• Please review our updated Terms and Rules here

XEBEC 1220 ROM contents

S

shock__

Member
Joined
Jan 30, 2011
Messages
43
Location
Berlin, Germany
Hi,
does anyone of you have this card and could dump the ROM chips on it? In case of the "lower" ROM, dumping the 2000h bytes following the adress from DOS might be an easier way (it's a MCM68766 EPROM with a semi-uncommon pinout).
I was recently gifted one of those, sadly one of the EPROMs appears to have come from another system (the other has a XEBEC Copyright message) making the whole card not work (hangs the system after memcheck - even tho I guess running it on a 486 class machine might add to the issue). If I disable the BIOS and attach a floppy to the corresponding port I can see the card accessing the drive, if I enable the BIOS after booting (via a switch) the upper ROM can be dumped fine using DEBUG
Mine is an "ASSY. 104856" in case that's important.

CRC-32 checksums I got (no idea if uploading my dumps would be allowed here):
ROM Low: 4B62FC09 (MCM68766 - 8K)
ROM High: A13276B3 (TMS2732A - 4K)
 
The 1220 is an 8-bit card (mostly known for being stock for a 5160 hard disk controller). Only one of the ROMs is for the 8088; the other is for the Z80 MPU that executes the card's firmware.

I believe that MZD has several images of both ROMs. for the 1210
 
Not quite the ROMs I was looking for (doubt they're compatible) but their starts/headers look quite similar, making me hope that the ROMs on my card are ok.
I'd assume due to starting with "55 AA" that the ROM closer to the ISA connector is the 8088 one (8K), while the other is for the Z80 (4K).
Thanks anyways :) This already helped a bit.
 
Some small steps have been made ... first of all the "upper" (Z80?) ROM checks out with 3 or 4 differences against the "IBM (XEBEC) Fixed Disk Adapter" ROMs leading me to believe that ROM should be good (which would be a huge relief as that EPROM was left without its' window covered outside of the machine for quite a while).
Also when inserting the board into a Commodore PC-10 III (8088) or a Schneider Euro-AT Tower (286) I get an error code "1701" leading me to belief the card is at least mostly working. With a ST-251 connected I also could see some attempts to access the disk (booting didn't work as the current state of the drive is unknown [I just remember low level formating it a while ago - no idea if I also installed an OS]). I should mention all those machines and the HDD are owned by a friend of mine, so I don't have constant access to them. Sadly accessing the internal ROM routines didn't work as the floppy controller on the card collides with the internal controllers of said machines making it impossible to boot from a floppy, as the 1220 doesn't allow disabling its' FDD portion (..ooOO(maybe I shouldn't have parted ways with my 5162 and 5170)).

Just for the sake of it, here are my dumps.
"Upper" is the 2732A EPROM that's probably for the Z80, while "Lower" is the MCM68766 that likely contains the 8088 code.
View attachment xebec 1220.zip
 
Moto 64Kb EPROMs are bit unusual in the wild on production gear--they were single-source and more expensive than the commodity 28-pin parts. Usually, you see them when someone has decided to replace a 64Kb masked ROM with an EPROM (e.g. on the original 5150 motherboard).

My recollection of the Xebec configuration was that they used with a 2732 in the 24 pin socket or placed a 2764 in a 28 pin socket (your board probably has the extra 4 pins on the PCB unpopulated). AFAIK, the hard disk section of the 1220 is the same as that of the 1210--the 1220 just adds a floppy controller.
 
I agree with your judgement on the MCM68766, especially since that chip originally had a handwritten label saying "AKRO2" on it. Maybe the BIOS was manually updated on the card at some point? But I agree, it's unusual to find an EPROM for which the datasheet says the primary use case would be pin compatible prototyping for a later production run of mask roms.

Taking a closer look on the PCB revealed something funny: The footprint for the 2732 has 4 additional pins, with Pin1 and the highest 2 being connected to VCC and Pin 2 (or -1 in relation to the current population) being a signal trace (which I'm assuming to be A12 allowing you to use 2764 EPROMs instead of 2732) those additional(?) pins are not represented on the silkscreen tho.

