• Please review our updated Terms and Rules here

XTIDE tech support thread

Finally thought I'd give my own XT IDE Adapter a try but running into some problems.

I got a kit for http://www.lo-tech.co.uk/wiki/Lo-tech_ISA_CompactFlash_Adapter_revision_2
Soldered it all up and noticed the BIOS saying it detected checksum errors when I ran the machine without turbo. Removing the SMT components afterwards seems to have fixed this at least.
Otherwise the card works fine in my 386/486 machines with several compact flash cards and they boot and run fine. But trying it on my 8088 it doesn't seem stable.
Occasionly it will detect the compact flash's IDE name but just fail to boot, I'm guessing there's some noise issue or having the compact flash drive the lines directly doesn't have enough power for the 8088.
Anyone have any ideas or should I go for a full XT IDE card not just a lite version and will that maybe buffer the data lines for better stability?
 
Not sure if this helps but Have you seen the "FAQ" page for this board:

"SMT components should not be populated for use in machines that do not support the IBM PC/XT slot-8 card select signal"

Hopefully James will be along or you can contact him via the lo-tech website.
 
The board is fully buffered on the IDE side. Check that the CompactFlash adapter is set to 5v operation and provide it with a direct power connection if you're getting inconsistent behaviour.
 
Last edited:
Tried multiple compact flash adapters and just power them from a floppy connector, measuring voltage seems good 5V as well.
Running the machine without turbo it won't even detect the compact flash card and with turbo it shows the name of the flash manufacturer but every time random letters are wrong.
Something like "S%MSUN CF/ATA" whatever comes out. Maybe some weird timing issue or noisy data lines, even though videocards seems fine with it.
Might be a good excuse to finally buy a scope :)
 
Malc said:
"SMT components should not be populated for use in machines that do not support the IBM PC/XT slot-8 card select signal"
Click to expand...
I am in a 'picky' mood today. The use of "should" indicates optional. If it is not optional, James should use "must" instead of "should".
 
modem7 said:
I am in a 'picky' mood today. The use of "should" indicates optional. If it is not optional, James should use "must" instead of "should".
Click to expand...
Well, you sure botched that one. The correct syntax is... The use of "should" indicates optional. If it is not optional, James must use "must" instead of "should".
 
Stone said:
Well, you sure botched that one. The correct syntax is... The use of "should" indicates optional. If it is not optional, James must use "must" instead of "should".
Click to expand...
As far as I am aware, James is not under any local/state/federal/international law that forces him to use "must".
Assuming that is the case, James could, for example, choose to use "are" instead.

"SMT components are not be populated ..."
 
modem7 said:
As far as I am aware, James is not under any local/state/federal/international law that forces him to use "must".
Click to expand...
Sorry. It has been brought to my attention that in the above, I left "intergalactic" out of "local/state/federal/international". Because the laws of that domain are unknown to me, I will need to do some research. Accordingly, the next time that I am abducted by aliens, I will be sure to make enquiries.

Again, my apologies. The omission was not intentional.
 
My apologies for any confusion on this one!

The SMT components provide two functions: slot-8 compatibility, and PC/AT ZWS operation (providing a speed boost for the ROM code and IDE reads). Generally, since the board is almost always destined for PC/XT class hardware, the components should only be added if the the board is destined for a real PC/XT because *some* other manufacturers used the line in question (ISA B8 ) for other purposes, like analogue audio in the case of some of the Tandy 1000 series.

Most machines however simply have the line unconnected, in which case it makes no difference whether the components are on or not. The PS/2 model 30-286 (and not doubt some other machines) won't boot from the adapter with the components in place for reasons currently unknown. This is however noted in the adapter FAQ. If there are other examples, please do let me know via PM, my site or post here so they can be added to the wiki as well.

Hope that helps.
 
I have two XT IDE controllers identically built, one for me and another for my friend.
Both are "classic" XT IDE v2 but with differend BIOS and jumper setting
1) Old XT IDE Universal BIOS 1.1.5 w/o "Chuck mod"
2) New XT IDE Universal BIOS 2.0beta3 with "Chuck mod"

I'm trying to test bunch of my old IDE drives and have a problem with Western Digital Caviar 11200 AC11200-00LA.
It works with first "slow" controller only and not detected at all with BIOS 2.0beta3 and "Chuck mod".
I have also tested both controllers with different drives:
- 1Gb TraxData CompactFlash with FG-ADIDE2CF-A2-01-CT21 adapter. Both controllers work fine.
- Quantum Fireball 1280AT. Both controllers work fine.

Did anyone see similar problem with WD Caviars ?

My test system is Taiwanese turbo XT-compatible motherboard, 8MHZ, Fujitsu 8088, AMI BIOS.
 
the caviar drives seemed to be a mixed bag. Most of them work, but at least one (and now two) didn't:
http://www.vintage-computer.com/vcforum/showthread.php?26435-XTIDE-TestResults

these test results were all from the rev1 cards (slow) so that may be why more of the caviar drives are in the working list and not the failing section.

