• Please review our updated Terms and Rules here

Headstart LX CD floppy boot disk image

I notice that in the LX CD thread that was referred to in post #5, that the OP tried various ROM addresses, such as D0000 and CA000.
Yes I used v1.1.5. For my Rev3 card, I can choose either D800h or C800h via a jumper on the card. I have tried both, but D800 causes one of the RAM counts to report 0kb, so I'm going to run with C8000.

I put 2 new caps in to replace the leaked ones and cleaned the board up on the original HDD, but still can't detect the miniscribe 8450XT in fdisk or Spinrite 4.0. Interestingly, I do hear the head do some seeking when I run fdisk (XTide removed).

'MODULE_VERY_LATE_INIT' is not included in the 'Official' builds of the XUB, So Yes a custom build will be required.

See https://xtideuniversalbios.org/

Since the miniscribe isn't behaving, I'm back to the XT-IDE. I will take a look at the custom build instructions for the universal BIOS Malc linked.

I don't understand how a "very late" initialization will also fix a problem solved by setting late initialization to "No". Setting late initialztion to "No" indicates an early initialization to me...

I see the very late module is already written, so it looks like I need to include that (maybe in place of the floppy control module as the built-in one is working fine) after installing all the compilers and programs at the start of the link above. Coding is outside my wheelhouse but I will try.

1680269230936.png
 
Yes I used v1.1.5. For my Rev3 card, I can choose either D800h or C800h via a jumper on the card. I have tried both, but D800 causes one of the RAM counts to report 0kb, so I'm going to run with C8000.
I just read back through the post's, v1.1.5 does not support the lo-tech ISA XT CF V3 adapter, Support was not introduced for the lo-tech boards until much later, You could try a Pre-configured binary based on r566 to see if that works or try the latest r625. See: https://www.lo-tech.co.uk/wiki/Lo-tech_ISA_CompactFlash_Adapter_revision_2b for the Pre-configured binary.
I don't understand how a "very late" initialization will also fix a problem solved by setting late initialization to "No".
It may not, Try the above first.
 
FWIW, I have two of the things (the original XT-IDE and the CF one that uses a CPLD and is SMT. I use the Hargle BIOS on those with no issue. The original uses an EPROM for the BIOS--it's the one that I hacked for the "chuck mod".
 
Does the information at [here] (just now written) help?
That is very helpful thanks. By the way, minuszerodegrees.net is a go-to site for me, as I'm heavy into ownership of early IBMs. I've used your site many times as a reference. Pic of my 5150 below.

I'm going to try making the custom build now so the XTide bios hits INT 13h, not INT 19h. The r566 build mentioned above didn't work.
 

Attachments

  • 5150.jpg
    5150.jpg
    624.4 KB · Views: 5
FWIW, I have two of the things (the original XT-IDE and the CF one that uses a CPLD and is SMT. I use the Hargle BIOS on those with no issue. The original uses an EPROM for the BIOS--it's the one that I hacked for the "chuck mod".
Interesting; model7 provided this link for the Hargle but I wasn't able to download it. Do you know which version you're using?

ftp://nostalgy.net.ru/pub/lib/XT-IDE%2520Prototipe/xtide_011.zip
 
Well, I'd suggest this is proof that the very late-init/13h is definitely required with this computer. However, you are also correct that it will fail to detect the attached IDE device due to the lack of Hi-Speed/Compatibility mode jumpers on my Rev3.

Tried a few known-good IDE-to-CF adapters and known good CF cards.
 

Attachments

  • 20230331_220815.jpg
    20230331_220815.jpg
    2.7 MB · Views: 7
If you want, I'll do a custom build Using the latest r625 with MODULE_VERY_LATE_INIT for you to try ?
 
If you want, I'll do a custom build Using the latest r625 with MODULE_VERY_LATE_INIT for you to try ?
That would be exceedingly generous of you; I've spent a couple hours trying to compile a custom build and uninstalling/reinstalling MinGW etc. with no luck!
 
Give this a try, Use IDE_XT.BIN if you have an 8088 cpu or IDE_XTP.BIN if you have an V20/30 cpu, They both include the MODULE_BOOT_MENU and MODULE_VERY_LATE_INIT
 

Attachments

  • XUB_r625_V_Late_Init.zip
    37.1 KB · Views: 2
I've read this thread and the other thread and I've also downloaded the system BIOS and analyzed it a bit in IDA Pro. There is an option ROM scan and it's called from several places. For example, one where it seems to be looking for a video card BIOS between C000h and C800h (in 2 kB increments) and another where it's looking for option ROM's at D000h, D800h, E000h and E800h. This latter scan appears to be used to determine if the 128 kB RAM in the D and E segments should be disabled - hence the message "RAM BUFFER SIZE: 0K" when an option ROM is found there. There is also a third option ROM scan between C800h and F600h (again, in 2 kB increments) and this appears to be the "real deal". A nice little bug I found; when an option ROM is found, a procedure is called to verify if its checksum is correct. The procedure works just like it should. However, the result from it is completely ignored.

