- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
XTIDE tech support thread
- Thread starter hargle
- Start date
MarsMan2020
Experienced Member
All of the information that was on the old PBWiki on the XT-IDE V2 & archived data on the original V1 can be found at - http://goo.gl/5kBOHz (I am sorry for the Google shortened URL - this is a link to a wiki page at http://www.retrobrewcomputers.org/ - for some reason the forum software was mangling the link when adding the ...'s in the middle)
Andrew L. did indeed shutdown the N8VEM sites and groups and ask us to discontinue use of his ham radio callsign for new board designs, but the old data from the PBWiki is being re-populated into this new wiki. We are using a FOSS wiki software on a server we control this time, to avoid having things locked into a proprietary site like PBWiki.
I populated the XT-IDE page after seeing this thread, I hope the information there is useful.
Andrew B.
Excellent, Great to see this info on the "Original" XTIDE cards archived, Would have been a great shame if it had been lost after all the hard work that's been put into the project, My R1 cards are still going strong today
I have a couple I purchased a while ago and never assembled, Finally getting time to use them. Good to see the information was not lost. I am also considering getting a couple more boards just to have. Is there information on how to get the BIOS on them still?
SquallStrife
Experienced Member
I recently bought a Lo-Tech ISA ROM board, with the intention of using the XTIDE BIOS plus a 16-bit IDE card in my 256KB 5150. And wouldn't you know it, the thing just won't play ball. This persists across two different ISA IDE cards, and a fistful of HDDs of various sizes.
My first action was to verify the functionality of the ISA cards, the cable and the HDDs in a different system, and they are all in good order.
The BIOS loads and attempts to detect the HDDs, but it always finds no disks. By fiddling with the HDD slave/master jumpers, it will occasionally be a little slower or quicker to "find nothing", but it hasn't actually detected anything so far.
I configured the BIOS to use "ISA card in 8-bit mode", on ports 1F0+3F0. The version of the bios is "R580" as found at the site Malc linked in the above post, I have tried ide_xt.bin and ide_xtl.bin to no avail.
Below is a photo of the IDE card I'm keen to use. As you can see, the only ASIC on there is a floppy controller, the IDE portion of the card seems entirely done with 74-series logic. I'd have thought this would be a perfect candidate for XTIDE usage! The other ISA IDE card I have is a multi-IO card with a Winbond W83757F + Winbond W83758P set. The symptoms are identical across the two.
My first action was to verify the functionality of the ISA cards, the cable and the HDDs in a different system, and they are all in good order.
The BIOS loads and attempts to detect the HDDs, but it always finds no disks. By fiddling with the HDD slave/master jumpers, it will occasionally be a little slower or quicker to "find nothing", but it hasn't actually detected anything so far.
I configured the BIOS to use "ISA card in 8-bit mode", on ports 1F0+3F0. The version of the bios is "R580" as found at the site Malc linked in the above post, I have tried ide_xt.bin and ide_xtl.bin to no avail.
Below is a photo of the IDE card I'm keen to use. As you can see, the only ASIC on there is a floppy controller, the IDE portion of the card seems entirely done with 74-series logic. I'd have thought this would be a perfect candidate for XTIDE usage! The other ISA IDE card I have is a multi-IO card with a Winbond W83757F + Winbond W83758P set. The symptoms are identical across the two.
SquallStrife
Experienced Member
Good to know. It's reassuring that it's not something I screwed up!!
Having said that, with this information in hand, I should be able to modify the card to use higher IO ports (thanks to its simplicity).
Edit: I buzzed out the decoder schematic with the ol' multimeter. Dead simple.
So now it should be a nice, simple case of breaking in to the circuit and tacking on some additional 74-series logic to futz around the 'middle' 7 bits.
Code:
** **** *
A: BA98 7654 3210
1F0: 0001 1111 0000
3F0: 0011 1111 0000
300: 0011 0000 0000
308: 0011 0000 1000Edit2: The address lines are "active low" right? So a TTL low represents '1'.
Last edited:
SquallStrife
Experienced Member
I'm close!! I can FEEL it!!
I used a 74LS04 inverter, and built the following "translation":
ISA A0...A2 -> Card A0...A2
ISA A3 -> Invert -> Card A9
ISA A4 -> Invert -> Card A4
ISA A5 -> Invert -> Card A5
ISA A6 -> Invert -> Card A6
ISA A7 -> Invert -> Card A7
ISA A8 -> Card A8
ISA A9 -> Card A3
Now, if I use "16-bit ISA IDE in 8-bit mode" mode in the BIOS (with the ports set to 300/308), the HDD is detected but the name is garbage. BIOSDRVS seems to display the correct parameters but trying to FDISK or format doesn't work.
As a crap-shoot, I let the config program do auto-detect, and it thought it saw an "XTIDE rev 1". When using this, every second character of the drive name is correct, and every other character is a zero. (E.g. WDC 21000-00H became 0D0 01000-000 0 0 0 0 0 0 0 0 0 0 0). Since it was able to read half of the HDD name bytes correctly, I feel like I'm SUPER close to having this running!!
Any ideas?
I used a 74LS04 inverter, and built the following "translation":
ISA A0...A2 -> Card A0...A2
ISA A3 -> Invert -> Card A9
ISA A4 -> Invert -> Card A4
ISA A5 -> Invert -> Card A5
ISA A6 -> Invert -> Card A6
ISA A7 -> Invert -> Card A7
ISA A8 -> Card A8
ISA A9 -> Card A3
Now, if I use "16-bit ISA IDE in 8-bit mode" mode in the BIOS (with the ports set to 300/308), the HDD is detected but the name is garbage. BIOSDRVS seems to display the correct parameters but trying to FDISK or format doesn't work.
