PDA

View Full Version : XT-Bios



k2x4b524[
July 23rd, 2010, 05:19 PM
Question, Which socket between u-18 and u-19 is the even and odd bits?

Chuck(G)
July 23rd, 2010, 05:27 PM
Trick question?

I assume you mean odd and even bytes...

The XT has an 8-bit bus; unlike the AT (16 bit), memory contents, bytes at both even and odd addresses are in the same ROM.

k2x4b524[
July 23rd, 2010, 05:33 PM
Yes, i have an upgraded XT xtyle bios, But the directions say the XT uses even and odd bits. I don't care to keep basic as i NEVER use it, i have GW-Basic anyway, so does that mean i only need 1 27256 Eprom chip for the bios or do i need the even and odd

Chuck(G)
July 23rd, 2010, 05:47 PM
Actually, on an XT, the entire BIOS will fit neatly in a single 2764. No even or odd. To figure out what you need, try pulling one ROM and see if the machine still boots.

Again, no evens or odds on an XT; straight address ranges. The naked BIOS lies between FE000 and FFFFF.

k2x4b524[
July 23rd, 2010, 05:50 PM
ok, i'll try that, and if it DOES boot, i can put the straight bios in one socket, then i've got the second open socket, since i don't care about basic, i can pretty much slap anything in there, like say, a bios for a TMC-850 that has no bios socket?

Anonymous Coward
July 23rd, 2010, 06:31 PM
Is it possible your XT board uses an 8086? 8086 has a 16-bit path to memory. My 8086 board has both even and odd BIOS chips.

k2x4b524[
July 23rd, 2010, 06:36 PM
Mine is a standard XT with a microsoft Mach 10 inside it, making it an 8086

Chuck(G)
July 23rd, 2010, 07:05 PM
Yes, but as I understand the Mach 10, it uses the host system BIOS (it will cache it in RAM for speed if you set the appropriate jumper). So my answer still holds, as access to memory off the Mach 10 board is still 8 bit.

Chuck(G)
July 23rd, 2010, 07:10 PM
ok, i'll try that, and if it DOES boot, i can put the straight bios in one socket, then i've got the second open socket, since i don't care about basic, i can pretty much slap anything in there, like say, a bios for a TMC-850 that has no bios socket?

Er, probably not with the TMC-850. The controller I/O space is memory-mapped, not port mapped. What this means is that the BIOS routines will deduce the card I/O space from the location of the BIOS ROM, which, of course, won't be right.

Maybe I'm wrong, but that's the way I read that controller.

k2x4b524[
July 23rd, 2010, 07:35 PM
That's all good. Ive ran into a problem with my burner, that or i'm not using it right, or i'm not letting my chips bake long enough in the erasure light, i'm doing it for about 10 minutes. The rom burner REFUSES to program a single chip, it reports all blanks, even with NO chips in the sockets.

It's a Sunshine EW-904B, uses an 8-bit card to interface with the burner. Thing is i have SEVERAL versions of the program, all of them wont burn, but they will report blanks.

Am i just simply NOT leaving them in under the light long enough? too long? I should have had several out and ready by now..

*Update* It may be the card is speed sensitive and my Mach-10 is just a hair too fast for it

Chuck(G)
July 23rd, 2010, 09:16 PM
*Update* It may be the card is speed sensitive and my Mach-10 is just a hair too fast for it

That's probably the case. I've got a Sunshine (the 4-socket gang programmer) and it won't run on anything but a 4.77 or 8 MHz real 8088.

k2x4b524[
July 23rd, 2010, 09:24 PM
*update* I got ANOTHER chip to program... It must be the duration inside the enclosure that isn't long enough. I figured 10 minutes would be sufficient, strangly though, the chip that did program, is the scrummiest, grimiest one, with half the label adhesive still on the window.. How long do you put yours in for? And are they SUPPOSED to make noise and get warm when they are programmed?

Youve probably got the same one i do, what version of the software do you use?

MikeS
July 24th, 2010, 07:54 PM
*update* I got ANOTHER chip to program... It must be the duration inside the enclosure that isn't long enough. I figured 10 minutes would be sufficient, strangly though, the chip that did program, is the scrummiest, grimiest one, with half the label adhesive still on the window.. How long do you put yours in for? And are they SUPPOSED to make noise and get warm when they are programmed?

Youve probably got the same one i do, what version of the software do you use?Umm, XTs with odd/even BIOS chips and EPROMs that make noise... are you smoking some illicit substance by any chance?
;-)

What kind of noise are the EPROMs making? Are they maybe talking to you, telling you that you're spending too much time with your old computers?

(Sorry, just being silly; no offense)

:rastarolleye:

k2x4b524[
July 24th, 2010, 08:07 PM
LOL, the even and odd blurb came from a readme.txt in a generic XT bios i found, and the noise the chip made sounded like electronic tinkling, so i think it may have been the chip getting programmed, it started when i started the programming process and stopped after the writing was done, and the chip got warm too. so i think that's the sound of eproms, i wish i had something seisitive enough to capture it too :(

and as for spending too much time with the computers, not really, i've been reading some interesting articles on ancient ruins, machu pichu, the bimini road, mahalipuram, stuff like that.. Good reads it is

Chuck(G)
July 24th, 2010, 09:44 PM
You know, that Sunshine is an old programmer. Which can be handy if you have old EPROMs to program. But maybe not so good in other respects.

One way EPROMs are different from other chips is that a bunch of chips may all be labeled 27128, but each manufacturer calls out the programming parameters for its own chips. Some may program at 25 volts; others may use 12.5. In addition, the algorithms used for programming may be somewhat different (e.g. different address sequences, pulse lengths, etc.). Sometime around the time of the first Seeq 27128s, something called "Silicon Signature", enabling a programmer to read a device ID from the EPROM electronically, thereby making programming foolproof. (This was Seeq's patent 44501903, whose eventual history showed that enough high-powered legal suits can break any patent, no matter how clear the patent language). But the Sunshine is ignorant of this little device and so it's up to you to make the proper guesses.

Fortunately, most older EPROMs do call out the programming voltage somewhere on the part; e.g. "PGM 12.5V". So you need to set the programmer to use this programming voltage. 25V will toast a 12.5V EPROM.

I'm not saying that this is what's happening; it's just to make you aware of something that might be going on.

k2x4b524[
July 25th, 2010, 05:27 PM
Problem Solved Through PM

Thread Can Be Closed.

MikeS
July 25th, 2010, 07:49 PM
Problem Solved Through PM

Thread Can Be Closed.Well, what was it? Inquiring minds want to know!

k2x4b524[
July 26th, 2010, 03:14 PM
Duration = 20 minutes
Noise = VPP doing it's thing to program the chip