• Please review our updated Terms and Rules here

xt-ide and wd1002a-wx1 won't play together

mikeyp

Experienced Member
Joined
Apr 16, 2019
Messages
84
Location
Leicestershire, UK
My IBM 5155 has a wd1002a-wx1 MFM controller with NEC drive. This works beautifully on its own. I would like to add an XT-IDE so bought the same Blue Lava XT-IDE Deluxe as I have already put in my 5160 with an IBM MFM controller. The XT-IDE works perfectly in the 5155 on its own but put together, neither will boot. I don’t believe there is a conflict with anything else in the system. I have however stripped it down to just the CGA card, floppy controller, MFM controller and XT-IDE for testing purposes.

Since there are not IRQ or I/O conflicts, I’m guessing as a ROM addressing issue but not sure how to calculate what each is using and therefore what to set each to. Can anyone advise me please? Thank you.

Looking at the manual for the WD card, it is set as follows:
BIOS version on the sticker is 62-000043-010 - it is socketed
W3 closed - BIOS enabled
W4 2-3 closed - I/O 320h
W5 1-2 trace wired - 32K or 64K BIOS ROM
W6 1-2 closed - 16 head hard drive supported (head select 3)
W7 1-2 trace wired - IRQ 5
W8 2-3 closed - Single controller setting, primary controller
S1/jumper 5 open - Not used
S1/jumper 6 open - Not used
S1/jumper 7 open - IRQ 5
S1/jumper 8 open - XT mode enabled

The XT-IDE is currently set as follows:
Separate “SW1” is set currently for low speed mode. I’m not changing this until it’s fundamentally working.
Switch block switches:
SW1 - off
SW2 - off
SW3 - off - I/O 0x300
SW4 - on
SW5 - off
SW6 - off
SW7 - off - ROM address D0000
SW8 - on
SW9 - on - 64k ROM chip
SW10 - on - boot ROM on
SW11 - on - write protect disabled
SW12 - on - used in slot 8

The system is configured as follows, noting that strike through lines are either disabled to not be used or planned and not installed yet, just with config mapped out.

CardFunctionMemory LocationsI/O Port HexIRQDMAPort Enable JumperRAM (k)
AST SixPackPlus Type 2/3COM1-3F8-3FF4-1-
AST SixPackPlus Type 2/3COM2-2F8-2FF3-2-
AST SixPackPlus Type 2/3LPT1-378-37A7-3-
AST SixPackPlus Type 2/3LPT2-278-27A7-4-
AST SixPackPlus Type 2/3GAME-201--5-
AST SixPackPlus Type 2/3CLOCK-2C0-2DF--6-
AST SixPackPlus Type 2/3RAM-----384
IBM CGACGAB8000H to BFFFFH-----
WDC WD1002A-WX1MFMC8000H to ?3205---
XT-IDEIDED0000H to ?300----
RTL8019ASLAN-3402---
IBM Floppy AdapterFloppy------
IBM Serial AdapterCOM2-2F8-2FF3---
Soundblaster 2.0Sound-22071--
 
Normally the XT-IDE occupies 8 or 16 KB.

I don't see any settings for the WD card so how do you know it starts at C800?

You didn't mention not being able to boot from floppy disk. Do so and check with DEBUG what data can be found at C800:0000 and D000:0000. If that is the correct data then the only thing I can imagine that something in the software clashes and causes both cards not to start at all.
 
Thank you for replying.

The XT-IDE ROM is AT28C64B so 8KB in this case.

C8000 comes from stuff modem7 said in this (https://forum.vcfed.org/index.php?t...t-will-not-boot-or-recognize-mfm-drive.71080/) thread. Unfortunately the conversation seemingly petered out without resolution.

A:\ does indeed boot (I only serviced and got it working yesterday!) with both installed. The CF card is accessible as C: but the MFM drive is not. D: returns invalid drive specification.

Debug returns the following:
XT-IDE Debug 1.jpeg
 
I have a WD1002A-WX1 and the motherboard that the IBM 5155 uses. I will try them together later today.

In the meanwhile, you could see what the RAYXTIDE tool thinks about the XT-IDE's ROM/XUB situation. Controller wise, try RAYXTIDE with only the XT-IDE fitted, then try RAYXTIDE (via boot floppy) with both controllers fitted. RAYXTIDE is available from [here].
 
I also have a WD1002A-WX1 which has the F300 SuperBios in ROM ( Not socketed) and an NEC D5126 20Mb drive. The NEC drive came with a WD1002S-WX2 controller and was setup using that controller.