As a crap-shoot, I let the config program do auto-detect, and it thought it saw an "XTIDE rev 1". When using this, every second character of the drive name is correct, and every other character is a zero. (E.g. WDC 21000-00H became 0D0 01000-000 0 0 0 0 0 0 0 0 0 0 0). Since it was able to read half of the HDD name bytes correctly, I feel like I'm SUPER close to having this running!!
Any ideas?
Do not expect normal harddrives to work in this mode, because very few drives do (if you find one that does work, please post the brand and model). Use a Compact Flash card or Microdrive instead.
SquallStrife
Experienced Member
OK, thanks for the advice.
I suppose the other alternative is to jury-rig the XT-IDE's bus mux arrangement, but that will get rather ugly. I think I'll just build an XT-IDE on protoboard.
It was an educational experience though, if it "just worked" then I wouldn't have learned about all those differences in the 5150.
Just another reminder why I love this place!
Send Bonedaddy a PM, I don't know if Todd has any left but worth asking.
SquallStrife
Experienced Member
This card is now working with an IDE->CF adaptor.
Here's the modded ISA IDE card. I desoldered the 74-series chips to "break in" to the address lines and do my translation. I also de-soldered the floppy controller IC since the FDC disable jumper seemed to only work intermittently. Another modification I did was to cut the trace to the "ALE" pin on the card edge, the CF adaptor was sinking it and causing the system not to boot.
Here's the card installed with the IDE->CF adaptor installed. This card is something I found kicking around in a drawer, I have no idea what brand it actually is. The identifier string simply says "Flash Module". *shrug*
BIOSDRVS correctly identifies the card. I can successfully partition and format the card.
My 5150 can now successfully boot from a CF card. Woo!
Great work!
This is interesting. I suppose this is why some combinations of CF adapters and IDE interfaces don't work together? Who is to blame for this? The IDE interfaces or the CF adapters?
SquallStrife
Experienced Member
It was definitely related to the CF adaptor. If I removed it but left the IDE card installed, the system booted just fine.
It seems like IDE pin 28 is connected directly to ISA pin B28 on some controllers (which is weird, because I thought pin 28 was CSEL). I ran out of time last night to see where pin 28 goes on the CF adaptor, I'll take a look tonight when I get home.
Edit: Did some more research, and some pinouts seem to show IDE pin 28 as "Host ALE". Fascinating. Is this something that changed over the years?
If that is the case, then that could explain it. The Master/Slave jumper on the CF adaptor probably shorts the CSEL pin on the CF connector to ground, but because these are cheap Chinese CF adaptors, they /also/ routed the CSEL pin from the CF connector to the CSEL/ALE pin on the IDE connector.
Edit2: My CF adaptor is shown on the XTIDE Universal BIOS's compatibility page as "Does not POST" for all tested configurations. https://code.google.com/p/xtideuniversalbios/wiki/DriveCompatibility
However as we've seen, by disconnecting ISA pin B28 the adaptor works. I think you might be on to something, Krille!
Last edited:
MarsMan2020
Experienced Member
Todd should have some. http://www.retrobrewcomputers.org/doku.php?id=boardinventory - under "Other".
I have that same CF adapter but different 16-bit IDE / Multi IO controller cards, That same CF adapter works fine on my older controller cards, However on the newer controller cards it can cause a no boot situation or just doesn't work ie: CF card not found.
I just grabbed a newer Multi IO controller card and covered B28 with electrical insulating tape and fitted the controller into my 5160 with CF adapter plugged into the IDE header of the card, With B28 "disconnected" The machine now boots, However i now get a 601 error even though i have the Floppy controller disabled on the 16-bit card, Pressing F1 to continue the machine boots successfully from floppy drive which is connected to the original IBM floppy controller. If i connect a cf Card it is not found and boots to Basic.
A long time ago when i was testing different controller cards with the XUB i remember reading something about the AT Spec ?, Something changed between spec 2 and spec 3 ? and this change made some cards incompatible but i can't remember exactly what it was all about now.
SquallStrife
Experienced Member
Fascinating results. I wonder if this is a 5150 peculiarity?
Edit: Also, does your ISA Multi-I/O card have a single "HDD+FDD" ASIC, or separate ICs? I wonder if some cards make use of the ALE signal somehow. In which case, how about isolating pin 28 of the IDE connector?
There's got to be a pattern!
Last edited:
Fascinating results. I wonder if this is a 5150 peculiarity?
Edit: Also, does your ISA Multi-I/O card have a single "HDD+FDD" ASIC, or separate ICs? I wonder if some cards make use of the ALE signal somehow. In which case, how about isolating pin 28 of the IDE connector?
There's got to be a pattern!
Success
This Multi IO controller uses Winbond W83758P and W837871F chips and looking at the traces on the card they are interconnected.
I removed Pin 28 from the IDE header, Removed the insulating tape from B28, Fitted the controller card with CF adapter back in my 5160 and switched on, The machine successfully booted from floppy with no 601 error, I switched off and fitted the CF card and switched on, The machine booted and the CF card was recognized and successfully booted into dos.
I'll have another look at my other Multi IO controllers / CF adapters if i can find them later.
Edit: Looking more closely at the CF adapter i can see that Pin 28 from the IDE header is connected to Pin 16 (A04) and Pin 14 (A06) of the CF header, Looking at another different CF adapter i can see that Pin 28 from the IDE header to the CF header is NC. So why the 2 address pins are conected to IDE pin 28 i don't know ?
Last edited: