP, 4MB Fast memory.
- Standard internal floppy as DF0: + DF1: external floppy.
- Phase 5 Blizzard 1230MkIV 030@50MHz, 64Mb Fast Memory. Blizzard SCSI card. SCSI CDROM.
- Mediator 1200TX, MMCD_UP_2.1, ATI Radeon 9250 256MB.
- FastATA MkIII, 30GB IDE drive, DH0: DH1:
- Subway USB Controller, USB Keyboard and Mouse. USB Media card reader.
- AmigaOS Version 3.9 + Boung Bags 1 and 2.
- Picasso96 version 2.1b software.
- "be quiet" Power supply unit, 300Watts (BQT SFX:300W).

Tests and results:


Test 1. Roll back the Radeon.card.
- Swapped the latest Radeon.card version 2.19 (MM_CD_UP_2.2) for the previous version 2.12 (MM_CD_UP_2.1).
- Still getting lockups when Loadmondrvs tries to load DEVS:Monitors/Radeon.
- Picasso96Mode not listing any Board in the "Attach To" menu.
- Revert files to original setup.


Test 2. Remove other *.card files in LIBS:Picasso96/
- Moved the CVision3D.card out of LIBS:Picasso96/
- Still getting lockups.
- Picasso96Mode not listing any Board in the "Attach To" menu.
- Revert files to original setup.


Test 3. Roll back PCI.Library
- Swapped latest PCI.Library version 9.9 (MM_CD_UP_2.2) for the previous version 9.7 (MM_CD_UP_2.1).
- Still getting lokcups.
- Picasso96Mode not listing any Board in the "Attach To" menu.
- Revert files to original setup.


Test 4. Check for contention between ATA3.Driver and Radeon Monitor.
- In the S:Startup-Sequence comment out C:ATA3.Driver RESIDENT QUIET

C:ATA3.Driver RESIDENT QUIET (old value)
;C:ATA3.Driver RESIDENT QUIET (new value)

- Reboot, then manually run DEVS:Monitors/Radeon in a Shell window.

- - Surpise! No lockup and the RadeonMem is also now active. 300MB FastMem available.

- Picasso96Mode now lists Radeon in the "Attach To" menu.
- Attach the Radeon board in Picasso96Mode and Save.
- After a reboot the new resolutions saved in Picasso96Mode are now visible in SYS:Prefs/Screenmode
- Testing different resolutions proves successful.


Test 5. Reboot and manually rerun DEVS:Monitors/Radeon to check repeatability.
- Reboot, then manually run DEVS:Monitors/Radeon in a Shell window.

- - Surpise! No lockup.

- Testing different resolutions proves successful.


Test 6. Uncomment LoadMonDrvs in Startup-Sequence.
- In the S:Startup-Sequence uncomment C:LoadMonDrvs.

;C:LoadMnDrvs (old value)
C:LoadMonDrvs (new value)

- After a reboot the new resolutions are still visible in SYS:Prefs/Screenmode
- Testing different resolutions proves successful.


Test 7. Verify possible conflict between ATA3.Driver and Radeon Monitor.
- In the S:Startup-Sequence uncomment C:ATA3.Driver RESIDENT QUIET

;C:ATA3.Driver RESIDENT QUIET (old value)
C:ATA3.Driver RESIDENT QUIET (new value)

- The system will not reboot.
- Revert the C:ATA3.Driver line in the Startup-Sequence

C:ATA3.Driver RESIDENT QUIET (old value)
;C:ATA3.Driver RESIDENT QUIET (new value)

- System reboots ok.
- Testing different resolutions proves successful.


Test 8. Check if the order of loading the ATA3.Driver and LoadmonDrvs makes any difference.
- Move the ATA3.Driver entry in Startup-Sequence below the LoadmonDrvs entry.

C:LoadMonDrvs
C:ATA3.Driver RESIDENT QUIET

- The system will not reboot.
- Revert Startup-Sequence entries.
- System reboots ok.
- Testing different resolutions proves successful.