I just tested the NEC D5126 drive using the WD1002A-WX1 controller in my 5160 and it worked fine, It booted into DOS with no problems, I then fitted my GW R4 XT-IDE controller using a CF and XUB r622, The combination worked flawlessly, The XUB reported the NEC drive as the 'Foreign drive' 80h and the CF 81h and i was able to boot from either drives or floppy with no problems.

What NEC drive do you have ?
 
I also have a WD1002A-WX1 which has the F300 SuperBios in ROM ( Not socketed) and an NEC D5126 20Mb drive. The NEC drive came with a WD1002S-WX2 controller and was setup using that controller.

I just tested the NEC D5126 drive using the WD1002A-WX1 controller in my 5160 and it worked fine, It booted into DOS with no problems
I've wondered for years if these two MFM cards were mutually intelligible, glad for some confirmation!
 
And my WD1002A-WX1 card and XT-IDE card (specifically, an XT-IDE Rev 4) and early IBM 5160 motherboard (with 11/08/82 BIOS) 'play together nicely'.

My WD1002A-WX1 differs to yours in only:
* BIOS version on sticker is 62-000094-002.
* W6 is set to 2-3 closed.

The difference in the W6 setting will not be the problem cause.

My XT-IDE Rev 4 contains release R622 of the XUB, and is at the default settings (base ROM address = D000h, base I/O address = 300h, hi-speed mode).

The motherboard's ISA slots contained only: IBM CGA card, IBM floppy controller, WD1002A-WX1 card, and XT-IDE card. Nothing in slot 8.

SW12 - on - used in slot 8
So, you have your Blue Lava XT-IDE Deluxe running in slot 8.
How about you try your XT-IDE/WD1002A-WX1 combination with the XT-IDE in a slot other than 8, just in case. Sometimes we see odd things. Obviously, you would need to turn switch 12 off.
 
I have done a lot, but know that each change was made sequentially and tested so I haven't changed more than one thing at a time.

OK. Slight change of behaviour. I moved the ROM address to the top end DE000. This results in the following behaviour.

Allowing to boot to 80h or C, it throws "Error 80h!" and boots to A.
Pressing "D", it now boots to DOS on the CF card.

I have also flashed the card and updated to r623, switching to hispeed mode in the process. This has yielded no change in the behaviour of either of the above modes.

Is the fact the machine is running a NEC V20 CPU significant? I only just realised this. This is what it came with, not an upgrade I have made.

The only other thing of note was that in flashing the XT-IDE card, the other utility provided, biosdrvs.com output the following:

Code:
BIOS Drive Information Tool v1.0.3
(C) 2012-2021 by XTIDE Universal BIOS Team
Released under GNU GPL v2
http://xtideuniversalbios.org/

-= Drive 80h   =-
ATA-information from AH=25h...
 Name         : TOSHIBA THNCF064MBA                    
 Cylinders    : 978     , Heads: 4   , Sectors: 32  

Will be modified to:
 Cylinders    : 978     , Heads: 4   , Sectors: 32  

 CHS   sectors: 125184
 LBA28 sectors: 125184
 Block mode   : Set to 1 from max 1 sectors
 PIO mode     : Max 2, Min cycle times: 240 ns, with IORDY -1 ns
XTIDE Universal BIOS v2.0.0á3+ (2022-06-15) generates following L-CHS...
 Cylinders    : 978     , Heads: 4   , Sectors: 32  , Mode: NORMAL
Old INT 13h information from AH=08h and AH=15h...
 Cylinders    : 978     , Heads: 4   , Sectors: 32  

 Total sectors: 125184
EBIOS information from AH=48h...
 BIOS returned error code 1h

Press any key to display next drive.

-= Drive 81h   =-
ATA-information from AH=25h...
 BIOS returned error code 1h
Old INT 13h information from AH=08h and AH=15h...
 Cylinders    : 1023     , Heads: 256   , Sectors: 17  

 Total sectors:  BIOS returned error code 1h
EBIOS information from AH=48h...
 BIOS returned error code 1h

I have a WD1002A-WX1 and the motherboard that the IBM 5155 uses. I will try them together later today.

