Home Page Forums MAME MAME Requirements?

yea.. I have it somewhere. I have the manual sitting here too at least for Omninet support. Flat cable support should be somewhere, I worked with the guy who wrote the DOS driver at Corvus at Apple around 1987. I've got the CP/M driver on an 8" floppy

CP/M stuff is in bitsavers.org/bits/Corvus/Floppies/corvus1.tar

From a2corvus.cpp:

5 MB: -chs 144,4,20 -ss 512
10 MB: -chs 358,3,20 -ss 512
20 MB: -chs 388,5,20 -ss 512

Those were taken from Corvus technical documentation on Bitsavers.

Thanks. There is a bit of confusion around the hard disks used.

The CP/M 1 tools shown in the screenshot only mention the (larger, later) type H drives.

How did Corvus achieve 20 sectors on MFM drives, by the way?

B drives used IMI 6 meg native flat cable drives
H drives are IMI 5006, 5012 or 5018 which are ST-412 and a formatter board

it is clocked at 11MHz

see corvus_mfm_decoder.c in Dave Gesswein's MFM emulator source for details of the sector format

hm.. that doesn't look right. H drives had 1,2 or 3 platter drives in them. you're missing the 6mb H drive, they have 306 cyls, and 2, 4 or 6 heads.

-- ah the 10 and 20 are the 8" "A Drive" values

Board revision B detected
Found drive at select 1
Drive RPM 3596.0
Primary transition period 180 ns, should be around 200
Matches count 24 for controller Corvus_H
Header CRC: Polynomial 0x8005 length 16 initial value 0xffff
Number of heads 6 number of sectors 20 first sector 0
Interleave (not checked): 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Drive supports buffered seeks (ST412)

No sectors readable from cylinder 306
Stopping end of disk search due to two unreadable tracks in a row
Number of cylinders 306, 18.8 MB

Yes, 306 cyls with 2, 4 or 6 heads are consistent with what i took from one of the many PDFs


I wasn't quite sure, given the 20 sector oddity...

Now that we emulate even exotic hardware from ancient times, i'd like to tackle long standing bugs

1. all hard disk boot loaders i tried crash our emulation (including, but not limited to latest WUTIL 3.2.x and DSKPREP).

It is possible to generate boot diskettes with DSKPREP for both MS-DOS and CP/M, yet neither 'autoboot' (from BIOS) nor a manual boot (W key) works correctly.

2. the ZFLIP flag or another Z80 register is wrong after CTRL/Setup (= standard method to warm boot the Rainbow)

I got some clues why 1. happens, and will report later

I have added info about 6 and 20 MB Corvus drives (and how to format / use them) under CP/M 86/80 version 1.

Al Kossow kindly provided driver source for Corvus on Lifeboat CPM 2.2, but this requires - more or less - skillful modification to work with DEC's CP/M 86/80 2.x.

There are wo obstacles: a specific patch address must be determined (RWMLOC & RWMDTA in old source from Maslin), and LDCOPY seems to be hard coded in version 2.x (will not write arbitrary loader files as in CPM 1). See HINSTALL.SUB:

(script from *, modifies 3 core CP/M 1.x files)
Maybe one of the few remaining CP/M experts can have a look (Robert...)?
The TRS-80 variant uses the same type H drives, by the way...

http://www.retroarchive.org/maslin/disks/dec/
* http://www.retroarchive.org/maslin/disks/dec/drcdutil.zip
(Al Kossow's link, from Lifeboat CP/M 2.2):
bitsavers.org/bits/Corvus/Floppies/corvus1.tar

Venix disks for the Rainbow have recently surfaced. I've been able to read them all, and have a running Venix on my real Rainbow 100B. Will leverage that to make sense of the files that I was able to read on the Rainbow. More to follow shortly.

The more obscure Unix variants, the merrier