A SURVEY OF CP/M ON SHARP COMPUTERS
written by Maurice Hawes / SUC/UK
Source: SUC-magazine March 1993, Volume 13 Number 1, p. 13 – 16
In recent months, we have enrolled several new members because they have just acquired their first computers from boot sales in Oxfam shops, and the computers happened to be by Sharp. In many cases these members know nothing about the CP/M operating systems. Therefore, at the risk of teaching a few grandmothers how to stew eggs, I offer this survey of the CP/M operating system, as it has been applied to Z80-based Sharp Computers; in other words, to MZ-80K, MZ-80B, PC-3201, MZ-80A, MZ-700, MZ-3500 and MZ-800.
These days, if you ask a young computer salesman about CP/M, he will probably look at you as if you had come from Mars. Ten years ago, things were different; CP/M was the standard operating system for all Z80-based computers; in practice that meant every serious business micro you could name, except for the Apple. The Z-80 CPU with its limited address range of 64K, was king; the 8086 still in the laboratory, and MS-DOS had not even been launched.
The CP/M system was invented by Dr. Gary Kildall. It was intended for Intel 8080-based machines; but as it happened, it also suited the more powerful and popular Zilog Z80 CPU, and this eventual settled to CP/M‘s success. In a nutshell, CP/M is a software interface between programs and hardware, which makes all Z80 hardware ‘look’ the same; or at least, as nearly the same as can be managed.
CP/M demands a machine with disk drives and at least 20K of RAM ( which must start at $0000 ). And although there is no theoretical lower limit on VDU width, less than 64 columns is impracticable. Some of Sharp‘s Z80-based machines meet all the criteria, but some do not. Given their divergent designs, one cannot generalise; so I now turn to the individual machines, in chronological order.
MZ-80K ( 1979 )
The MZ-80K was designed as a direct competitor to the Commodore ‘PET‘, and Sharp probably never intended it as a serious business micro. As manufactured, it does not meet two of the criteria; its screen is only 40 columns, and it has a fixed ROM at $0000 – $0FFF.
Crystal Research of Torquay devised a ‘fix‘ for the ROM problem and wrote a 46K CP/M system for the MZ-80K which, initially, was designed for the MZ-80K‘s 40-column screen. But the package never really took off; this was probably due to the unhelpful attitude of Digital Research, the limited capacity of the MZ-80K‘s 35-track 5.25″ disks ( 143K ), and difficulties in upgrading the 40-column screen, rather than any deficiencies in Crystal‘s CP/M software.
The Sharp Users Club has done a lot of work in an attempt resuscitate the idea of CP/M on the MZ-80K, but we have recent concluded that the problems are almost insurmountable.
This does not mean that the ‘K‘ is no use. Other ‘K‘ DOS systems ( e.g. Brian Gladman‘s SDOS and ZEN DOS ) offer facilities similar to CP/M without demanding expensive and difficult hardware modifications and they turn a disk-based ‘K‘ into a useful software platform.
Note by the author of this site: My MZ-80K is able ( by hardware modifications ) to run CP/M Crystal Electronics version 2.21 40-columns ( 46K ), CP/M Crystal Electronics version 2.23 80-columns and its modified version for 80 tracks ( both 48K ).
MZ-80B ( 1980 )
Although it appeared only a year after the MZ-80K, the MZ-80B is very different, especially in the context of CP/M. As manufactured it has a switchable 40 / 80-column screen, 64K of RAM, and 35-track 5.25″ disk drives with a capacity around 300K. Therefore it does not have to be modified in any way, to run CP/M.
Nevertheless, it appears that Sharp did not see the MZ-80B as a business machine; at any rate, they didn‘t produce a CP/M system for it. The job was left to MicroTechnology Ltd.; and a very good job they did, too. They not only produced a reliable and friendly system, they also increased the capacity per disk, from Sharp‘s original figure of 286K, to 350K ( and that was done using Sharp‘s 35-track drives – on 40-track drives it would have been 400K ).
As a result, the MZ-80B was accepted as a business machine, and MicroTechnology went on to produce a hard disk system which gained many customers. They also implemented CP/M PLUS on the MZ-80B; but this was overtaken by events, and was never properly debugged.
