INMC 80 News


September 1980 – January 1981, Issue 2

Page 10 of 59

If the C/B board is connected to the N1 with both clocks to the external clock pin, then the output clock will cause the receive clock to slave to its own frequency, and one might as well not have a phase locked loop at all. The clocks should be connected to the UART by a single pole double throw switch, which is thrown to the receive side to receive, and to the transmit side to transmit. Alternately, a bit of surgery on the board will allow a track to be cut which joint the two clock pins on the UART, and the correct leads attached to each pin. In my cases, I took all the connections I required to a connector I mounted on the N1 board, and made a connection through a multiple switch to allow either Nascom Tape Standard or 1200 CUTS to be selected. The reliability this modification (of seperate clocks) has given is unbelievable – I’ve almost forgotten about tape errors.

With regard to tape quality, I use Microdigital mostly, and for very large files BASF LH60 or FUJI FL30, with no problems at all. If available, use a frequency meter to set output clock from 555 to 76.8Hz, and if thought desireable, check the capture range of the PLL as described by Mr. Cottis in PCW December 1979.

Arising from my notes on implementing a Tiny Pascal, I have since optained from LP Enterprises a copy of the 8080 runtime support and P code to machine code converter written in Basic. These, bound together with a copy of the three original articles so that they form a complete unit, cost 11.95, and would form a good starting point for someone wishing to get a Pascal up and running. Also of interest are articles in the LSG first four issues on the subject of getting this language running on a O*mm*d*r* P*t (I don’t use dirty words, unlike Dr Dark). As my own machine has packed up, and is currently having a holiday in Nascom’s repair Dept, I have not been able to progress any further with the development of the Pascal. I have however discovered that there is available from Springer Verlag: “A Concurrent PASCAL Compiler for Minicomputers” by A.C. Hartman, which describes a seven pass compiler for Concurrent Pascal in detail, and from McGraw Hill: “ the BYTE Book of PASCAL”, with “two versions of a Pascal Compiler, one written in BASIC, and the other in 8080 assembly language”. I have both of these books on order, and will report on them when they arrive.

General Points:

Weller’s assembler has a nice feature – it can read source from a paper tape, and assemble it, which is a nice trick for very large programs. I must have a go at that. It could be implemented on Zen fairly easily.

Has anyone given any thought to the inelligance of designing for the Z80 as if it were an 8080 with a few extra instructions? In my view the Z80 is a very different machine to the 8080. For example, in the case of a Nascom, which is not running under interrupt control in general, the ancillary bank of registers is there and empty most of the time. Zeap uses this to good advantage. The indexing instructions are very rarely used, and yet they can very easily lead to a nice neat compact solution to a problem. The relative jumps can allow position independant code to be written, which can have advantages if you run out of memory on a big assembly. The Nascom BASIC seems to be an 8080 Basic with input/output patches all over it, which are concerned with deciding which monitor is available, and even emulating it (?). I have discussed the disassembly with a fellow Nascom fanatic (a trusted associate) and we are of the opinion that were it written in Z80 code from the beginning, it could be made about 20 o/o smaller and about 15 o/o faster. For example, it could on initialisation look to see which monitor it had available and set a flag which it would look at each time it wanted input or output, instead of doing all the tests it seems to do. As I am working towards a Pascal compiler, I am not inclined to have a go at rewriting it – I think I’ll save the 35.00 for the XTAL BASIC and put it towards the price of CP/M V2.2 and a floppy disc controller. Can anyone help with details on I.B.M. 33FD 8" floppies – I have two on indefinite loan, and would like to get them running, but have no further details other than that they are in working order. I understand them to be single sided single density, rather old-fashioned and outmoded really, but still 250K bytes on each !

This is an OCR’d version of the scanned page and likely contains recognition errors.

Page 10 of 59