In the meanwhile, you could see what the RAYXTIDE tool thinks about the XT-IDE's ROM/XUB situation. Controller wise, try RAYXTIDE with only the XT-IDE fitted, then try RAYXTIDE (via boot floppy) with both controllers fitted. RAYXTIDE is available from [here].
Here is the output of RAYXTIDE with the XTIDE alone. The output with the MFM card installed is identical.
Code:
*******************************************************
*  RAYXTIDE, version 2.8    (August 2022)             *
*  Does some XUB discovery.                           *
*  May be freely used and distributed, but not sold.  *
*******************************************************

+--------------------------------------------------+
| Part 1 of 4 - Motherboard class                  |
+--------------------------------------------------+
CPU type           = 8086/8088/80186/80188/V20/V30
Motherboard class  = PC or XT class (based on CPU type)

+--------------------------------------------------+
| Part 2 of 4 - ID the motherboard's BIOS ROM      |
+--------------------------------------------------+
Model byte at FFFFE   = FE        (XT)
Date at FFFF5         = 11/08/82  (in clones, do not trust this date)
Checksum of last 8 KB = Calculating ...
Checksum of last 8 KB = 000E9A00  (simple byte addition)
CRC32 of last 8 KB    = Calculating ...
CRC32 of last 8 KB    = 19F624A8       
CRC32 indicates       = IBM revision 11/08/82
                        For IBM Portable PC (IBM 5155)
                           and early IBM XT (IBM 5160)
                        Contains "1501512 COPR. IBM 1981"

+--------------------------------------------------+
| Part 3 of 4 - Look for XT-IDE card's BIOS ROM    |
|               (XTIDE Universal BIOS, a.k.a. XUB) |
+--------------------------------------------------+
Searching for the XUB .....................................................................................................

>>> Found

Base address           = DE000
Main string            = "XTIDE Universal BIOS (XT+"
Version string         = "v2.0.0á3+ (2022-06-15)"
Version release        = R623

Declared size          = 8 KB      (third byte is 10)
Checksum of that size  = Calculating ...
Checksum of that size  = 000B7600  (simple byte addition)
CRC32 of that size     = Calculating ...
CRC32 of that size     = A17FF3B8

+--------------------------------------------------+
| Part 4 of 4 - Some IBM 5162/5170 stuff           |
+--------------------------------------------------+
IBM BIOS for IBM 5162 ?  = No
IBM BIOS for IBM 5170 ?  = No

I also have a WD1002A-WX1 which has the F300 SuperBios in ROM ( Not socketed) and an NEC D5126 20Mb drive. The NEC drive came with a WD1002S-WX2 controller and was setup using that controller.

I just tested the NEC D5126 drive using the WD1002A-WX1 controller in my 5160 and it worked fine, It booted into DOS with no problems, I then fitted my GW R4 XT-IDE controller using a CF and XUB r622, The combination worked flawlessly, The XUB reported the NEC drive as the 'Foreign drive' 80h and the CF 81h and i was able to boot from either drives or floppy with no problems.

What NEC drive do you have ?

Without pulling it from the machine, I believe it's a NEC D5126 20MB 5.25" half height drive. It has the nice feature of snapping back the heads to park when it powers off. Ideal for a portable machine.

And my WD1002A-WX1 card and XT-IDE card (specifically, an XT-IDE Rev 4) and early IBM 5160 motherboard (with 11/08/82 BIOS) 'play together nicely'.

My WD1002A-WX1 differs to yours in only:
* BIOS version on sticker is 62-000094-002.
* W6 is set to 2-3 closed.

The difference in the W6 setting will not be the problem cause.

My XT-IDE Rev 4 contains release R622 of the XUB, and is at the default settings (base ROM address = D000h, base I/O address = 300h, hi-speed mode).

The motherboard's ISA slots contained only: IBM CGA card, IBM floppy controller, WD1002A-WX1 card, and XT-IDE card. Nothing in slot 8.


So, you have your Blue Lava XT-IDE Deluxe running in slot 8.
How about you try your XT-IDE/WD1002A-WX1 combination with the XT-IDE in a slot other than 8, just in case. Sometimes we see odd things. Obviously, you would need to turn switch 12 off.

Confirmed, I am only running the CGA, Floppy, WD and XT-IDE card, which is now moved to slot 5 and switch 12 off with no change.

I noted you're running a newer XUB so, as mentioned above, re-flashed to r623, specifically to ide-xtp.bin in hi-speed mode as the machine has a V20 CPU. No change.
 
Is the fact the machine is running a NEC V20 CPU significant? I only just realised this. This is what it came with, not an upgrade I have made.
The 8088-to-V20 upgrade can sometimes cause a problem.

I brought out my early IBM 5160, then upgraded its 8088 to a V20. Unexpectedly, the problem that I recorded at [here], did not appear. Thinking that the particular V20 I used might be a rebadged 8088, I ran CheckIt, and saw CheckIt both report a V20, and show a better CPU benchmark.

I then added both my XT-IDE Rev 4 and the WD1002A-WX1. At power-up, the XUB appeared on-screen, but the computer hung when going to boot from the MFM drive. I then tried the XT-IDE Rev 4 and the WD1002A-WX1 individually, and there was no problem individually. I then reverted to both the XT-IDE Rev 4 and the WD1002A-WX1 being fitted, but this time, things worked - the XUB appeared on-screen, and a boot from the MFM drive occurred. I tried many power cycles with both the XT-IDE Rev 4 and WD1002A-WX1 fitted, but the computer booted to the MFM drive every time.

