Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: amigakit on February 16, 2007, 11:16:16 PM
-
Further to the discussion here the other day- we have just received a quantity of new v40.70 Kickstart 3.1 mask roms manufactured by Amiga:
http://amigakit.leamancomputing.com/catalog/product_info.php?products_id=476
-
These ROMs are only use for A4000 T.
-
I think someone (keropi?) has tried them succesfully even in A4000D. Of course there might be issues with IDE if mb one is used, or with workbench.library if one is not present in libs:, but otherwise they might work with A4000D. For only A4000T owners the market for these would be extremely small - weren't they all sold with KS3.1 already installed?
-
For the A4000T there would be no market, they all came with 3.1 Kickstart. Whats the point?
-
What if your roms went bad? Wouldn't that be of any use?
-
How often do Kickstart ROMs go bad? I had it happen to me once in 18 years, a set of Amiga 3000 Kickstart 1.4 Boor ROMs, I wasn't upset. LOL
-
The only different between 40.70 and 40.68 is the missing workbench.library and builtin ncrscsi driver. So you can use it of course also in an a4000 desktop if you modifiy your install disks and add the missing library.
-
£28 quid... what is the world coming to?
You can almost buy a pair of 27C400 chips and a budget EPROM programmer for that :-)
-
I had an A4000D that had 40.70 roms, from C.
it had workbench library in, and just the scsi.device... but it was a little picky on the hd+cd combo it needed to work, that is why C went back to 40.68 as I found out later.... so these roms might not be for T machines...?
-
alexh wrote:
£28 quid... what is the world coming to?
You can almost buy a pair of 27C400 chips and a budget EPROM programmer for that :-)
I can easily get 27C400 chips on eBay. However, I wanted to buy a good EPROM programmer to program these but all newer programmer's don't support these chips.
Can you suggest a good programmer? Please note that I would prefer NOT to use a Willem as these are rather low cost and their reliability is sometimes questionnable.
Thank you.
-
If I understood this post correctly, it looks like there might have been 2 different Kickstart 3.1 40.70 versions:
THE FIRST is Kickstart 3.1 40.70 ROMs for A4000T released in 1994:
These A4000T Kickstart 3.1 40.70 ROMs have scsi.device replaced with NCR scsi.device and workbench.library has been removed and added to Workbench 3.1 A4000T floppy to make room for NCR scsi.device code. NCR scsi.device is needed for A4000T onBoard SCSI.
THE SECOND is Kickstart 3.1 40.70 ROMs for A4000T or A4000D produced somewhere in 1995 and 1996:
These A4000T Kickstart 3.1 40.70 ROMs are reported to have workbench.library and normal scsi.device, just like in previous 40.68, however Commodore/Escom/whatever noticed it was a little picky on harddisk+CD-drive combinations and was reported to cause problems with some harddrives and accelerators so after producing some they gave up and switched back to 40.68 to avoid those problems.
Assuming because of the bankrupt the developers responsible of Kickstart coding stopped working in a situation where they could soon solved those issues. So we may never find out if there is just a little typo in 3.1 40.70 code that could be corrected by some skilled hobbyist. Or who can say if using 40.70 in A4000D really just got a bad reputation mainly because of bad design in some old harddisks, CD-drives or accelerators or the well known bugs A4000D Buster version 7/9 or even, because A4000T has Buster 11 and it had they decided to continue using 40.70 there. Ofcourse no software or hardware is 100% free of design errors, so what I'm saying is that there must be bugs in all Kickstart versions but there are no proper documentation about how likely it is that all these problems were caused just because of A4000D with 40.70.
It is also a shame they didn't move to use bigger capacity PROMs, so removing something like workbench.library to fit in new code would not be necessary. As programmers know when something new is added to the code something old may break and after adding something new the code would have needed full testing if every old feature works.
Anyway I think the Amigakit product code: AMIROM007 with the following chips...
MX J9648
391657-01
V3.1 (V40.70)
(C) AMIGA 1995
42747JY9
MX J9648
391658-01
V3.1 (V40.70)
(C) AMIGA 1995
42706JY9
...could be one of those second versions. At least the fact that they are manufactured 1996 on week 48 supports this theory. Or why would they produce any Kickstart 3.1 ROMs for A4000T at that date, since all A4000T were indeed shipped with Kickstart 3.1 40.70 anyway and while in the edge of the bankrupt I don't believe that they would dare to produce these chips just for spares just for A4000T models that were quite rare in the first place.
For A4000D IDE normal scsi.device is needed and for A4000T NCR scsi.device is needed and both are used by system with the same name scsi.device. So how could any Kickstart work in both A4000T and A4000D models when only one scsi.device can exist in one Kickstart?