This is one of those things where we could spend a lot of timing digging into exactly why this one drive doesn't work, or you could get any number of other drives which do work and save a lot of time... :)
 
alecv said:
I have two XT IDE controllers identically built, one for me and another for my friend.
Both are "classic" XT IDE v2 but with differend BIOS and jumper setting
1) Old XT IDE Universal BIOS 1.1.5 w/o "Chuck mod"
2) New XT IDE Universal BIOS 2.0beta3 with "Chuck mod"
Click to expand...

Both very old versions of the Bios now, The latest Revision of the Bios is R568, If you can't build your own James has the R567 binary's on his site if you wanted to try them.
 
alecv said:
I'm trying to test bunch of my old IDE drives and have a problem with Western Digital Caviar 11200 AC11200-00LA.
It works with first "slow" controller only and not detected at all with BIOS 2.0beta3 and "Chuck mod".
Click to expand...
First try the latest revision (r568) as Malc said, if that doesn't help I would suggest trying a special build of r568 but without ATA-ID validation. PM me your e-mail address if you need me to do this for you.
 
I've successfully build new SVN BIOS under Cygwin. The result :

- 1Gb TraxData CompactFlash with FG-ADIDE2CF-A2-01-CT21 adapter. Works
- Quantum Fireball 1280AT. Works.
- Western Digital Caviar 11200 AC11200-00LA (DCM: ASBAECW0CAZ) - NOT DETECTED
- Western Digital Caviar 26400 AC26400-00RN (DCM: CRBBKMVB0BAB) - NOT DETECTED

Both WD drives are working and tested with USB-ATA box.
It seems really there is an issue with WD Caviar HDDs.
 
alecv said:
It seems really there is an issue with WD Caviar HDDs.
Click to expand...

Indeed. Alec helped me verify that these drives fail ATA-ID validation as I suspected so it will be disabled as of next revision and likely removed completely at some point in the future. Case closed. :D
 
I've found another issue with XT-IDE controller. I'm trying to use SD-to-CF adaptor card with XT-IDE rev2
and current XT-IDE Universal BIOS.

The adaptor is
http://www.aliexpress.com/item/New-...Card-Adapter-33MM-CF-ADAPTER/32279608662.html
as far as I understand it's original name is Micreal MR-C053
cf.jpg sdcf.jpg
The adaptor uses KTC FC1307A chip as convertor from SD to CF.
To connect to IDE I use IDE2CF-A1 board.

XT-IDE controller can't fully identify this adaptor, it can't even read full IDE ID name,
and cuts it to 3..5 characters. BIOS determines different disk size on every resets.
MS-DOS's fdisk can't partition it e.t.c.
This adaptor works VERY unstable.

It seems there are timing issues with this KTC FC1307A chip. This adaptor
works well and stable in my USB CF reader.

I tried many SD/uSD cards, with the same result. Inserting my real 1GB TraxData
CF Card all works fine.



DO NOT BUY this "red" adaptors at e-bay or aliexpress, it WILL NOT WORK with XT-IDE.
There are "blue" and "yellow" similar adaptors, not tried yet.
 
These adapters also appear to NOT implement the (required) 8-bit transfer mode. On the back of the board you will find a small EPROM as well.
 
I have one of these here with the KTC chipset, and it also had issues with the XTIDE BIOS on my 486's onboard IDE. Eventually FreeDOS would just stop working, as if the disk had failed. Taking the card out and reading it on a USB reader worked fine, though.
 
Hi,

I have bug with latest v2.x.x firmwares and SYSINFO.EXE from Norton Utilities v5.0 - when i chose "Disk Benchmark" from menu, "Disk Speed Benchmark - Testing ..." window appear but nothing happens. No IDE activity, "Testing ..." message blinking forever. On the other hand at v1.1.5 XT-IDE BIOS this benchmark worked very well.

ce26d313348f40bdb936c62380f23105.png


v1.1.5 - worked
v2.0.0b3 - not work
v2.0.0b3 latest svn build r582 - not work

My hardware is custom XT-IDE based on altera CPLD with "Chuck mod", very similar to XT-IDE v2.
 
Tronix, are you benchmarking CF drives or similar fast media? I'm asking because you are the third person to report this specific problem and the other two, alecv and HoJoPo, were both using CF drives.

I just ran the disk benchmark in SI.EXE in Bochs using the latest version of XTIDE Universal BIOS and it didn't hang for me when benchmarking a serial harddrive so I think it's a problem with SI.EXE and very fast drives. If you haven't already, try benchmarking a real harddrive and see if that works?
 
You must log in or register to reply here.
Back
Top