Anyway, I am not able to determine why interrupt 19h fails to initialize XUB. For what it's worth, I would like to bring back "early initialization" as an option because it would make XUB compatible with more machines, (and not just the subject machines of this thread) but I have a feeling it will require extensive changes at this point. And I'm not sure when I'll get the time to do this.

BTW, I am a bit perplexed why that attempt at building the XUB failed so spectacularly. I'm guessing it was mingw32-make.exe that failed?
 
Well, had a busy week but had time to mess with this. While the very late bin that Malc kindly provided did work in that it would allow me to see the menu and C:, the setup was plagued with errors I couldn't resolve. I could fdisk the compact flash card and set an active dos partition in DOS 3.31 (my target dos as that is the dos version Headstart included), but it didn't seem to stick. For example, I could install msdos 6.22 completely, but upon boot it would be an unformatted drive. Once, fdisk found 4 active primary dos partitions despite only having 1. I tried formatting on a modern computer and transferring it, but couldn't see any files on the Headstart. Vice versa was the same. I tried 3 different manufacturers of flash cards and a small IDE hard drive (141mb I believe). Also tried the compact flash card with 6.22 I use on one of my 486s (known good card, known good installation) and the fdisk reported it had no active partitions.

All the above was after a few evenings trying things, so don't take as a hard sequence of events as it's a bit mushed together in my memory.

After endless frustration, I raided my hard drive stash and found a Seagate ST351A/X. I put the jumpers in XT mode and attached to the original controller and everything was perfect. I installed DOS 3.31 on the hard drive, edited config.sys for the palmzip.sys driver, and attached a 100mb parallel ZIP drive with the entire "LX" folder. Xcopied it onto the hard drive, and now it boots with all original Headstart menu system, software, video/sound/CD drivers. I can play music CDs on this 10mhz machine!

So, I'm left thinking either something is still causing an issue with this computer and the XT-IDE BIOS or my Rev3 card is bad. In any event, I'm up and running. While I was hoping to find the startup floppy image, which must have drivers to get the CD drive working during the install process to restore the hard drive contents, really all it does is copy data off the CD. If anyone has this computer and can get a hard drive working and another way to get data onto it, everything you need to get it to original software is included on the CD rips I put on Archive.org.

I believe other known VCFED users with this machine are Dirtbag and Al Kossow.
 
... but it didn't seem to stick. For example, I could install msdos 6.22 completely, but upon boot it would be an unformatted drive. Once, fdisk found 4 active primary dos partitions despite only having 1.
Also tried the compact flash card with 6.22 I use on one of my 486s (known good card, known good installation) and the fdisk reported it had no active partitions.
You stated earlier that you have "few known-good IDE-to-CF adapters".

Could it be that the recent activities of yours involved only one of your IDE-to-CF adapters? If so, was it the one shown at [here], one that cannot be powered from the XT-IDE card.
 
You stated earlier that you have "few known-good IDE-to-CF adapters".

Could it be that the recent activities of yours involved only one of your IDE-to-CF adapters? If so, was it the one shown at [here], one that cannot be powered from the XT-IDE card.
I did try a few. I do keep them powered. I'm willing to do more troubleshooting to solve the xt ide for myself and others. I'm considering buying another xt ide to continue troubleshooting.
 

Attachments

  • 20230408_095947.jpg
    20230408_095947.jpg
    2.2 MB · Views: 5
Some time back i remember someone having a similar problem with partitions not created and known partitioned CF cards not being seen. It turned out to be a bad +5V supply (Noise).
Have you tried this XT-CF v3 adapter in another computer.
 
I grabbed my Compaq Portable, removed it's Hardcard 20, and put in the XT-CF v3 adapter. It was also giving me weird behavior. I'm not sure when it went wrong, but I'm going to say it's no longer working properly. I'll keep an eye on prices of future XT-CF adapters and will retest the Headstart when I get one.
 
Finally found the disks to my families Lx Cd that I hope to finish repairing one day, if I can find all the missing parts lol. Here is a WinImage image of the floppy you're looking for.
 

Attachments

  • LxCd_install_floppy.zip
    481.7 KB · Views: 7
Back
Top