Conclusion:
- It appears there is a definite conflict between DEVS:Monitors/Radeon and C:ATA3.Driver.


I'll send this over to the Elbox Support folks and see what they can suggest.

Hopefully these tests might help some other Mediator owners in the future.

Great testing Ed! Will mess with mine next week once I'm back home from vacation

ok back from little vacation and back to Mediator testing! seems I finally got the Radeon 9200-256Mb to be reported correctly by pciinfo now but when I double click on the Radeon driver in devs:monitors I get a an immediate sytem hang. starting to wonder if with all the renaming, deleting, moving stuff around i got the radeon, radeon.info and radeon.card files messes up??

John/ Ed/ anyone else with a Mediator 1200 and Radeon 9200 with 256mb, can you please send over your working radeon, radeon.info and radeon.card files so I can attempt to get rid of my Mediator paranoia:unsure:

sent you a pm klx

Hope you had a good break Klx :thumbsup:

Just to add another pointer, I ended up removing my FastATA1200 and my setup is working ok now on the Blizzard SCSI. I'm not sure if you have a FastATA Klx but there are some serious issues with the combination of hardware/software, specially the ROM sockets which are flaky as hell. I also did an uninstall of Picasso96 and reinstalled fresh after doing the Elbox MediaCD 2.2 update.

I'll send a PM too. er... if I can figure out how to add attachments to PM's :oops:

just ask him for his email in private and send it that way.:)

no problems with fastata here,then again i changed the rom sockets on the motherboard for machined pin type,no problems after that bonus was i dont ever need cable ties now...

half the problems with adding expansions on these is mechanical connectivity issues.
anyway,must sleep,its been a long day:)

no problems with fastata here


or use dropbox

thanks John:) hopefully I can finally see something with my Radeon tonight

@ Ed.D

just enough to recharge the batteries a bit mate BUT alas when I've learned since I had my kids 9 years ago that there really is no relaxation for papa on vacations

well guys I think Im with Ed taking a break from fighting with this mediator before I smash stuff I know I'll regret smashing soon arghhh. Appears my problem is either a flaky connection or damaged mediator board as I just tried to load the files received from johim and surprise surprise pciinfo only seeing garbage yet again! Didn't touch or move the case at all.....Ed yes I do have a Fast ATA MKIII installed and it appears to be working fine but like you said who knows....

More testing after a stress break away from the 1200T :-)

Noticed a pattern in that the first boot up (after not using for a day or two) pciinfo always correctly finds my Radeon and every subsequent boot pciinfo reports garbage.

Tonight at first boot up I immediately went to devs:monitors and double clicked on the Radeon, went into Prefs:Picasso and board was attached. Went to test standard 800x600 screen at system hung. Every single attempt to reboot the system only brings up nonsense in pciinfo. It seems I only have one shot a day at initial boot up to have the Mediator see the Radeon correctly ?????

@ Ed

What version of the Fast ATA MK III you running ?

I've removed my FastATA for the moment and the system's been fine ever since. But the software version was the latest available from Elbox for the Mk III. Version 9.4a3.8 according to their website. link: http://www.elbox.com/downloads_fastata.html

I'll be replacing my FastATA once I've soldered in some new ROM sockets with the machined pin type that Roy mentioned and see if it makes any difference.

Just a random thought ~ I also wondered if the PFS3DS filesystem I've installed has any impact. On a Zorro 2 board like the blizzard the direct SCSI apparently can only use the first 16MBytes of Fast memory for disk access. Maybe something else tries to use that memory area as well e.g. the FastATA or the Mediator !?!?

if you have an apollo make sure the first thing the startup sequence loads is the fastata then the apollo libs then anything else including the mediator...

it can be a bit confusing as most of these ask to be first in the docs...

