• Please review our updated Terms and Rules here

XT-FDC project level of interest

I started to populate the PCB, only 82077AA-1 and its related components. Plugged a 3 1/2 HD floppy drive and tried some DD disks, it gives random "Drive not ready" and "General Failure" errors. I will check and test it in more depth tomorrow.
 
"Drive not ready" is actually PC-ese for "I issued an operation and it didn't complete in a set time". The FDC will hang if, say, an index signal is missing or the drive isn't spinning. "General failure" is a catchall and can be anything from "what I got back isn't what I expected" to "stupid controller's out to lunch".

Check the smaller pullup resistor pack for soldering issues. You can also check that with the controller out of the machine by measuring resistance at the drive connector between the +5 supply line on the FDC and pins 8, 26, 28, 30 and 34 on the connector (no drive attached).

Finally, make sure that any existing floppy controller is disabled. If you're trying to test this thing on a P3 or later system, some of the whacko subtractive-arbitration chips special case DMA 2 and do not make it available on the ISA bus.
 
I have checked all FDD header pins and they are connected where they should be. Did the same with the PLCC68 IC and the ISA connector. No apparent shorts or bad connections. Replaced the small resistor pack with a 300R one, same results. Put back a 1K one.

Tested in three machines: XT Clone 4.77 Mhz, Pentium 100 Mhz and Pentium II 233 Mhz. Same results.
Using three floppy drives: two 1.44MB 3 1/2 and one 360KB 5 1/4. Same behavior.

Some info:
- The P2 BIOS perfoms a drive seek test, and it gives no error (and floppy drive makes typical seek sound).
- The XT clone has a XTIDE installed. When booting from floppy drive, the XTIDE Universal Bios returns a 80h error if no floppy is inserted and 02h error if there is a floppy inserted.
 
Last edited:
I needed to take a break from something, and since the balance of ordered components had arrived today, I decided to solder them in and see if the 8477 floppy section worked.

It did. I can boot from a 360K floppy and read its entire surface okay. That was done on an XT clone.

In the photo below, the jumpers are:
8477 floppy: red jumpers selecting 3F0h, IRQ 6, and DMA 2
Optional ROM drive: white jumpers selecting 'rom enable', 'write enable', and address C8000
Optional boot ROM: blue jumpers selecting 'rom enable', 'write enable', and address D0000

The crystal and three capacitors between the 8477 and U1 were moved to the other side of the board.

I'll do some more testing when I get time.

xt-fdc_at_24oct12_1.jpg


xt-fdc_at_24oct12_2.jpg
 
Further testing:
Testing of 8477 chip and primary internal conector (P18).

360K in XT clone

On my clone XT, I ran the floppy drive check (random seek/read/write) in CheckIt.
For each of the following types of 360K drives, CheckIt ran that check successfully for about an hour.
1. Matsushita JA-551-045
2. Panasonic JU-455-7 (Matsushita)
3. Tandon TM-100-2A

Cable length of 28 inches.

360K in IBM PC (5150)

I ran the floppy drive check (random seek/read/write) in CheckIt for an hour against the as-supplied Tandon TM-100-2A.
Cable length of 22 inches.

360K/1.2M in IBM AT (5170)

Problems here. The drive lights are coming on as expected, but 'nobody is home'. I've tried various configurations. I'll continue this later.
 
Hi,

It looks that I am a bit late to the party. Although I was aware that Andrew designing this board, I missed this thread completely.
Anyway while ago I wrote a BIOS for my XT board, and it includes AT-like INT 13h floppy disk support. I tested it with 5.25"/360 KB, 5.25"/1.2 MB, 3.5"/720 KB, and 3.5"/1.44 MB disks. Feel free to reuse it for this board.

The BIOS is released under GPL terms, and the latest version can be found here: http://n8vem-sbc.pbworks.com/w/file/60420678/bios-0.7e.tar.gz (floppy1.inc and floppy2.inc contain the floppy part).

Regards,
Sergey
 
Further testing:
Testing of 8473 chip (U2) and primary internal conector (P18).

1. Removed PC8477BV chip at U16.
2. Inserted DP8473V chip at U2. (Note: used DP8473V, not DP8473AV.)
3. Installed jumpers JP3 and JP6.
4. Fitted card to my IBM 5160 (XT) in place of the standard IBM controller card.
5. Booted successfully from A: drive, a Tandon TM-100-2A.
6. Ran the floppy drive check (random seek/read/write) in CheckIt against A: for an hour - no errors.

Cable length was 22 inches.
 
Woo Hoo! Thanks! That's great news!

I think all the major subsystems have been tested now except the external floppy drive connector and the 8" floppy internal connector.

Chuck was working on a BIOS and apparently Sergey has one too.

Are there any other changes except for the 7 noted earlier?

I am thinking to move the DP8473 chip to the left and up to make more room for the PC8477. Also to move the ROM chips lower to make some room for the three 7407 buffer chips.

Thanks and have a nice day!

