• Please review our updated Terms and Rules here

Baffling High Density Floppy/Controller Issue

Oh, I know. Early IBM PS/2 1.44M drives were typically host-sense. Early Teac 3.5" models had a plethora on a grid of jumpers (e.g. FD325J). If you look at the docs for the Intel 82077 FDC, there are several models, including one where the normal media type select is inverted from the usual 5.25" 5170.

Makes for an interesting life.
 
Chuck(G) said:
The DTK card takes a standard floppy cable (A:=twist, B:=flat) and standard PC-configured drives (i.e. second drive select). Your jumperless Teac is hardwired that way.

Your problem is that the add-on BIOS isn't present. You can verify this by using a utility such as Dave Dunfield's ImageDisk to see if it will read a high-density floppy (it doesn't use the BIOS for access). There are also a couple BIOS additions that run as installable device drivers, rather than ROM-based BIOS additions.
Click to expand...

OK, Ive got ImageDisk on floppy now - what do I need to do with it exactly to run a test?

Mike
 
Chuck(G) said:
Also grab the TESTFDC utility (from the same place) and start with that. Let's see what it says your drive can do.
Click to expand...

OK, I ran FDCTEST with the parms TESTFDC A: 1.44

Got back:

TESTFDC 1.16 - Copyright 2007 Dave Dunfield - All Rights Reserved.
BIOS reports drive A: as Unknown
Testing as 1.44 HD, 80-track
250 Single-Density: Format Write Verify OK
250 Double-Density: Format Write Verify OK
250 Double/128byte: Format Write Verify OK
500 Single-Density: Format Write Verify !error (0) NoAdrMark
500 Double-Density: Format Write Verify OK
500 Double/128byte: Format Write Verify OK

Report on FDC capabilities, issued 1/01/1980 1:36:42:
Single-Density at 250 kbps...............................Passed
Single-Density at 300 kbps...............................Not tested
Single-Density at 500 kbps...............................Failed
Double-Density at 250 kbps..............................Passed
Double-Density at 300 kbps..............................Not tested
Double-Density at 500 kbps..............................Passed
Double-Density at 250 kbps / 128 byte sectors....Passed
Double-Density at 300 kbps / 128 byte sectors....Not tested
Double-Density at 500 kbps / 128 byte sectors....Passed

Thoughts?

Regards,
Mike
 
I get all the same errors as you do and even more!!! And guess what? The floppy performs flawlessly. :) So, I wouldn't put too much stock in that test as it definitely seems to be flawed.
 
Of course you're going to post up your results for comparison purposes because you're testing it on the same hardware nc_mike is using right?
 
Regardless, you now know that high density (500kpbs) works on this card and your drives are correctly wired up.

So now, it seems it's just a matter of the BIOS.
 
The important line is:
Double-Density at 500 kbps..............................Passed
That means the FDC chip is reading/writing high density, used by 1.44mb disks, just fine.

Although Single-Density is not relevant to the issue as hand, the fact that it passes single density tests at 250 kbps as well as the 128 byte sector tests means that this chip is pretty good for interoperating with other non-PC 5.25" disk formats. The bottom line is, it is certainly desirable to repair this card if possible.

So now, the issue is finding out why the floppy BIOS ROM isn't doing anything.
 
I might have found the conflict... I just tried a different HD FDC controller - VTG FA-100 and ran into the same symptoms, it could boot a 720K disk and read it but failed reading a 1.44 disk. I tried it in a 2nd PC XT with the only other adapters being an EGA card and a Lo Tech XT-IDE. I thought I had pulled the XT IDE before in one of my scenarios but maybe I hadn't because after I pulled it an rebooted the XT, I could read the 1.44M floppy.

I am not sure how to resolve that conflict. I recall I tried three different addresses on the Mini-Micro and that didn't clear up any address conflict. Not sure if its a IRQ conflict maybe between the two cards? I see there is only one jumper on the XT-IDE card but have no doc for it.

Thoughts?

Regards,
Mike
 
IRQ (and DMA) has nothing to do with the failure. It's the visibility (or lack thereof) of the BIOS ROM contents in memory. And you know how to check that. You're not using an expanded memory (EMS) RAM card, perchance. If you remove the accelerator card and go with the native 8088 CPU, does that change things?