PC-3201 ( 1980 )
Sharp made it clear that they intended the PC-3201 as a business machine, and they co-operated with Digital Research to produce a CP/M system for it; but it was such a cockeyed version that it was bound to fail. The PC-3201 failed the RAM criterion because it had a BASIC-in-ROM at $0000 – $4000, and Sharp ‘solved‘ this problem by creating an unorthodox CP/M system which started at $4200 instead of $0000, with programs loading at $4300. This was quite unlike anybody else‘s version of CP/M, and therefore virtually useless.
The proper solution to the problem came from MicroTechnology; they designed a ‘Relocator Board‘ which plugs into one of the I/O slots at the rear of the PC-3201. The BASIC-in-ROM chips are then removed from the main PCB and plugged into the ‘Relocator Board‘, after which they are automatically switched out of the main 64K area whenever the user loads the M-T version of PC-3201 CP/M.
M-T‘s version ‘looks and feels‘ identical to their MZ-80B CP/M; so anyone used to M-T CP/M will feel at home on either machine. But for some reason the two systems use different disk formats – 350K / disk on the MZ-80B, but only 256K / disk on the PC-3201. As it happens, this is not a serious problem today; MicroTechnology‘s CP/M system for the later Sharp MZ-3500 can handle both of these ( and some other ) ‘foreign‘ disk formats.
MZ-80A ( 1981 )
The MZ-80A, as manufactured, does not have an 80-column screen; but MicroTechnology ( and others ) soon discovered that it is easy to fit an 80-column option. M-T then modified their 64K MZ-80B version of CP/M, to form a 48K system on the MZ-80A; the two systems appear identical to the user, and they use the same disk format ( 35-track, 350K per disk ). But there are some important differences between the system files, see later in this article.
We can legally supply our members with Kuma‘s MZ-80A 80-column kit. We also offer an enhanced version which allows you to restore the MZ-80A to its original 40-column MZ-80A / MZ-80K configuration.
MZ-3500 ( 1982 )
The MZ-3500 is related to the PC-3201 and, like that machine, it was seen by Sharp as a business Computer. By this time, Sharp knew a bit about CP/M, and the MZ-3500 satisfies the main criteria; but once again, surprisingly, Sharp made a bit of a mess of writing a CP/M system for it. This is evidenced by the fact that there are several different versions of Sharp‘s MZ-3500 CP/M, and none of them is really satisfactory. Again, MicroTechnology came to the rescue, and produced a more standard, user-friendly, and versatile version of MZ-3500 CP/M. It looks and feels like M-Ts earlier CP/M‘s for Sharp computers, except that it has extra utilities which enable it to handle ‘foreign‘ CP/M disk formats, including MZ-80B / A and PC-3201. However, because it uses all 40 tracks, its own ‘native‘ disk format is different again – 320K per disk.
MZ-700 ( 1983 )
Neither Sharp nor MicroTechnology offered a version of CP/M for the MZ-700. The stumbling-block is the screen hardware, which is built around a customised LSI chip, and is meant for 40 columns.
Tim Cowell, a member of the Sharp Users Club, and the Sub-editor of the MZ-700 section in 1985 – 1988, designed a plug-in module which solved the problem by utilising the Z80‘s external ‘ports‘; Peterson Electronics showed considerable initial interest in this device, but for some reason they did not follow it up. Sharpsoft and Kuma also announced 80-column kits ( possibly the same kit in each case ), but nothing concrete ever seems to have appeared.
In 1989 the SUC took the first step towards a full-scale CP/M for the MZ-700, when member Dave Bagshaw modified the 48K MZ-80A version of CP/M to form a 52K system on the MZ-700, albeit in 40-columns only. This was a good start, enabling us to offer MZ-700 owners a disk version of our ever-popular WDPRO Word Processor.
Then, in 1990, John Edwards devised a very ingenious 80-column modification, involving a small additional PCB with flying leads and a few alterations to the main PCB. The SUC can supply the mod., in kit form for around £15 ( to members only ); once this is fitted, the MZ-700 screen may be switched into 80 columns.