Andrew Lynch
 
lynchaj said:
I think all the major subsystems have been tested now except the external floppy drive connector ...
Click to expand...
At post 173, Chuck wrote, "I can add that the external drive connector (DC37F) works just fine with my external drives."
So it will be the secondary internal connector (P2) that is untested.

lynchaj said:
... and the 8" floppy internal connector.
Click to expand...
I'll look at that once I work out why my XT-FDC card isn't functional in my IBM AT (5170).

lynchaj said:
Are there any other changes except for the 7 noted earlier?
Click to expand...
Per http://www.minuszerodegrees.net/xtfdc/photos.htm

1. Label positions of JP7 and JP9 changed.

2. Unless physically smaller caps can be found for C2A-1, C2B-1, and C2C-1, then more space for those caps. Whilst what I show is within the capability of builders, unless the caps are glued/secured in place somehow, then the possibility of leads shorting out exists. Insulation could be put on the leads, but more 'real estate' for the caps (to suit their size) is a much cleaner solution.

Note that item 2 above conflicts with Chuck's suggestion to move the 8" drive connector to the left. I'm guessing that the extra 'real estate' will appear due the card being sized increased to cater for the to-be-added 7407 chips.
 
Still no luck with Intel parts. I unsoldered a 82077AA from an IBM motherboard and put a PLCC68 socket instead:
- The 82077AA-1 works in the IBM motherboard.
- The 82077AA doesn't work in the XT-FDC controller, behaves exactly the same.

img20121029235106.jpg
 
Crypticalcode0 said:
nestor i see you forgot to do the fix on the address decoder.
Or is that resistor it?
Click to expand...
To me, that looks like he had a 5-pin resistor SIP for use in position RR1 rather than a 6-pin, and added a standard quarter watt resistor to compensate.
The question remains. Has the decoder fix been applied?
 
modem7 said:
To me, that looks like he had a 5-pin resistor SIP for use in position RR1 rather than a 6-pin, and added a standard quarter watt resistor to compensate.
The question remains. Has the decoder fix been applied?
Click to expand...

You are right I replaced all parts trying to discard any failing component but I only had a spare 5 pin resistor SIP, so I put a in the empty hole a standard resistor tied to VCC.

The decoder fix is applied: I cut the socket pin 16 and isolated its PCB hole so the pin makes no contact with it. Then I soldered a wire from the socket pin 16 to socket pin 17. As this mod is made under the socket, it is almost invisible.
 
Last edited:
nestor said:
You are right I replaced all parts trying to discard any failing component but I only had a spare 5 pin resistor SIP, so I put a in the empty hole a standard resistor tied to VCC.

The decoder fix is applied: I cut the socket pin 16 and isolated its PCB hole so the pin makes no contact with it. Then I soldered a wire from the socket pin 16 to socket pin 17. As this mod is made under the socket, it is almost invisible.
Click to expand...
How about I post you a PC8477BV-1 (free for prototype testers)? It should be thin enough that I can get it through as a letter. PM me your postal address in Spain.

Of course, if the problem cause is not your chip, there is always the chance that you may determine the cause between now and when my chip arrives.
 
Ok, PM sent.

Current status: The floppy disk LED lights up when it should and I can hear the head motor noise when floppy seek tests are performed... but any access will return a error ("Drive not ready" / "General failure" from DOS, 02h or 80h errors from XTIDE Universal BIOS boot).

At the moment I think there are two possibilities that can explain the troubles:
-PCB defect (shorted traces or similar problem). I will check deeply with a continuity tester.
-82077AA-1 incompatibility. I have downloaded the datasheet and seems pretty similar to PC8477BV-1... but maybe there are small differences.
 
nestor said:
Ok, PM sent.

Current status: The floppy disk LED lights up when it should and I can hear the head motor noise when floppy seek tests are performed... but any access will return a error ("Drive not ready" / "General failure" from DOS, 02h or 80h errors from XTIDE Universal BIOS boot).

At the moment I think there are two possibilities that can explain the troubles:
-PCB defect (shorted traces or similar problem). I will check deeply with a continuity tester.
-82077AA-1 incompatibility. I have downloaded the datasheet and seems pretty similar to PC8477BV-1... but maybe there are small differences.
Click to expand...

Hi
Thanks!

There are some small differences. The 82077 requires a small capacitor on the filter pins that is not required on the PC8477. There are some other differences as well. Compare the schematic against the PC8477 reference circuit in its datasheet and you will see them. I think Sergey is familiar with the differences as well.

Thanks and have a nice day!

Andrew Lynch
 
Continuing with the problem I saw with the XT-FDC in my IBM AT (5170).
What I have discovered so far is:

BIOS SETUP for A: = 1.2M, A: drive = 1.2M, boot disk = 1.2M: LED comes on, spindle motor activates, heads seek, but won't boot (jumps into BASIC)
BIOS SETUP for A: = 1.2M, A: drive = 1.2M, boot disk = 360K: LED comes on, spindle motor activates, heads seek, but won't boot (jumps into BASIC)
BIOS SETUP for A: = 1.2M, A: drive = 360K, boot disk = 360K: Boots okay

Same data/control cable used in the three scenarios above.

Multiple 1.2M drives (known working) tried.

Same results with either the 8477 chip or the 8473 chip.

The code controlling the XT-FDC is the POST and BIOS code in the 5170 motherboard, code that 'knows' all about 1.2M drives and disks.

Compared to the IBM floppy controller supplied in the 5170, maybe the 8477/8473 chips require extra ininitialisation and/or support code for HD operation.
 
You must log in or register to reply here.
Back
Top