• Please review our updated Terms and Rules here

XT-IDE tech Support

I'm also interested in trying this version. I had a problem with my late model FDC controller. I couldn't boot off the A drive. But I could boot off the B: drive, but then the hard drive would be undetected. Maybe this fixed one or more of my issues as well. I could also take that opportunity to get you the ROM dump of that controller like you asked. I just needed a break to enjoy the system as working with the stock controller. I'm ready to dig back in soon!
 
I've done some quick testing and so far every 16-color Tandy games have worked with v0.11 BIOS. I'll do some more testing once i receive the CF-to-IDE adapters that i have ordered for Tandy.
 
Awesome news! Thanks for testing this.
Glad it appears to be working as expected, and is an improvement over the last version.

I just got sideswiped with a few other things that I need to take care of, so I might just archive up this version of 0.11, release it officially so there's no confusion or need to update again later, and just wait on 0.12 before doing the code cleanup that I wanted to do. No new functional changes were expected in the cleanup, so it was purely for readability anyway.

Can anyone kick off a few things in BASIC to make sure that didn't break?
 
XT-IDE Woes

XT-IDE Woes

Hargle - Agent Orange here -
Here's the rundown on the card for my 1000SX:
1. Installed card first in my 486 after the build - worked great -no problems.
2. Installed card in the 1000SX - appreared to boot but hung-up in the
config.sys, because the HD was DOS 6.20 on a 8-bit machine. (my mistake).
3. FDISK'd the Maxtor 7645AT (500 MB) and attempted to format for DOS 3.3 - The activity LED flashed but the processed failed. I would see "Format Unsuccessful".
4. About that time my 360 KB floppy failed.
5. I checked the card setup and found that I had JP-2 staked, or write enabled. I believe this contributed to corrupting the EPROM. Another mistake on my part. I set the thing up according to the VCWiki guide but didn't give a second thought to JP-2 (cranal-anal inversion on my part).
6. How about I ship the drive & card back to you for a look-see?
 
Agent Orange said:
2. Installed card in the 1000SX - appreared to boot but hung-up in the
config.sys, because the HD was DOS 6.20 on a 8-bit machine. (my mistake).
Click to expand...
dos 6.22 shouldn't be a problem on the 8bit machine. that's all I'm testing with. Granted, something in your config.sys might be trying to do 386 instructions, so I could see the hangup.

3. FDISK'd the Maxtor 7645AT (500 MB) and attempted to format for DOS 3.3 - The activity LED flashed but the processed failed. I would see "Format Unsuccessful".
Click to expand...
that may be the issue of too big a drive for old DOS that I still need to look into. Still seems that a 500mb drive should work, but we are talking about FAT-12 here, with a 32mb limit per partition... I need to check into that.

6. How about I ship the drive & card back to you for a look-see?
Click to expand...

Nah. Just flash it again, then write protect it. You'll want to try BIOS version 11 anyway. Eventually mine will fail and I'll look into it, but for the moment, this is kind of a non-issue, or at least not a show-stopper.
 
I realize that 6.2 will work okay - what I meant to say was that it stubbed its toe on the HIMEM.SYS line in the CONFIG.SYS, which will not work on a 8088.
 
Agent Orange said:
Can you give me the run down on flashing the EPROM?
Click to expand...

http://wiki.vintage-computer.com/index.php/XTIDE_project
Download the BIOS and the utilities package.
Put oprom.bin and flash.com onto a bootable disk
write enable your BIOS
boot to the disk
type flash
it'll ask you the filename, type oprom.bin
it should detect your card, ask you one last time if you want to flash, and then do it. after a reboot to verify that the new bios is installed, power off the machine and write protect the eeprom again.
 
I flashed my board with v .11 and it successfully fixed losing the c: drive when booting from a floppy. I have a late model floppy controller with 2 x 1.44m drives and one 360k drive. I still cannot successfully boot from the A: drive. I can, through the XT-IDE menu, boot B:, C:, or D: Once in DOS, I can access all drives. But just can't get it to boot off of A: The drive spins, but no loading. This is of course still a very workable system. I can boot off of a 1.44M, 360k, or the hard drive. Hargle, I dumped the ROM of my FDC per your suggestion and will email it to you tonight in case you want to take a look.
 
I think you're the first to have a 3 drive floppy system. I'm pretty happy that the boot menu appears to be picking up the 3rd floppy properly and displaying it. I have many, many revisions of the boot menu code where I try and determine how many floppy drives there are; the biggest problem I found was that a lot of machines say there are 2 drives even when there's 1, and sometimes even 1 drive when there are non at all!