Thanks Roy, I'm using a blizzard 1230 MkIV and matching SCSI interface myself. Not sure what Klx300r is using.

I've got an Apollo 060@50 with 32 Mb ram. Will try modifying the S-S later tonight

Put Fast Ata first in my startup-sequence but still same issue. One boot pciinfo see the Radeon and the next it says 'no PCI devices in mediator'

the apollo 1260 can be funny beasts i have a blizzard 1260 in mine

have you the cpu 060 command

i have sent a link to my dropbox with my system in

Ya I do have all my ss commands and libs up to date and it's working great. I've hear d about the Viper 060 having issues with the mediator but not the Apollo 060?

i have heard people have problems with apollos and busboards,i only had minor problems though..

like i said,i had to put the fastata first and the anything apollo straight after it.i think it had something to do with the reset it does.
cant remember exactly its been a really long time.but it is a power hungry monster...one problem was the power supply.

as john says,they are funny cards sometimes.

heres the first part of mine

Quote:

---

C:ATA3.driver QUIET
C:stack 8192
PsdLoadModule DEVS:input.device QUIET ;Added by Poseidon-Installer
;BEGIN IDEfix
;If Exists C:IDEfix
; You may want to exchange the following line with LoadIDE reset quiet
; to make IDEfix reset resident, or LoadIDE start quiet
; to start IDEfix immediately and have it reset resident
; from the next reboot. Make sure to put LoadIDE behind any Kickstart mappers!
; C:IDEfix
;EndIf
;END IDEfix

---

Thanks John, will try a slimmed down version of yours later.

@ all Apollo 060 & Mediator 1200 Rev1.0 owners

Do you have poolmem, remapollo, mmumap, mufastrom on your SS??

Please let me see your SS

No luck with playing with SS so it's got to be a Defective mediator board sadly. Tonight I tried 2 PSU's too but alas it is what it is.

I'll be sending it off to Elbox and hopefully they can fix it.

Do you have any other acellorator to try with it eg anything 030 +

John I was just about to write that I'm heading over to my parents place this afternoon to pick up my 030@50 that's in my 1200 wedge. Absolutely last thing I can try so fingers crossed !

make sure its not a apollo1230 they definitely conflict with mediators mate.

yes i am using remappollo with a 1260.

What else do you have in your SS that's Apollo 060 specific Roy?

Same craziness with pciinfo reporting all kinds of strange stuff with my E-Matrix 030@50 :-( time to send the mediator off to Elbox & in the mean tune I can get my mancave cleaned up a bit since testing this 1200T it's looked like a hurricane hit it.

Btw, wow almost forgot the speed difference in WB the 060 makes over the 030!

One last thing you could try if you haven't already would be to temporarily remove your FastATA.

No luck ED with or without FastAta board, mind you mine is a MKII & not a MKIII like yours

Wondering if it's worth my while to just keep the board unpopulated and put my IndiAGA from my desktop in it ? But I guess I'll see how much Elbox wants to repair it first

Ok, I got a bit of a problem as well, I have a Mediator 1200 + Radeon 9200 and starting with a fresh install of everything after my hd (with all recent back ups on it) stopped working. So, as i am rebuilding the 1200 in a bigger tower I added the cards 1 by 1. Setting the Radeon up worked without flaw in P96 (at first) but it didnt recognize my network card anymore when I added that. fiddling with it it did show something in pciinfo sometimes. After a lot of frustration (and thinking it may be broken as well as the hd, perhaps during moving). I tried cleaning the contacts on the card with contact spray and... That worked :-). Now it is recognized correctly and goes online.
The problem now is screen corruption on Radeon screens (not just mouse but the whole screen), I am trying out different settings on the radeon tooltypes, perhaps that will help.
The second problem is that it doesn't add the memory to free memory (I tried setting it to 8 and 31 in prefs:env-archive/mediator/radeonmem but no luck so far.).
Are there any other settings that should be changed ?