Shortly afterwards John wrote a CP/M patch ( CPM780.COM ) which modifies Dave Bagshaw‘s 40-column CP/M to suit the 80-column hardware. In essence, the MZ-700 can now run almost any standard CP/M software; but to take full advantage of the 80-column screen you really do need a proper computer VDU, rather than a TV set.
The story does not end there. As reported in the 700 section of this issue, we now have an MZ-700 version of the 64K PCP/M system that Sharp marketed with MZ-800. By the time this issue appears, anyone who uses an 80-column MZ-700 with 40-track drives will be able to choose between TWO different CP/M Systems – one based on M-T‘s MZ-80A CP/M, and the other based on Sharp‘s MZ-800 PCP/M. The PCP/M disk format is 40-track, 320K ( but with some odd-ball features, see Vol.12 No.1 p.60 ); but PCP/M has a DISKDEF utility which allows Drive B: to read other Sharp-CP/M 35 / 40-track formats ( MZ-80B, MZ-3500, MZ-5500 ); so there is good inherent flexibility.
MZ-800 ( 1984 )
The MZ-800 was the last of Sharp‘s Z80 computers, and to go with it Sharp / D.R. introduced their 64K ‘personal CP/M‘ ( PCP/M ). This, with its ‘Visual Console Command Processor‘ ( VCCP ) and its ‘Status Line‘ display, was a belated attempt to revitalise CP/M. The VCCP presents a ‘facade‘ at which system operations may be performed by cursor-controlled selection from a list. In some cases this offers no advantage over the ‘command line‘; but in others it certainly does, and PCP/M itself also includes three VERY useful ‘extra‘ utilities which work similarly ( DISKDEF, DISKEDIT & SETUP ). The 320K disk format used by PCP/M is discussed under ‘MZ-700‘ above.
COMPATIBILITY and PROGRAM TRANSFER
Thanks to MicroTechnology Ltd., we have ‘look-alike‘ versions of CP/M for the MZ-80B, the PC-3201, the MZ-80A, the MZ-3500, and the MZ-700. And we also have Sharp‘s PCP/M on the MZ-800 and MZ-700.
There are points to watch. M-T‘s various CP/M‘s do not all use the same disk format, and PCP/M‘s disk format is different again. Having said that, there is no problem transferring CP/M software between machines. The ‘key‘ is in M-T CP/M on the MZ-3500; this can read / write M-T CP/M disks from the PC-3201 and MZ-80B / A / 700; and its own disks can be read / written by PCP/M.
DIFFERENCES IN SYSTEM FILES
Finally, a warning. The above-mentioned disk-transfer facilities allow you to transfer CP/M ‘applications‘ programs from one Sharp machine to another, with little fuss. For example, I now have my favourite W.P. ( no, no, they cry, not PEACHTEXT again !! ) running on ALL my Sharp computers except the ‘K‘. But you CANNOT transfer systems utilities between machines. So after any program-transfer session, you must ensure that all systems utilities files remain on their original machines. I fell for this recently – FORMAT.COM on the MZ-80B is the same size as FORMAT.COM on the MZ-80A / 700, ( it even comes up with the same version number, 1.3 ). But when I accidentally tried to FORMAT a disk on my MZ-700, using MZ-80B ‘FORMAT.COM‘, I kept getting ‘VERIFY ERROR‘. It took me an hour to find out why, and in the process I discovered that sometimes you can tell which version of a program you are in from its screen messages. For example, MZ-80B FORMAT.COM prints the message:
Format disk B, when ready type return to start
all on one line; but MZ-80A FORMAT.COM prints the same message on two lines. Having discovered this, I compared the corresponding CP/M systems files on the MZ-80B and the MZ-80A. The results were:
PIP, SUBMIT, XSUB, ED, ASM, DDT, LOAD, STAT and DUMP, SYSGEN, DEL, EJECT, and BUGFIX are all absolutely identical.
MOVCPM, BACKUP, CMT, CONSOLE, CONFIG, COPY, FILES, FORMAT, IODEFS, and TIME are different.
In the next issue, I shall discuss the individual files in M-T‘s various versions of CP/M, and in PCP/M, including those files which M-T or Sharp added to the standard Digital Research list.