Amiga.org
Amiga computer related discussion => Amiga Software Issues and Discussion => Topic started by: Amiduffer on May 21, 2009, 06:32:14 PM
-
Things are progressing after my A3000 crapped out.
Now that everythings in the new A3000, and its without a 3640 accelerator, I'm now having a Ramlib error during boot up, which sucks.
Another thread said that Ramlib errors were caused by having the 68040.library in LIBS, since I'm now running with the stock 68030, but after I removed the 68040.library, I'm still getting Ramlib errors. Anything else I have to do?
On the other hand, games loaded from floppies work!
-
Just comment everything out in s:startup-sequence & s:user-startup that might make ramlib crash, by putting a ";" (without the "") infront of the programs. After this is done and you run startup-sequence 'light', if it doesn't crash anymore you can enable line by line again.
x303 :D :D :D
-
Errr. That's a lot of "work", but, I suppose I'll have to do it.
On the other hand, what you think about this?
I've been running OS3.9 so far. Now that I'm back to a 68030 (until I can get another accelerator), should I re-install a plain 3.1 WB? Its got two, 2 gig HD's in it that are each divided into two partitions, so I shouldn't have trouble, or will I?
Its too bad those new 060 GVP accelerators weren't slightly more affordable. :)
-
Have you got 3.1 ROMs in your new A3000?
-
My a3000 does/did the same. Drove me nuts....until I cleaned the pins and connector
on the power supply cable. You must remove the shelf and power supply to get at them
Use a good contact cleaner and a very small something to get between the pins.
Amiganaut
-
My a3000 does/did the same. Drove me nuts....until I cleaned the pins and connector
on the power supply cable. You must remove the shelf and power supply to get at them
Use a good contact cleaner and a very small something to get between the pins.
Amiganaut
Yes Moto, I transfered the 3.1 Roms, the entire drive tray, and all the Zip Ram into this spare 3000.
What do dirty pins on the power supply have to do with Ramlib errors??
-
Dirty connectors, especially ones to the psu and ram can cause all manner of interesting and difficult to otherwise pin down issues. Likewise, ropey PSU's can do the same.
-
Dirty connectors, especially ones to the psu and ram can cause all manner of interesting and difficult to otherwise pin down issues. Likewise, ropey PSU's can do the same.
OK. Sounds strange, but, I'll do it.
-
My A3000 gave me a ramlib error (still does occassionally) and I had to disable Amidock in WBSTARTUP.
Then, I found if I played with the maxtransfer in the hard drive settings it would reduce the occurence of RAMLIB errors and allow Amidock to work without causing RAMLIB errors. It was bizarre, and I don't really understand the cause.
-
Things are progressing after my A3000 crapped out.
Now that everythings in the new A3000, and its without a 3640 accelerator, I'm now having a Ramlib error during boot up, which sucks.
Another thread said that Ramlib errors were caused by having the 68040.library in LIBS, since I'm now running with the stock 68030, but after I removed the 68040.library, I'm still getting Ramlib errors. Anything else I have to do?
On the other hand, games loaded from floppies work!
Hve you set the CLOCK jumpers to INTERNAL when you removed the 3640? (IIRC the 3640 uses EXT clock)
and ... does your plain A3000 have the FPU chip installed? maybe theres something in WBStartup that requires it ..... just guessing
-
OK. Sounds strange, but, I'll do it.
It's even more strange when you take a supposedly unstable system, give it a good clean and it cease to be unstable.
Even if it doesn't work for this particular issue, it is something that should be done with a computer every once in a while anyway.
Hope it does work out for you however :)
-
My A3000 gave me a ramlib error (still does occassionally) and I had to disable Amidock in WBSTARTUP.
Then, I found if I played with the maxtransfer in the hard drive settings it would reduce the occurence of RAMLIB errors and allow Amidock to work without causing RAMLIB errors. It was bizarre, and I don't really understand the cause.
I arranged some of the jumpers, and its a little more stable. Or it might be because of the cleaning of the pins on the PSU connector. I still randomly get a RAMLIB error though. Taking out Amidock might be a good idea. That is the cause of crashes at boot up too. So strange. I never used to get RAMLIB errors before this.
Have you set the CLOCK jumpers to INTERNAL when you removed the 3640? (IIRC the 3640 uses EXT clock) and ... does your plain A3000 have the FPU chip installed?
