here is another my experiment

it's not uefi, but works

This would replace your tool:

Could you post some instructive about how to test this on our homes?

[edit]Changed "store byte -sum and $FF at $$+5" to "store byte (-sum) and $FF at $$+5"

thanks, your calculation can be useful. i've been thinking about this, but supposed it can be executed too long for, say, 64kb.

i made a board of an old ISA netcard. i made it "universal" - to support ROM from 2716 to 27512 (but have test only with 2716 yet)






i think, it is possible to use any netcard with ROM socket, just you need to know what kind of ROM they support (i met 2764). also once i met multycard with ROM socket (a combined ISA card with IDE, Floppy, LPT etc.), but only once... if you wish to use an old motherboard as "embedded" system - rom on any vga card can be replaced with your

of course, you have to have appropriate EPROM and tool to burn program into it!

well... it's a wide field for hobby

regards!

Oh cool, the old ISA connector. I never thought I would be seeing one of those again.

BTW: Very neat work with the wiring, good job.

I've just tried with a 127.5 KB binary and the speed was below the measurable threshold (no time was given by FASMW). BTW, you'll have to change "store byte -sum and $FF at $$+5" to "store byte (-sum) and $FF at $$+5" because the former is interpreted as "store byte -(sum and $FF) at $$+5" (not sure if this behavior is backed by the documentation or not). I realised this because when I've tried the 127.5 KB ROM a "value out of range" error was triggered at that line.

I have a S3 Virge PCI with socketed Diamond ROM (if I remove it the card still works but a different ROM message is shown and the brand Diamond is not found anywhere), I suppose I can use it too? I'm not sure how to figure out the type of EEPROM I have to buy and I also have the problem that I don't have an EEPROM writer

Thanks for posting the instructive, I have some of these cards and an old motherboard with an AMD 368 embedded so I can test on that too (after solving the problem commented above of course).

[edit2]Chanced time to type on second paragraph

i like it.

then if i do it, i can lanch any code from a 27C512?
cool.

i have all the tools to do so, just maybe a lack of machine as the old ones gave up.


have you got the shematics of your modified board please? just to try to do it industrially.

@Loko
where you tried? i still use P-24 100MHz at mine garage (btw, this utility was exactly written at it )

yes, this board was made not to write stupid messages before boot. it let to use an old motherboard as diskless system, possibly also video-less - just program and IO ports - own-made or standart LPT

Binary "-" is stated as lower precedence than "and" but unary "-" is left without definition in fasm.pdf. So it looks like your test show us that unary "-" is also lower precedence than "and".

i'm reminding related project: http://board.flatassembler.net/topic.php?t=8189

My post was incomplete I pressed Send instead of Preview, two paragraphs were added on the edition.

shoorick, an Athlon 64 2.0 GHz But does it really take too much time even on hardware like 486SX 33 MHz?

you can buy it for tens of dollars only ( billions of $ZBW), or make it for less.

i can mail you a programmed UVPROM if you want.

Jesus! time?? I wanted to say "the type" not "the time"

Actually if I remember right the ROM that came with the card already was an UVPROM because when I removed the tag there was a glass that I covered immediately. The card still works as a "Diamond Stealth" so the short and weak exposure to the UV rays wasn't enough to erase. I removed the tag around five years ago some I may in err (and the fact that there was a glass that allowed me to see inside the chip is enough to proof that the chip is an UVPROM?).

Thanks edfed, but the mail service is one of the services that I trust the less and there is no money return for the loss. A certificated latter is more likely that will reach my hands but are so expensive that I think that better I buy the equipment myself here.

sure, it is not a matter how many seconds it will compile for real project but i noticed, that bios_sum in first version, when it was read file byte-by-byte, time was noticable on P 100MHz, thus i made it 512 byte block-by-block reading. and it is win32 application, last time i have run win95 on 486dx40 was at 2002 under dos it can be faster possible winasm is loading 60 seconds on P100, but i'm waiting, as it will save me time while i'm working

here is schematic and mask for PROM used as selector. sure, it even can be simplified (replace 7404 element with spare element from 7400). A16-A19 selector selects "D" hex value, so, used range starts at 0D0000h. other range selected by PROM selector depending on selected EPROM type.

ok!
thank you.
maybe a set of eproms can be selected using the selector.
hem, great!

forgot to mention: OE is connected to ISA MEMR pin

btw, if FLASH ROM will be used, it can be more convenient, as it is possible to provide in-system programming, and FLASH ROM persistent for much more cycles instead old UV EPROMs.

minimizing soldering
here is examples of usage single obvious chip 74LS138 as ROM selector:



it is applicable for 27512, 27256 and 27128 types, for roms of less size additional chip will be needed.

but in general, it is possible to use any of these schemes for rom chips of size below 32kb, just it is necesary to know, that image will be "duplicated" some times in selected range. to avoid many loads of same image it is possible to prevent it programmatically, say, like this:

-----------------------------------------
hope, nobody has build this yet
swap A16 and A18 lines at diagram!

i have made another card - more simple. you can see it here with high resolution images

regards!

I'm just wondering if its easier to draw and scan an image or is it better to draw it in MS Paint

But the project itself is worth noting and I sometimes love myself getting to the nuts and bolts of things.