My setup: A1200+Blizz060/603 + BB2 + MMCD_UP 2.2.
Jumpers on the Mediator: 8Mb, Wait: closed, config: open.
Changing the jumper on the radeon has no effect other than pciinfo reporting 1 or 2 radeon cards as far as i can see.
(before cleaning the contacts pciinfo sometimes reported up to 9 different pci cards :P
I have a FastATA as well but the driver is not installed yet.

Edit: Not sure about which/if boing bag is installed as cd-rom doesn't work without ata driver :-P, (confused with fs-uae where i did install both)

Hi WW,

When you say corruption on screen is it like an almost visible workbench with horizontal streaks going through various icons and even the mouse pointer? If so then it's most likely a loose Radeon card, just reseat it and it should be ok. If it's no picture at all then I'm not sure as it could be other issues. I'm still convinced (sticking my neck out here) that the mediator and other Elbox products are not 100% compatible on all Amiga's all the time.

Usually the Radeon memory is added to FAST memory after you've completed the Picasso96 setup. If the Radeon monitor driver isn't loading then I don't believe you're going to see the extra memory.

I just noticed you have an 060. That's a whole different world I've not experienced yet, perhaps the other guys can advise if you need to do anything in that area.

Good luck :) hope you can get it all going ok.

The Radeon works (800x600, 1280x1024 etc. ) at first, then after a while it sometimes gets corrupted (all over the screen). But this happens at random i will post a picture of the screen when it happens again. now it is online, with aweb open, dopus5 open and playing a mp3. (and its fine when i have camera ready :P).
ill try to clean the contact for radeon and for the accelerator as well. maybe you are right with that not 100% compatible all the time, perhaps they can't test every combination and version of 1200, accelerator , other setting etc.

Right, today the problem of screen corruption hasnt appeared (yet.., fingers crossed). The screen corruption looked a bit like an overheating gfx card, but would dissapear right after a reboot and the Radeon is barely lukewarm ( I added a heatpipe cooler to it). Right now my 1200 is on 800x600 16 bit on the radeon, playing a 192 kbit mp3 and copying files from my ftp :-). Two different things I did was; not opening aweb (old version from 0s3.9 cd) and iconifying genesis (which icon went missing somehow and only displays a dot icon on workbench). So there is a small chance that might be it or still a bad contact somewhere. Also, the blizzard is not very tight connected to the mediator (easy to put on or take it off).
Still, that leaves the question of not adding the memory to fastmem..
Another thing is that I never had screen corruption last year, before I moved. After i moved in this house the Amiga would not start and it turned out my hd had crashed and does absolutely nothing now. So for now i am going to let it play some music for a while :-) and mail Elbox if it should be possible with this old mediator as well to have that memory added.

As promised, a picture on the screen corruption. This appeared right after I opened workbench prefs to set icon borders to none and use other mem. but after a reboot ( I didn t save settings) the excact same steps did not corrupt screen. So, next I installed Roadshow, I wanted to try that for some time and to rule out something perhaps corrupted in the Genesis install (missing icon). I also tried another screen (a 24 inch lcd and a 19 inch one) and created new screenmodes with P96, sometimes corruption still appeared but not so often after I cleaned the edge connectors (of all pci cards and mediator - accelerator)
All was working fine for a while, I copied a large mp3 from my ftp, playing it in AmigaAmp with software decoding (put that 060 to work).
I did notice that with the 24 inch screen ( very nice, wb in 1920x1080 :-)) the Radeon got a bit warm, this is because of the higher setting for pixel clock right?
Next step; install the fastata software. the installer didn't work. it hanged when it was supposed to copy startup-sequence to ram:
Then I did a manual install, but it just hangs when the ata3 driver or the ata3prefs are started. and to make it worse it now doesnt even recognise the cf-card anymore.
I took it all apart, cleaned the clip-on thing of the fast ata, reassembled it and still no cf-card (nothing in early startup-screen). However, it does recognise a normal hd (the cf is only 2 Gb, and the only ide device, i took it off and connected the hd to see if the fastata still did work).
I think Im just going to give up on the CF as hd idea, it is giving me trouble (it did that as well last year, before I moved, only working in pio 0 mode) and just install a normal hd for now, perhaps try with a ssd later (with ide-sata converter).
So, to go back on this thread topic: The Radeon with Mediator. When the screen corruption appears and the screen is still readable a bit, it is possible to switch back to AGA (pal screen on 1438 monitor) and there is no corruption there (but still on the Radeon screen when switching back, in any resolution). It dissapears when I reboot.
There are two possibilities, hardware or software, (likely both i think). about the software: I remember reading something about settings in the documentation for the Radeon but cant find it in the MMCD_UP 2.2 docs and all of my older versions dissappeared when my previous hd crashed. Also, where do I set it to use the mmu? Perhaps there are other settings (toolypes) in PVS or something?
about hardware: I think I got the jumper settings right, perhaps someone with a working setup can confirm that?
Attachment 76573