Sources:
http://amigakit.leamancomputing.com/catalog/product_info.php?products_id=476
http://www.gregdonner.org/workbench/wb_31.html
http://wiki.classicamiga.com/Kickstart_Roms_Explained
http://www.amiga.org/forums/showthread.php?p=302371
http://eab.abime.net/showthread.php?p=260038
-
Please note that I would prefer NOT to use a Willem as these are rather low cost and their reliability is sometimes questionnable.
Thank you.
Willem are not bad at all, I use older version of it (parallel port, iirc its 3.x).
what you should be worried is quality of second hand 27c400... those EPROMs turn out to be some kind of analog devices. they have limited amount of write/erase cycles. and erasing for too long time can kill them.
http://wiki.xtronics.com/index.php/How_EPROMS_Work#Why_do_I_need_an_EPROM_eraser.2C_can.27t_I_leave_them_out_in_the_sun.3F
-
If I understood this post correctly, it looks like there might have been 2 different Kickstart 3.1 40.70 versions:
THE FIRST is Kickstart 3.1 40.70 ROMs for A4000T released in 1994:
These A4000T Kickstart 3.1 40.70 ROMs have scsi.device replaced with NCR scsi.device and workbench.library has been removed and added to Workbench 3.1 A4000T floppy to make room for NCR scsi.device code. NCR scsi.device is needed for A4000T onBoard SCSI.
THE SECOND is Kickstart 3.1 40.70 ROMs for A4000T or A4000D produced somewhere in 1995 and 1996:
These A4000T Kickstart 3.1 40.70 ROMs are reported to have workbench.library and normal scsi.device, just like in previous 40.68, however Commodore/Escom/whatever noticed it was a little picky on harddisk+CD-drive combinations and was reported to cause problems with some harddrives and accelerators so after producing some they gave up and switched back to 40.68 to avoid those problems.
Assuming because of the bankrupt the developers responsible of Kickstart coding stopped working in a situation where they could soon solved those issues. So we may never find out if there is just a little typo in 3.1 40.70 code that could be corrected by some skilled hobbyist. Or who can say if using 40.70 in A4000D really just got a bad reputation mainly because of bad design in some old harddisks, CD-drives or accelerators or the well known bugs A4000D Buster version 7/9 or even, because A4000T has Buster 11 and it had they decided to continue using 40.70 there. Ofcourse no software or hardware is 100% free of design errors, so what I'm saying is that there must be bugs in all Kickstart versions but there are no proper documentation about how likely it is that all these problems were caused just because of A4000D with 40.70.
It is also a shame they didn't move to use bigger capacity PROMs, so removing something like workbench.library to fit in new code would not be necessary. As programmers know when something new is added to the code something old may break and after adding something new the code would have needed full testing if every old feature works.
Anyway I think the Amigakit product code: AMIROM007 with the following chips...
MX J9648
391657-01
V3.1 (V40.70)
(C) AMIGA 1995
42747JY9
MX J9648
391658-01
V3.1 (V40.70)
(C) AMIGA 1995
42706JY9
...could be one of those second versions. At least the fact that they are manufactured 1996 on week 48 supports this theory. Or why would they produce any Kickstart 3.1 ROMs for A4000T at that date, since all A4000T were indeed shipped with Kickstart 3.1 40.70 anyway and while in the edge of the bankrupt I don't believe that they would dare to produce these chips just for spares just for A4000T models that were quite rare in the first place.
For A4000D IDE normal scsi.device is needed and for A4000T NCR scsi.device is needed and both are used by system with the same name scsi.device. So how could any Kickstart work in both A4000T and A4000D models when only one scsi.device can exist in one Kickstart?
Sources:
http://amigakit.leamancomputing.com/catalog/product_info.php?products_id=476
http://www.gregdonner.org/workbench/wb_31.html
http://wiki.classicamiga.com/Kickstart_Roms_Explained
http://www.amiga.org/forums/showthread.php?p=302371
http://eab.abime.net/showthread.php?p=260038
They are the one's without the workbench library.
They are exactly the same as 40.68 in all other respects.
If these are used in a Desktop A4000, then you need to put the workbench.library on the HDD in the libs folder. If you dont then it will hang on the Workbench screen.
-
Assuming because of the bankrupt the developers responsible of Kickstart coding stopped working in a situation where they could soon solved those issues ...It is also a shame they didn't move to use bigger capacity PROMs, ...
These are some of the reasons I'm contributing bucks to the AROS Kickstart Replacement projects - 15 years is long enough since an update. I want CDROM boot from Kickstart!
-
40.68 and 40.70 exists for both A4000 version (desktop & tower)
difference between 40.68 & 40.70 is a different version from the scsi.device
(40.70 scsi.device where reported to make more problems with some drives, thats the reason for shipping 40.68 instead 40.70)
difference between "tower" & "desktop" is the missing workbench.library (replaced by wb.find and library outsourced to disk) and the additional ncr_scsi.device for the scsi.controller