View Single Post
Old 23 September 2006, 13:49   #40
Sandgunner
Kill everything in a 360
 
Sandgunner's Avatar
 
Join Date: Sep 2006
Location: Near West Chester PA. Made the pilgrimage to the former Amiga assembly plant.
Posts: 23
The kickstart oddity extravaganza

[quote=Jope]Those eproms have both a byte mode and a word mode.

The explanation above is not 100% thorough, as I'm again feeling very lazy.. But at least you can get a rough idea what it's about.. If you have a 256kx4 part, then it's (256 * 4)/ 8 to get the amount of bytes it stores.

Thank you for the EPROM 101. It's been a looooong time heehee

Here is where I am at. The chips are indeed 512K
40000h X 16bit

Now here is where I'm going nuts.
Inserted a working 2.04 kickstart rom in the programmer (this is not
an elcheapo unit) select the exact device type in the dropdown
menus. In this case the 2.04 kickstart is a Macronix 27C4100
Read the contents into the buffer.
Remove the 2.04 kickstart, insert AMD 27C400 (identical layout
part pin for pin and capacity)
Burn binary image to EPROM exactly as read from the source so no
byte swapping is required here as this is not a UAE or other image I'm
working with.

Take newly burned EPROM to my testbed, an A500 with a revision 6A
motherboard. This revision requires no jumpers or patches whatsoever
to work with these EPROMS.

Fire her up. I'm now getting the Kickstart insert screen. GREAT...almost
loading workbench or a self booting floppy like Lemmings causes a system
crash, flashing power light and reboot.

Comparison details. I take the supposedly bad EPROM that is crashing
back to the programmer and read it's contents saving it to a binary
file. I then re-insert the original working 2.04 kickstart and read its contents saving to a binary file.
I do a file comparison on the two. They are identical!

WTF?

Do note that during the burning process the buffer contents, original kickstart read and the new EPROM programming pass all verification checksums.

Here are my options in the programmer that "might" have something to
do with the failures at the device level when read in the actual Amiga.
The programmer has config options to do the following.
Gang split mode has the following config options
Byte, Even, Odd, Word0, Word1, Word2, Word3

I have used the Byte option and I get parlaysed A500 that
can boot directly from floppy but crashes without a floppy and
you can not actually run anything without crashing.

When burning I have the default buffer value set to FF as this
is its default state anyway.
In the buffer edit options I have the buffer set up as
8bit byte (hex), I also tried 16 bit word but this causes
only half the binary file to be written.

Anyone have any thoughts on what I'm doing wrong?
Key points. A500 rev 6A motherboard (no jumpers/wires required)
EPROMS used: AM27C400 256kb X 16
binary image used : taken from actual working 2.04 kickstart rom
read/write and verify all pass and match the checksums.
Read from burned roms matches identicaly to the original roms.
YES the orginal rom works perfectly in the A500.

I'm off to UV the eight EPROMS I have once again and try some
more experiments. That or go bald figuring this out.

Anyone here successfully burned a single chip kickstart for
an A500-A2000? If so what exactly was your option
setup for your programmer.

I greatly appreciate the help.
Sandgunner is offline  
 
Page generated in 0.07020 seconds with 11 queries