Amiga 500 Kickstart 3.1 error #8000000B

Hello dear felllows,

I just received my Amiga 3.1 upgrade kit ( which contains the kickstart and the 6 floppies )
I removed the old 1.3 kickstart and replaced it with the 3.1 ( i linked the pins 1 and 31 on the socket and the bend up 31 and 21 on the ROM , because i have a REV 5 motherboard )

The workbench launches but i have a #800000B very VERY often, every 2 or 3 minutes, why ?
Is it a RAM issue or something ?
( i have a 512 Ko ram expansion )
is there any solution ?

Thanks

Regards.

[clenched]

Just a theory on my part: Maybe Workbench is trying to invoke some function your custom chips aren't capable of. For example, I have A500 with rev5 board that came with the OEM 8370 Agnus, so it lacked PAL capabilities in screenmodes. I suspect that since you booted ok, the ROM installation is not at fault. Seems like the package should state upfront all the machine requirements.

[Jope]

The WB detects the old agnus+denise, the additional screen modes will not be available.

I guess it'd be best to try it without the ram expansion, see if it keeps on crashing.

[yaqube]

Originally Posted by grommodore

I removed the old 1.3 kickstart and replaced it with the 3.1 ( i linked the pins 1 and 31 on the socket and the bend up 31 and 21 on the ROM , because i have a REV 5 motherboard )

I don't know if the description of what you have done is correct.

On rev 5 motherboard you should connect the pin #1 to the pin #31 on the ROM socket which you have done.
You should bend up only the pin #31 of the ROM chip and connect it to the pin #21 (which is not bent up).


Some technical info:

On any rev 5 motherboard (with 40-pin ROM socket) the pin #1 is not connected but 512 KB ROM chips use this pin as the most significant address line (ROM chip A17) which should be connected to the A18 line of the Amiga CPU address bus. This signal is present at the pin #31 of the ROM socket.

The 512 KB ROM chips (and 27C400 EEPROMs which most likely this 3.1 ROM chip is) use pin #31 as _BYTE/WORD mode selection line. It should be permanently high and that's why we should connect it to the pin #21 which is 5V power line.

[Jope]

I don't understand the pin31 issue, the 256kB rom also runs in word mode..

When I got 2.04 for my rev5 a500 as a kid (around 1992 or 1993), I just soldered a wire between pin 1 and 31 and plugged it in.

[yaqube]

Originally Posted by Jope

I don't understand the pin31 issue, the 256kB rom also runs in word mode..

When I got 2.04 for my rev5 a500 as a kid (around 1992 or 1993), I just soldered a wire between pin 1 and 31 and plugged it in.

When I read my 2.04 ROM in an EPROM programmer (I set the device type as 27C400) it reported a warning that there was probably no connection between the device pin #31 and the socket. It seems that the mask ROMs have this pin not used and always work in 16-bit mode. That's why you hadn't to connect that pin to VCC and could leave it in the socket.

But most 3.1 ROMs for A500/A600/A2000 (all I have seen) are just programmed EPROMs and they have the pin #31 functional as byte/word mode selection. That's why you have to connect it to VCC.

[Doobrey]

Originally Posted by yaqube

I don't know if the description of what you have done is correct.

Maybe he just described his mod badly?
Surely if it wasn't done right, he'd get either the red screen caused by a bad ROM checksum (address bus not fixed, but at least that'd show the ROM was working in word mode) , or just a black screen and constant rebooting ?

IIRC, #8000000B is the Line-F error, either the OP is running code compiled to use an FPU on a machine without one, or running buggy programs that love to jump into random memory, or as others have suspected it could be dodgy RAM corrupting code.

Originally Posted by yaqube

I don't know if the description of what you have done is correct.

On rev 5 motherboard you should connect the pin #1 to the pin #31 on the ROM socket which you have done.
You should bend up only the pin #31 of the ROM chip and connect it to the pin #21 (which is not bent up).
.

That's exactly what i did

I tried whitout the RAM exp, same problem.

Here's a video of my WB :

http://www.dailymotion.com/user/kami...cra_videogames

Thanks !

[Toni Wilen]

Originally Posted by grommodore

The workbench launches

1.3? Or did you update to 3.x?

Most likely 2.0/3.x+ incompatible program in your startup-sequence. (if not hardware problem)

[Jope]

Originally Posted by yaqube

But most 3.1 ROMs for A500/A600/A2000 (all I have seen) are just programmed EEPROMs and they have the pin #31 functional as byte/word mode selection. That's why you have to connect it to VCC.

Ah, that is indeed a valid point. :-)

[yaqube]

@grommodore

I can see in the video that you have an external drive. Does the same happen when you try to use the internal one? Or is your internal drive not functional?

Does it fail when the drive is being accessed? Or maybe moving a window on the workbench screen can also cause a failure?

I can only think about degraded capacitors on the motherboard and in the power supply. My A500 is now waiting for new caps, especially the power supply. I had similar issues when an FPU was installed on the turbo board and the system was powered from a stock power supply. Replacing the power supply with an ATX one or removing the FPU has cured the problem.

Both of them are working properly, it crashes wherever i insert the Workbench 3.1 floppy disk, the capacitors issues is perhaps the way...

thanks

[Photon]

I'd say power failure, ram failure, or rom failure. Gurus usually come from executing random bits (that should have been code), or jumping to the wrong place in code.

Disconnect all add-ons (lessen power consumption). Run a ram check on the memory you have in the system. If it still gurus, verify the ROM contents and its connection to the motherboard.

[Don_Adan]

Originally Posted by grommodore

Both of them are working properly, it crashes wherever i insert the Workbench 3.1 floppy disk, the capacitors issues is perhaps the way...

thanks

For me it looks like programmers bug (perhaps 68020 command), better load Workbench 2.0 and check effects, if it works you can try to replace copy of Workbench 2.0 disk with some parts from Workbench 3.1. For example as first replace all commands from C: dir, later Libs: etc.

[oissery77]

Hi all,
I have same type of issues with my A500 rev6a from 1988, i replace my KS 2.04 by a KS3.1 rev 40.63 it boots ok. But after several minutes the system can freeze or have guru errors (with floppy or CF). If i softkick the same KS all is perfect.
I was not having issues with KS 2.04.
Maybe it's my PSU/A500 caps?
Thanks for your help !

[voyager_1701e]

This may have nothing to do with your problem....but I had huge problems getting a 3.1 ROM to work on a Rev6a board. I found a reference to missing resistor network RP107. I found my board was missing RP107....so I fitted it and suddenly the system works perfectly.

My issue wasnt the guru...but black screen. Does your board have RP107 fitted?

[oissery77]

i've just checked and it don't have RP106 RP107 fitted but without it the KS 3.1 is recognised, it's not the same behaviour but thanks it's maybe a clue.