• Please review our updated Terms and Rules here

HeadStart Explorer and XT-IDE v1.0

Maybe that's a problem with the script, then.
Have you tried using debug to manually call an INT 13H service, like 08H Read Drive Parameters ?
patscc
 
nope, the script is fine. I can execute the same script on my machines and they work. in that thread, chuck and modem7 suggests that it may be an inability to DMA out the data to the file, so a modification to the script was put in place to work around that. Have not heard back if it was successful.

INT13 for HDD support is not installed because the BIOS never ran, as there is no sign-on/boot menu.

The only things I can think of that would do this are a corrupt rom or a mainboard BIOS that refuses to load option roms.

Putting a different card with an option ROM in the headstart machine would tell us if the mainboard is capable of running external ROMs.
Putting the XTIDE in a different machine would tell us if the ROM on the XTIDE is corrupt or not.
 
You should be able to use debug to manually 'point' the INT 13H handler to the routine in the BIOS. Just because the script works on machine x doesn't necessarily mean it'll run on machine y.
If you can execute debug -d d000:0 l 8192 and get a dump of the EPROM, then I'd say there's a problem with the script.
patscc
 
Well, I had some tome to experiment with the Explorer... I tried some other settings and a normal ATA harddisk (527MB IBM one) and that wouldn't work either. FDISK keeps reporting "no fixed disk present" :-(

So, I took the controller and the harddrive out of a HeadStart II and put them in the Explorer. Boots without a problem, starts DOS and I am able to play Sopwith ;-) Never changed anything on that controller, just took it out of the HeadStart II and put it into the Explorer (it sat around so long I actually thought the disk would be dead). I don't know what controller this is, but the harddisk is a Seagate ST-238R (RLL right? So not an ST-225 as I thought at first).

So that means I can at least use a third party controller with RLL disk in the Explorer... so it DOES call external ROMs??

I've tried some other settings on that XT-IDE v1 controller but that would only end up the BIOS being "in the way" of some other ROM (I think), because either the machine would hang, or it would start but without video (I think that was the C8000 setting). Also what is weird: WITHOUT any drive in it, this thing boots to DOS 3.31. Not a complete DOS but I have some files (some with .CMD extension, others with W8 extension or something). So I think there's a ROM in there somewhere with DOS 3.31 (like the IBM Cassette BASIC on the PC/XT/AT and PS/2).

So uhm... any ideas? Dunno what to think of this, since I expected the RLL controller/disk also to not be recognised at all TBH. Confused here and somehow I think this means I should be able to get that XT-IDE v1 controller working in the Explorer?
 
It's great that it's running. It does seem to be calling extension rom. It might be a problem with where your onboard DOS-thing is sitting. Can you list a few more of the files ?
patscc
 
So that means I can at least use a third party controller with RLL disk in the Explorer... so it DOES call external ROMs??
Yes. This is good news. So we have either a corrupt rom, a ROM conflict, or some other weird thing (eeprom speed?) which is keeping the headstart from liking the ROM.

I've tried some other settings on that XT-IDE v1 controller but that would only end up the BIOS being "in the way" of some other ROM (I think), because either the machine would hang, or it would start but without video (I think that was the C8000 setting).
Do you know where the ROM is located in memory on the RLL controller? The ROM on that card is probably 4k in size. the one on the XTIDE is 8k, so it is possible that there is something in the headstart in the same range as the XTIDE, perhaps clobbering the 2nd half of the XTIDE ROM and keeping it from checksumming properly during POST.
There are 16 possible settings for where the XTIDE can live, you've tried a few of them. You can either try all of them one at a time, or poke around in debug to find a free space that is big enough and then set the card there. I seem to recall microsoft also having a utility which can show you free spaces. (MSD.exe or something?) which you could also try.
 
Well, while poking around in the ROM addresses I did sometimes see other stuff. I remember one being "MiniScribe wrong interleave set for drive $" as that struck me as strange. How do I find 8K of free space? When checking some addresses sometimes I only saw dots with the occasional < in between.

I'm not that familiar with debug. BTW: would this mean that the default address of 300h is OK, and only the ROM address needs to be changed?
 
Well, while poking around in the ROM addresses I did sometimes see other stuff. I remember one being "MiniScribe wrong interleave set for drive $" as that struck me as strange.
that's fine, that's just an error message that could come up. It doesn't mean it will come up. that message is also not inside the XTIDE BIOS.