Judging from this picture I found via eBay the placement of the ROMs on my board seems to be correct tho: https://www.picclickimg.com/d/w1600/pict/141646645113_/Xebec-1220-8-Bit-Isa-Mfm-Hard-Drive.jpg
What also confuses me is the fact that most cards I found appear to have a label saying 104833A indicating a BIOS revision (that label was missing on my card) - yet the _other_ ROM of mine has a string 104830A3 inside which I interpret as some kind of revision number. This could be perfectly normal indicating another revision of the BIOS or some bits being randomly swapped/flipped/corrupted. I'm assuming the first case for now.

To my judgement the 1220 also appears to have an additional jumper block for setting the ROM adress (W4 - http://arvutimuuseum.ee/th99/c/U-Z/21080.htm) but I guess that should be done indenpendantly from the actual ROM code (likely setting a chip enable via the 74LS688 chips) so swapping ROMs with the Xebec 1210 dumps might be worth a try (since the floppy controller is doing more harm than good currently).
 
Ok, some progress.
Turns out the FDD portion of the controller is working (360k - also works with 3.5" drives)
Using the HDPREP 4.5 tool I was able to read out the configurations set by W5

Turns out my BIOS has the following settings (all 17 sectors per track)
4 heads, 306 cylinders (ST-412, 10mb total)
8 heads, 320 cylinders (Rodime 204, 22mb total)
4 heads, 612 cylinders (no idea what drive has those specs, 20mb I assume)
6 heads, 306 cylinders (no idea either, probably 15mb)

So in case anyone of you has a dump of the "104959/102168" BIOS prominently mentioned in TH99, I'd be highly interested (currently I only have ST238R and ST251 drives around for testing). Programming and adapting EPROMs to be used on my card is no problem here.
 
In a pinch, you can modify the drive table in the BIOS for the drives that you want to use, re-calculate the checksum, burn a new PROM and Bob's your uncle.
 
shock__ said:
... (currently I only have ST238R and ST251 drives around for testing).
Click to expand...
Just in case you are unaware:

It is possible to use your drives with the present BIOS in the 1220, although at reduced capacity. For example, your ST-251 has 6 heads and 820 cylinders. If the W5 setting for '6 heads / 306 cylinders' is selected, then the 1220 will only use 306 of the ST-251's 820 cylinders.

Although your drives may have been low-level formatted elsewhere, there is reasonable possibility that the particular low-level format that was laid down cannot be used by the 1220. If that is the case, you will need to low-level format the drive when it is connected to the 1220.

I expect that the cabling diagrams at [here] apply.
 
modem7 said:
It is possible to use your drives with the present BIOS in the 1220, although at reduced capacity. For example, your ST-251 has 6 heads and 820 cylinders. If the W5 setting for '6 heads / 306 cylinders' is selected, then the 1220 will only use 306 of the ST-251's 820 cylinders.
Click to expand...
I was just gonna post, that low level formating the drive @ 612 cylinders & 4 heads worked ... it's currently copying over PC DOS 3.3 as I'm typing. I guess that confirms the controller as working. Hoorj.

modem7 said:
Although your drives may have been low-level formatted elsewhere, there is reasonable possibility that the particular low-level format that was laid down cannot be used by the 1220. If that is the case, you will need to low-level format the drive when it is connected to the 1220.
Click to expand...
That's also correct. The drive was previously formated on a 16bit controller as type 40, low level formating via HDPREP (at reduced capacity) worked nice, followed by FDISK, FORMAT and SYS - booting also works.

modem7 said:
I expect that the cabling diagrams at [here] apply.
Click to expand...
Correct once again - at least as far as I can judge ... connecting the twisted end while the drive is jumpered as DS0 results in lots of noise from the head banging.
 
You must log in or register to reply here.
Back
Top