A few days after the last issue went to press, John Edwards came to Weymouth for his customary Easter Break, and we spent a couple of days sorting out problems with 3.5" disks on the MZ-80K. I had previously told John, over the ‘phone, of my experiences with 3.5" 720K disks that I had recently bought new from Dixons in Weymouth, which gave no problems on the MZ-80B, MZ-80A or MZ-700 but became unreliable when used on the MZ-80K. So when he came down to Weymouth brought with him several small boxes of 3.5" 720K disks, some new and some used, of various different makes.

A long series of tests with these disks on my MZ-80K produced the same results as before i.e. disks seemed to format and copy O.K., albeit with seemingly minor hiccups which cause occasional ‘retries‘; but most of them later refused to boot up first time, and some would not boot up at all. We tried various tricks, such as increasing the number of ‘retries‘ allowed, but
nothing seemed to work, so we swapped the 3.5" 720K drives on my MZ-80K for a pair of 3.5" 720K drives that had been working satisfactorily on my MZ-80B for about 2 years. But the problems still persisted.

Completely frustrated, we decided to try some new 3.5" 1.44MB disks that I happened to have handy, also bought from Dixons in Weymouth. To our utter amazement the first disk out of the box formatted and copied with no signs of any problems, and re-booted consistently every time; and this remained the case after a large number of subsequent experiments with other disks from the same box. So we decided to stick to 3.5" 1.44MB disks, and from then on all our formatting, copying and re-booting problems disappeared.

We then turned to other problems. Some disk routines were crashing out because the drive heads were not being given enough time to move back from very high tracks; and the QUICKCOPY routine was returning the drive heads to HOME before every read or write operation. The first problem was solved by inserting a delay; but unfortunately the second problem did not respond to the type of solution we had already used successfully on the MZ-80B, MZ-80A and MZ-700, and we reluctantly bad to admit defeat on this point.

Nevertheless, we now had a version of SUPERCOPY which could format and copy reliable 3.5" disks for the MZ-80K, and all we had to do now was to decide how to utilise the extra storage space these disks provided.

The MZ-80K disk System differs in many ways from the disk System used on later MZ-machines such as the MZ-80B, MZ-80A and MZ-700. In this context the most important differences are as shown in the following Table:

The effects of these MZ-80K parameters are discussed on the next page.

The Theoretical Capacity Limits of various Disks on the MZ-80K ( p. 5 )

On the MZ-80K, an original Sharp 5.25 ‘35-Track‘ disk contains 70 x 16= 1120 sectors. Using similar calculations a ‘40-Track‘ disk contains 1280 sectors and a 3.5" ‘80-track‘ disk contains 2560 sectors. But on all these disks the first 4 Tracks ( 64 sectors ) are used by the Boot Program, FAT, and Directory; so the theoretical user capacities of 35-track, 40-track and
80-track disks are 1056, 1216 and 2496 sectors respectively.

On the other hand, whatever the size of the disk, the FAT always contains 256 bytes, of which 4 are used for system info ( Vol. No. & Sectors Used ). So there are always 252 x 8 = 2016 bits available to ‘map‘ used sectors.

Therefore the theoretical capacity of a 35- or 40-track disk can be fully utilised, because all the sectors can all be mapped in the FAT with plenty to spare. But the 2496 sectors on an 80-track disk cannot all be utilised because the FAT is limited to 2016 sectors.

Theoretically it might be possible to increase the size of the FAT to allow it to map all the sectors on a 3.5" disk; but this would involve software changes that would make the new 3.5" disks incompatible with the old 5.25" disks, and this was considered to be an impracticable step at this stage.

Setting the Actual Capacity of the new 3.5 MZ-80K disks

To provide a buffer between the FAT on Track 00 and the Directory on Track 01 it was decided to utilise only 2000 sectors ( rather than the theoretical maximum of 2016 ). In DISK BASIC SP-7082 the ‘Sectors Available‘ parameter is held at $4A45-$4A46 and on the new 3.5" disks it is set to $07D0 ( 2000 ).

SP-7082 also has to know the first prohibited Track No. This is held at $5304, and on these new disks this location is set to $81 ( 129 ); this ensures that any attempt to write to Track 129 gives a ‘Disk Full‘ error.

Summarising all this information:

The Capacities of the New 3.5" Disks expressed in Bytes

If you do DIR on a newly-formatted 3.5" disk, using a Basic that has been set up for 80-track disks as shown above, the DIR display should consist only of the header ‘VOL. xx ( S.2000 )‘, indicating 2000 free sectors. As each sector contains 128 bytes, this is equivalent to 256,000 bytes.

A 3.5" ‘Master‘ disk with no other files on it shows fewer than 2000 free sectors because the SP-7082 Disk Basic File occupies 160 sectors and these are mapped as ‘occupied‘ in the FAT. So if you do DIR on an empty 3.5" Basic Master Disk ( i.e. one from which all other files have been deleted ) the DIR display should consist only of the header ‘VOL.MASTER ( S.1840 )‘, indicating 1840 free sectors, equivalent to 235,520 bytes.

Both these figures are vast improvements over the original Sharp figures for their 70-track 5.25" disks ( Slave 135,168; Empty Master Disk 114,688 ) and we therefore feel that the exercise has been very much worth while.