Anyway, which drive is your A: drive? (1.44 or 360?) and are all of these drives attached to that single controller? Do you get any error messages when the A: drive is attempting to boot? You say it spins but doesn't load, then what happens? Does it come back and say "Press F1 to retry or ESC for boot menu"?

(and apologies for making you answer my questions both in email and in the forum, but the forum is better for tracking)
 
Anyway, which drive is your A: drive? (1.44 or 360?) and are all of these drives attached to that single controller? Do you get any error messages when the A: drive is attempting to boot? You say it spins but doesn't load, then what happens? Does it come back and say "Press F1 to retry or ESC for boot menu"?
Click to expand...

The drive spins, then I get "no boot device available" followed by "Press F1 to retry or ESC for boot menu"

FDC has two connectors, each supporting 2 drives. The two 1.44M are on the first connector and labled A:. B: THe 360k is by itself on the second connector. I have tried various combinations of the above and the drive assgned as a: is not bootable regardless of what drive it is.

I had a small breakthrough this morning. I can boot from the A: drive from a dead cold boot. No kind of warm boot does the trick, even though the FDC and XTIDE ALWAYS correctly identify the dirves. I can make it work by shutting down the PC, and waiting 15-30 seconds to simulate a cold boot. I have had reset problems that I discussed on another thread that may relavant, so let me summarize.

With a NEC v20 installed, I do not get a post from a cold boot. I had to flip the switch on and off to get normal post. With Chuck's help, I implemented a circuit which replaces th 5 volt power good signal to the motherboard and has an external pushbutton so I can use it. This is only a problem during cold boot. Ctrl-alt-del works fine. So I don't know if this is related, but didn't want to leave it out. I would think that you are expecting the XTIDE to function from a warm boot.

To clarify, my normal startup routine is: 1) turn on power switch (no post) 2) push the button on my home made reset circuit (get post and XTIDE works and can boot A: )

XTIDE doesnt boot A: with any kind of warm boot 1) ctrl-alt-del 2)hitting my home made reset button 3)turning the power off and on without waiting at least 15-30 seconds.

Confused? I am!
 
I had a small breakthrough this morning. I can boot from the A: drive from a dead cold boot. No kind of warm boot does the trick, even though the FDC and XTIDE ALWAYS correctly identify the dirves. I can make it work by shutting down the PC, and waiting 15-30 seconds to simulate a cold boot. I have had reset problems that I discussed on another thread that may relavant, so let me summarize.
Click to expand...

Very interesting. Here's an idea:
The FDC controller is checking the warm boot flag, and then not installing/hooking INT 13 services as a time saver, thus something doesn't get initialized/reset to a proper setting, and thus it doesn't work. Now that I have a dump of the BIOS, I can do a quick scan through it to see if it does stuff different on a cold/warm boot.

I am now blaming the FDC BIOS entirely for this issue. :)
I win!
 
I started testing with CF-cards. Tandy DOS 3.2 fdisk and format works with 64MB and 256MB CF cards, so drives at least up to 256MB works with DOS 3.2. I was not able to format any of the real hard disks that i tested with DOS 3.2 and they were all over 500MB. I wasn't able to format 512MB CF card just like i wasn't able to format 500MB hard disks.

I had problems with Apacer 64MB CF card (no speed rating). Tandy DOS 3.2 format froze at cylinder 488 and head 2. I expected some problems since those cheap slow cards don't work properly on my 286 either (drive not ready errors when writing). Slow CF cards simply cannot be used as hard disks.

Apacer Photo Steno Pro II (100x) 256MB works. I was able to format with Tandy DOS 3.2 and copy files to formatted drive. I was not able to boot from it. XTIDE drive detection reappeared and nothing happened after that.

Next card i tried was Apacer Photo Steno III (88x) 512MB. I was able to create full sized partition with DOS 3.31. Format went well but booting didn't work. Garbage appeared on the screen. Boot was successful after i formatted (same DOS 3.31 partition) with DOS 6.22. I tried again with DOS 3.31 but this time i created 256MB partition to the same 512MB CF card. Boot worked this time. I'd like to note that DOS 3.31 fdisk just froze when i previously tried it with 1GB real hard disk. Maybe DOS 3.31 just don't like big drives and partitions even though it should support them.

Tandy DOS 3.3 works well. No problems since it only supports partitions up to 32MB. Still, >=500MB drives and DOS 3 just doesn't seem to be good combination.
 
I was finally able to create bootable Tandy DOS 3.2 partition and other partitions as well!