That screen corruption looks just the same as I get and I don't think it's anything to do with settings or modes. A quick reseat of the Radeon and a reboot and it goes away ...for a while. I'm afraid I can't help with the CF card as I don't use them, but some of the folks here have suggested that some work and some don't with Amigas.

The MMU setup is in Medconfig when you do the configuration.

Ah, I think so too, cleaning the connector of the network card helped a lot already as well.
The cf cards i have work with the cf adapter but only at pio0 in the Amiga, I previously had those in my pc, working fine with euae, they were left over when I upgraded the pc. I also have a ssd left over so next plan is to try that (need to wait for a suitable ide cable I ordered from Amigakit, I also have a ide-sata adapter lying around.)
I have the mmu set to no with medconfig, was just wondering if this also perhaps needs to be set elsewhere (perhaps in PVS in prefs or somewhere else) .
The annoying part is that that corruption sometimes appears when I am making some change in settings, letting me think there is something wrong with that :-P

A big "AHA, Thats how it works" moment today. I always had MMU set to no (as medconfig recommended) and was limited to 8mb for RadeonMem. so today i decided to be a rebel :-P and set it to MMU = yes, and suddenly the valid range for RadeonMem is 0 - 98 :-) (i set it to 31). and set the jumper on the Radeon to 256 MB and behold, after reboot...:
.Attachment 76717
On another note I seem to be missing pciinfo (where should that be? it isn't in mmcp_up2.2?). and the screen corruption occurs less often now ( i also took off the fast-ata and even cleaned the mach chips on that)

What's Dutch for Well Done! ? :)

Pciinfo is on the original CD in "Mediator system Files/C" ...I just checked and it's not included in the updates.

Goed gedaan! :-) .In hindsight, is it just me or is the documentation a little unclear about that, (it does say it needs to work in mmu mode), medconfig threw me off by recommending the mmu set to no so I was wondering where to set that.

Thanks, found pciinfo on the cd. I have started from scratch with an empty hd, i think the CF-card is dying or something (it is also at least 7 years old).

well well after the headaches with my setup it actually was a defective Mediator all along http://www.amibay.com/showthread.php...l=1#post652248

so now back to the battle ground with my 1200T thansk to drewlio over at AW.net sending me a Mediator 1200 LT4 as a spare until Elbox gets mine back to me. So far so good with my Radeon finally going fine but few issues:

1- when setting up my Spider2 USB card Port 2 installs fine but 1 and 0 cause immediate system reset?? tried removing the card and placing in other pci slots but no difference. also tried reinstalling Poseidon v4.4 and also fat95 but same issue??

2- not really an issue as the system seems stable but my fast ram is reporting 258 MB when my Radeonn9200 has 256MB and my Apollo 060 has 32MB so total should be close to 288MB as this is what I had last time my 1200T was up (mind you with another HDO)