Yes, I set that pin. Not sure about the FPU. I'll have to use the viewer to see what I have now that the 3640 is out.
-
Ramlib errors can also be the result of ramlib having a too small stacksize.
The original ramlib stacksize is 2k. MCPramlibpatch increases the ramlib stack to 8k.
There also is a blizkick module that fixes it.
You might want to try one of these patches to see if it fixes your problem.
-
@MozzerFan
I believe the stack size problem is fixed with newest SetPatch...
SetPatch (V44.38) has already been installed. Patch list:
68060 Support Code Loaded
Enabled Advanced Graphics Modes
ramlib signal usage
stack size check for CON:, RAW:, RAM:, PAR:, PRT:, SER:
graphics/WritePixelLine8()
graphics/WriteChunkyPixel()
Enabled data cache
@Amiduffer
I would recommend staying with AmigaOS3.9 especially if you have a RTG gfx board. If you don't have a RTG board, I would recommend you get one. It makes a huge difference even on a 68030.
-
@Matthey
Nope, you can check it with scout.
Goto tasks and doubleclick on ramlib.
The stacksize is still 2k without mcpramlibpatch (or a similiar patch).
-
@MozzerFan
@Amiduffer
I would recommend staying with AmigaOS3.9 especially if you have a RTG gfx board. If you don't have a RTG board, I would recommend you get one. It makes a huge difference even on a 68030.
I have a C64/3D RTG board, but without a Buster 11, it doesn't work in the A3000.
-
@MozzerFan
I checked Scout and my ramlib stack is still set to 2k. I am using all the latest patches to a custom 3.9 BlizKicked ROM (MAPROM). I don't have stability issues. Did you get crashes from the ramlib before installing the patch? I'll have to dl and read MCP's docs now.
-
@Matthey
I can vaguely recall my system giving ramlib errors with Amidock.
MCP ramlibpatch used to be included with MCP, but with the latest releases it's not included anymore.
You can find it in the RTGMaster archive on aminet though.
A similiar module is included in the blizkick archive, and the romfixes module of blizkick also includes this patch.
Anyway, it can't hurt to give it a try when your system is
giving you this error.
-
I have a C64/3D RTG board, but without a Buster 11, it doesn't work in the A3000.
I'm pretty sure I ran my CV64/3D in my 3000 without Buster 11. It was rock solid on my Amiga although I did upgrade to the Buster 11 later. I used P96 and it was more stable and bug free than the Voodoo driver I use now. The boards performance matches the 3000 well and provides a huge improvement over ECS. Maybe the problem you are having with the crashes is related to the CV64/3D not working. I think it's time to do some cleaning. You should also push down on all the socketed chips on the motherboard, especially memory chips. They work up over time from heat cycling. It could also be the power supply getting old but I would try everything else first. You might blow out the power supply though.
-
@MozzerFan
Thanks, I'll look into the ramlib patch. I thought the BlizKick romfixes was unnecessary with the latest patched AmigaOS 3.9 modules. I'll have another look.
-
I'm pretty sure I ran my CV64/3D in my 3000 without Buster 11. It was rock solid on my Amiga although I did upgrade to the Buster 11 later. I used P96 and it was more stable and bug free than the Voodoo driver I use now. The boards performance matches the 3000 well and provides a huge improvement over ECS. Maybe the problem you are having with the crashes is related to the CV64/3D not working. I think it's time to do some cleaning. You should also push down on all the socketed chips on the motherboard, especially memory chips. They work up over time from heat cycling. It could also be the power supply getting old but I would try everything else first. You might blow out the power supply though.
The CV64/3D runs fine in my A4000, so I know it works. I originally bought it for the A3k, but, it refused to work in it, and from the info available online, I concluded that the Buster was the issue. All socketed chips have been firmly pushed down and I've cleaned up a lot of the dust.
-
@Amiduffer
Any change in the errors?
A Buster 11 may be worth tracking down. It fixes some bugs and it's faster at Zorro transfers.
-
Just comment everything out in s:startup-sequence & s:user-startup that might make ramlib crash, by putting a ";" (without the "") infront of the programs. After this is done and you run startup-sequence 'light', if it doesn't crash anymore you can enable line by line again.
x303 :D :D :D
An easier approach is probably to boot without startup-sequence, enter the line
"set echo on"
and
"execute s:startup-sequence"
and see at which line the ramlib guru comes up.