It seems that Tandy DOS 3.2 partitions over 25MB are not bootable. This could also be some CHS calculation issue with Tandy DOS 3.2. Anyway, i used the 256MB CF card with 980 cylinders, 16 heads and 32 sectors per track. Partitions with 95 or more cylinders (on this CF card) are not bootable and partitions with 94 or less cylinders seem to boot just fine.

DOS 3.2 supports only single partition but Tandy DOS 3.2 supports more partitions with mlpart.com (for partition), mlpart.sys (driver) and mlformat.com (for formatting). Mlpart only supports three additional partitions per drive. This means that i can only use about 120MB from 256MB CF card.

Mlpart.sys takes 1088kB so that leaves 592320 bytes free after boot. Tandy DOS 3.3 does not need mlpart but it leaves 586816 bytes free after boot. There is no partition limit with Tandy DOS 3.3 (but last possible DOS drive is z).

So it is possible to use Tandy DOS 3.2 with XTIDE. I recommend fast 128MB CF card for it.
 
Where are you getting the smaller CF cards? I can only find them in much larger sizes. Must be missing something in my searches.

Thanks,
Kelly
 
I got my CF cards from a small Finnish web store. Unfortunately they don't appear to ship internationally. That 256MB 100x CF card cost 3.10 euros. It certainly seems that 256MB CF cards are hard to find, especially fast ones.

I have two new XTIDE problems. 512MB CF card doesn't work as slave (i tried 256MB card as master). XTIDE bios doesn't find slave drive at all. Both cards are found when 512MB card is master and 256MB card is slave. I also tried 256MB card as master and 64MB card as slave. XTIDE founds both of them.

Second problem is about booting from D-drive (512MB as master and 256MB as slave). When i select D from boot menu, booting starts but from wrong drive! I have Tandy DOS 3.3 on 512MB card and Tandy DOS 3.2 on 256MB card. Selecting D should boot to DOS 3.2 but instead DOS 3.3 is being loaded. It doesn't get very far and ends to Bad or missing Command Interpreter error message.
 
aitotat said:
I have two new XTIDE problems. 512MB CF card doesn't work as slave (i tried 256MB card as master). XTIDE bios doesn't find slave drive at all. Both cards are found when 512MB card is master and 256MB card is slave. I also tried 256MB card as master and 64MB card as slave. XTIDE founds both of them.
Click to expand...
weird. all you're doing is moving the card back and forth between the IDE-->CF adapters, and not moving the adapters themselves?
I have no idea what could cause that.

Second problem is about booting from D-drive (512MB as master and 256MB as slave). When i select D from boot menu, booting starts but from wrong drive!
Click to expand...

I wonder if something broke here when I fixed the floppy B: booting problem?
You could try going back to 0.10 BIOS and seeing if that works. I know I tried booting to D: in previous releases, but probably not before 0.11 was released.

BTW: I'm updating the wiki with some of these reports. Feel free to add your additional findings.
http://wiki.vintage-computer.com/index.php/XTIDE_TestResults
 
BTW: what happens when you remove a CF card from the IDE adapter when you're in DOS and it is the current, active drive?

There's no cache involved with any writes in the XTIDE, so any data written should be on the CF when you remove it, so I don't think there is a data corruption issue there, although DOS and certainly smartdrv type caches could cause problems.

I'd also think that since CF cards have no wear leveling, that the FAT table areas would wear out pretty quickly on the flash, so you'd almost want to make sure you had some cache installed, so the number of writes would be minimized; this may be a whole new thread discussion.

When you re-insert the CF, does it just pick up and go where you left off?

Just wondering if additional utilities or functions like "park" and "reinitialize" might be required to signal that we're ejecting or re-inserting the media.

Certainly swapping a CF with another card, particularly one of another size, would cause all kinds of problems. I do not think that we can re-initialize to the point where DOS would be happy, as it scoops up the drive parameters at boot time...
 
Actually, I think that (lucky for us) there are at least some CF cards that implement wear-leveling on the card itself.

A quick Google brought up this vendor:
http://www.stec-inc.com/product/cf.php

That's an industrial-grade card, but I seem to remember seeing this in others as well.
 
Up & Running

Up & Running

BIOS rev 11 took care of all my 1000SX problems. Everything works. The XT-IDE BIOS co-exists perfectly with my Western Digital WD1002A-WX1 MFM controller for the ST-225, as well as the Trantor SCSI controller with the Maxtor 7245SR.
The XT-IDE checked good with the 512 MB Maxtor 7245AT IDE as well as Western Digital 80 GB IDE. All partitions were about 31 MB.
Hargle is the man!
 
You must log in or register to reply here.
Back
Top