Then, just in case, I removed the SixPakPlus card, but still, it all worked as expected.
 
Thank you for going to all this effort to help me.

I’ve ordered an Intel D8088 to rule the V20 out as the issue. I’ll report back when it arrives. In the mean time, I’m open to any other suggestions to try.

I was wondering if there is any mileage in flashing a ROM chip to the same version as your WD card? I would just need to know what chip to order and a copy of the ROM if you’re able to dump it please? I’ll double check your site later as I’m about to head in to work and will edit this if I find it.
 
I was wondering if there is any mileage in flashing a ROM chip to the same version as your WD card? I would just need to know what chip to order and a copy of the ROM if you’re able to dump it please? I’ll double check your site later as I’m about to head in to work and will edit this if I find it.
It's worth a shot, The F300 BIOS is on Modem7's website, http://minuszerodegrees.net/rom/rom.htm Scroll down to Western Digital WD1002A-WX1, The Eprom is a 2764, You could probably use your original Eprom if you dump it first before erasing it. If you do dump your original Eprom can you post a copy here.
 
I was wondering if there is any mileage in flashing a ROM chip to the same version as your WD card? I would just need to know what chip to order and a copy of the ROM if you’re able to dump it please? I’ll double check your site later as I’m about to head in to work and will edit this if I find it.
It's worth a shot, The F300 BIOS is on Modem7's website, http://minuszerodegrees.net/rom/rom.htm Scroll down to Western Digital WD1002A-WX1, The Eprom is a 2764, You could probably use your original Eprom if you dump it first before erasing it. If you do dump your original Eprom can you post a copy here.
Printed on the BIOS ROM of my WD1002A-WX1 is '62-000094-002'.
I can see that the first of the two 'WD1002A-WX1' rows on the referenced web page includes the comment of "62-000094-002 has the same content".
 
The sticker on my BIOS ROM is barely eligible now but it is "62-000094-032" a couple of pics:
wd1.jpg
wd2.jpg
 
Thank you both. I have also ordered "2 x D2764C-4 64K 8KX8 64K CERAMIC UV-EPROM, INTEL NEW UNUSED 2764, 28 PIN" so watch this space. I'll update as things arrive.

Interesting you both have the 2 numbers listed as having the same content. It'd be interesting to confirm it.

I'll see about dumping my ROM when I next get it apart unless I can do it with debug in the meantime.
 
I'll see about dumping my ROM when I next get it apart unless I can do it with debug in the meantime.
I use Mike Brutman's 'PCJRCART' utility for dumping ROM's, Download " PCJRCART.ZIP " It's easy to use, Stick it on a DOS boot floppy and run:
pcjrcart -h ( For help )
pcjrcart -noheaders ( To dump the extension ROMs it finds without the extra header )
 
Thanks Malc. Here's my ROM. It is way older, copyrighted 1984 as opposed to modem7's which is dated 1986. Interestingly the ROM of unknown version is also copyright 1984.

modem7, please add it to your site.

Looking in HxD I found the following date stamps. Sizes and md5 sums included too as interesting comparisons.
FileSizeDatemd5sum
unknownrevision.bin4KB29 Jan 85f442ece01c29d50ce30e29c19f59d592
62-000043-010.bin8KB01 May 85adea694ea8485e6ec1d542379b87fb98
62-000094-032.bin8KB15 Jul 8648437f554ed81e787cd5f44a9c824bdc
 

Attachments

  • 62-000043-010.zip
    3.8 KB · Views: 4
I just tried your ROM with my WD1002A-WX1, I disabled the ROM on my controller and flashed your ROM to a ROM board set to C800h, With just the VGA / Floppy / ROM board and WD controller + NEC drive i switched on and my 5160 booted into dos just fine. Though it did think my 20Mb NEC drive was a 10Mb drive, I didn't bother re-setting up.

I then fitted my GW XT-IDE R4 controller set to D800h and CF card, I switched on and my 5160 FAILED to boot from the NEC drive with error 80h but booted fine from the floppy and CF card.

It looks like your WD BIOS is not happy working alongside another controller in the same system.
 
Ooh. Interesting result. I'm feeling confident that this may be the answer. I await the arrival of my ROM chips. I have however found and dusted off my programmer.
 
...and I have received some PROMS... but the seller has sent me one time programmable PROMS, not the UV-erasable EPROMS. *sigh*
 
Back
Top