You already know how to check for the presence or absence of the BIOS in your system using DEBUG.
 
Chuck(G) said:
IRQ (and DMA) has nothing to do with the failure. It's the visibility (or lack thereof) of the BIOS ROM contents in memory. And you know how to check that. You're not using an expanded memory (EMS) RAM card, perchance. If you remove the accelerator card and go with the native 8088 CPU, does that change things?

You already know how to check for the presence or absence of the BIOS in your system using DEBUG.
Click to expand...

No EMS card on the 2nd XT, but when I pull the XT-IDE out of the 2nd XT using a totally different FCD controlled (the VTG FA-100 IDE controller - which has its own BIOS), it works with the XT-IDE pulled and not when the XT-IDE is installed. I think if the BIOS was bad on both IDE controllers that would be rather ODD, nor would the card be working in HD mode at all - but it does when I pull the XT-IDE out. That leads me to believe there is a conflict between the XT-IDE and either of the two HD FDC controllers, no?

I'll run the same test with the mini-micro on the 2nd XT machine with and without the XT-IDE board and let you know - it could be the BIOS is indeed bad in the mini-micro, but then I still can't explain the conflict between the VTG FA-100 and the XT-IDE - maybe that one is just a matter of changing the DMA of the VTG card to make that pairing work.

Mike
 
No, it's not DMA. What I am wondering is if the XTIDE isn't using the same BIOS address space as your floppy cards. What do you have the XTIDE BIOS address set to?
 
Chuck(G) said:
No, it's not DMA. What I am wondering is if the XTIDE isn't using the same BIOS address space as your floppy cards. What do you have the XTIDE BIOS address set to?
Click to expand...

I can't set the XT-IDE BIOS address - it has no jumpers other than to enable or disable the XT-IDE BIOS. When the XT-IDE loads is says MASTER at 800h.

Its confirmed now. When I put either of the two FDC controllers in the 2nd XT they both clearly load the FDC controller BIOS as long as I don't have the XT-IDE card installed. I then tried every address combo I could on both cards with the XT-IDE present but no address allows the floppy controller BIOS from either card to load. So, maybe you've nailed it - maybe the XT-IDE is using the same BIOS address as the FDC cards. I don't know if/how to change that on the XT-IDE to eliminate the conflict, if that is possible.

Regards,
Mike
 
Chuck(G) said:
Okay, we're getting there. "XT-IDE" unfortunately, describes a number of versions of the XT IDE controller. Which one do you have?
Click to expand...

Its a Lo-Tech. The PCB says ROM 32KB at C800h, IO 300-31Fh XTIDE Universal BIOS, Adapter Type CF-XT
Uses a WINDBOND W29C020C-90B

I have two of them; on one of the XTs it has the backside components for use in slot 8, but other wise they are the same and exhibit the same results. It doesn't have rocker switches - just one jumper (closed).

Yah, I was wondering if I can somehow flash it with a different address.

Regards,
Mike
 
Last edited:
"Its a Lo-Tech. The PCB says ROM 32KB at C800h" explains the problem quite precisely. That says that the XT-IDE ROM occupies from C8000-CFFF0, right on top of the floppy BIOSes.

This one? Try turning SW2 off on the XT-IDE. That should move your XTIDE BIOS to D0000-D7FFF.

I've never understood why the XT-IDE BIOes are so large. 4KB should be sufficient.
 
Chuck(G) said:
"Its a Lo-Tech. The PCB says ROM 32KB at C800h" explains the problem quite precisely. That says that the XT-IDE ROM occupies from C8000-CFFF0, right on top of the floppy BIOSes.

This one? Try turning SW2 off on the XT-IDE. That should move your XTIDE BIOS to D0000-D7FFF.

I've never understood why the XT-IDE BIOes are so large. 4KB should be sufficient.
Click to expand...

Ahh, there's the rub, it doesn't have a SW2 - no rocker switches. Not sure if it can be changed. I've got one of these

Mike
 
Last edited:
I see that there is a flash utility composed of a FLASH.EXE program and a FLASH.PAS code file. Can I load that onto a bootable low density DOS disk with the XT-IDE adapter installed and set it to different ROM base address? Not sure if I can do that with these boards as they don't have the referenced jumper #2

Regards,
Mike
 
Last edited:
You must log in or register to reply here.
Back
Top