How do I find 8K of free space?
the MSD.exe program I was referring to is the easiest, otherwise repeated D(umping) in debug can get you there eventually.
Starting at
-d d000:0
<screen fills with data>
-d
<more data>
...etc
until you've dumped up to D000:2000 (that's 8k)
If you don't see anything other than dots and the occasional <, that area should be free. If you see garbage and occasional text messages, that area has a ROM in it already and is not free.

you can do the same thing at C800:0 --> c800:2000 or any other address you want to explore. You've got 16 potential areas to check, so get busy! :)

BTW: would this mean that the default address of 300h is OK, and only the ROM address needs to be changed?
yes. Do not change the default address of 300h (yet). If you do, you have to reflash the BIOS. We have to get the BIOS to show up on the screen before we can see if 300h IO space is also taken up. that's the next problem we will get to deal with. ;)
 
Something that I created today to aid in the diagnosis of ROM related issues in the XT-IDE:

In [this] ZIP file are two binary files for the XT-IDE card's ROM.
3TEST_S.BIN: Small - 512 bytes
3TEST_L.BIN: Large - 8 KB (the normal size for code in the XT-IDE V1)

The code within both does something simple. It displays a "3" character and then halts.
For example, flashing 3TEST_S.BIN into the XT-IDE V1 within my XT clone results in the screen that follows.

3test_photo.jpg


Above, the BIOS expansion ROM in my hard drive controller card was executed first and displayed the first line of text.
Then, the BIOS expansion ROM in my floppy drive controller card was executed and displayed the next two lines of text.
Then, the BIOS expansion ROM (containing 3TEST_S.BIN) in my XT-IDE card was executed, displayed the "3", and then halted the CPU.

So, in using 3TEST_S.BIN, the appearance of the "3" informs me that:
1. The computer does call BIOS expansion ROMs, and
2. Writing and reading of the first 512 bytes of the ROM is working.

And if required, visual comparison between the ROM contents (using DEBUG) and the 3TEST_S.BIN contents doesn't take long, because the code is only 512 bytes long.

If use of 3TEST_S.BIN results in the "3" being seen, then the next step would be see if the long version, the 8 KB sized 3TEST_L.BIN, also displays a "3".

Important

Once 3TEST is flashed into the XT-IDE's ROM, re-flashing the ROM requires a particular procedure:

1. Remove the ROM-enable jumper on the XT-IDE.
2. Power up computer.
3. Whilst the computer is running, replace the ROM-enable jumper.
4. Flash ROM.
 
Well, for now I've tested every memory address setting and nothing comes up :-( It just boots directly into the DOS ROM or a floppy disk.
 
Well, for now I've tested every memory address setting and nothing comes up :-( It just boots directly into the DOS ROM or a floppy disk.
It will be interesting to see how your use of 3TEST_S.BIN (per post #34) works out, should you decide to give it a try.

After flashing 3TEST_S.BIN into your card, expect DEBUG to show the first 512 bytes of the ROM content to be:

55AA 01EB 0300 0000 B40E B033 B700 CD10 F400 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 00
E5
 
Well, I tried to flash the EEPROM today with a downloaded V1.1.5 BIOS (from the google code page). Created a 720K bootable DOS 5 floppy disk, copied XTIDECFG and the BIN file (for XT) onto it. Removed the JP1 jumper, and put it back on while the machine was on (to erase the ROM??). Then rebooted off the floppy, checked with debug that there's nothing on D0000 address, then ran XTIDECFG, loaded the BIN and tried flashing the BIOS. It starts flashing and almost immediately comes back with some "Timeout while polling EEPROM! EEPROM not succesfully flashed!". I kept the settings at default in the menu, so SDP enabled, EEPROM 2864, address e000h, Page Size = 1, Generate report YES. The card is currently set for 300h/e000-e1ff...

Weird? Or am I doing something wrong?
 
And at address E000 I find something with "ROMDOS 1.0" in it's description... probably the DOS 3.31 ROM? So this probably interferes with flashing?
 
OK, been able to flash the BIOS at address CC00 with success. After a reboot still no menu but I can see the BIOS again with debug (d CC00:0).

Any ideas? I should think a successful flash at CC00 should mean the BIOS should be available and called at boot?
 
Compare the ROM to the bin file. But most likely the board BIOS isn't looking for ROM extensions.
 
Back
Top